chp04_aggregation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e3fdfa8e-aef2-4cc3-b562-8ee6bcf66f33",
   "metadata": {},
   "source": [
    "#### Top"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc7057d3-38ff-42bb-a82a-80d4eb1e9f18",
   "metadata": {},
   "source": [
    "# Chapter 4 Aggregation with Polars\n",
    "\n",
    "* [4.0 Imports and Setup](#4.0-Imports-and-Setup)\n",
    "* [4.1 Introduction](#4.1-Introduction)\n",
    "* [4.2 Loading the Data](#4.2-Loading-the-Data)\n",
    "* [4.3 Aggregations](#4.3-Aggregations)\n",
    "* [4.4 Lists](#4.4-Lists)\n",
    "  - [4.4.1 Using `.with_columns_seq`](#4.4.1-Using-.with_columns_seq)\n",
    "* [4.5 GroupBy Operations](#4.5-GroupBy-Operations)\n",
    "* [4.6 Multiple Aggregations](#4.6-Multiple-Aggregations)\n",
    "* [4.7 Grouping by Multiple Columns](#4.7-Grouping-by-Multiple-Columns)\n",
    "* [4.8 Pivoting with Multiple Aggregations](#4.8-Pivoting-with-Multiple-Aggregations)\n",
    "* [4.9 Aggregating over the Original Rows](#4.9-Aggregating-over-the-Original-Rows)\n",
    "* [4.10 Grouping to Lists](#4.10-Grouping-to-Lists)\n",
    "* [4.11 Chapter Methods and Functions](#4.11-Chapter-Methods-and-Functions)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "312acbea-ea37-4c7c-b1eb-4bbdfb1b6d75",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.0 Imports and Setup\n",
    "\n",
    "[back to Top](#Top)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "745bc82d-2cb6-4b12-a562-3f0b8ca20c29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "(function(root) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  var force = true;\n",
       "  var py_version = '3.4.3'.replace('rc', '-rc.').replace('.dev', '-dev.');\n",
       "  var reloading = false;\n",
       "  var Bokeh = root.Bokeh;\n",
       "\n",
       "  if (typeof (root._bokeh_timeout) === \"undefined\" || force) {\n",
       "    root._bokeh_timeout = Date.now() + 5000;\n",
       "    root._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  function run_callbacks() {\n",
       "    try {\n",
       "      root._bokeh_onload_callbacks.forEach(function(callback) {\n",
       "        if (callback != null)\n",
       "          callback();\n",
       "      });\n",
       "    } finally {\n",
       "      delete root._bokeh_onload_callbacks;\n",
       "    }\n",
       "    console.debug(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(css_urls, js_urls, js_modules, js_exports, callback) {\n",
       "    if (css_urls == null) css_urls = [];\n",
       "    if (js_urls == null) js_urls = [];\n",
       "    if (js_modules == null) js_modules = [];\n",
       "    if (js_exports == null) js_exports = {};\n",
       "\n",
       "    root._bokeh_onload_callbacks.push(callback);\n",
       "\n",
       "    if (root._bokeh_is_loading > 0) {\n",
       "      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls.length === 0 && js_modules.length === 0 && Object.keys(js_exports).length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    if (!reloading) {\n",
       "      console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "    }\n",
       "\n",
       "    function on_load() {\n",
       "      root._bokeh_is_loading--;\n",
       "      if (root._bokeh_is_loading === 0) {\n",
       "        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n",
       "        run_callbacks()\n",
       "      }\n",
       "    }\n",
       "    window._bokeh_on_load = on_load\n",
       "\n",
       "    function on_error() {\n",
       "      console.error(\"failed to load \" + url);\n",
       "    }\n",
       "\n",
       "    var skip = [];\n",
       "    if (window.requirejs) {\n",
       "      window.requirejs.config({'packages': {}, 'paths': {}, 'shim': {}});\n",
       "      root._bokeh_is_loading = css_urls.length + 0;\n",
       "    } else {\n",
       "      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length + Object.keys(js_exports).length;\n",
       "    }\n",
       "\n",
       "    var existing_stylesheets = []\n",
       "    var links = document.getElementsByTagName('link')\n",
       "    for (var i = 0; i < links.length; i++) {\n",
       "      var link = links[i]\n",
       "      if (link.href != null) {\n",
       "\texisting_stylesheets.push(link.href)\n",
       "      }\n",
       "    }\n",
       "    for (var i = 0; i < css_urls.length; i++) {\n",
       "      var url = css_urls[i];\n",
       "      if (existing_stylesheets.indexOf(url) !== -1) {\n",
       "\ton_load()\n",
       "\tcontinue;\n",
       "      }\n",
       "      const element = document.createElement(\"link\");\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.rel = \"stylesheet\";\n",
       "      element.type = \"text/css\";\n",
       "      element.href = url;\n",
       "      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n",
       "      document.body.appendChild(element);\n",
       "    }    var existing_scripts = []\n",
       "    var scripts = document.getElementsByTagName('script')\n",
       "    for (var i = 0; i < scripts.length; i++) {\n",
       "      var script = scripts[i]\n",
       "      if (script.src != null) {\n",
       "\texisting_scripts.push(script.src)\n",
       "      }\n",
       "    }\n",
       "    for (var i = 0; i < js_urls.length; i++) {\n",
       "      var url = js_urls[i];\n",
       "      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (var i = 0; i < js_modules.length; i++) {\n",
       "      var url = js_modules[i];\n",
       "      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (const name in js_exports) {\n",
       "      var url = js_exports[name];\n",
       "      if (skip.indexOf(url) >= 0 || root[name] != null) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      element.textContent = `\n",
       "      import ${name} from \"${url}\"\n",
       "      window.${name} = ${name}\n",
       "      window._bokeh_on_load()\n",
       "      `\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    if (!js_urls.length && !js_modules.length) {\n",
       "      on_load()\n",
       "    }\n",
       "  };\n",
       "\n",
       "  function inject_raw_css(css) {\n",
       "    const element = document.createElement(\"style\");\n",
       "    element.appendChild(document.createTextNode(css));\n",
       "    document.body.appendChild(element);\n",
       "  }\n",
       "\n",
       "  var js_urls = [\"https://cdn.bokeh.org/bokeh/release/bokeh-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-3.4.3.min.js\", \"https://cdn.holoviz.org/panel/1.4.5/dist/panel.min.js\"];\n",
       "  var js_modules = [];\n",
       "  var js_exports = {};\n",
       "  var css_urls = [];\n",
       "  var inline_js = [    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "function(Bokeh) {} // ensure no trailing comma for IE\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    if ((root.Bokeh !== undefined) || (force === true)) {\n",
       "      for (var i = 0; i < inline_js.length; i++) {\n",
       "\ttry {\n",
       "          inline_js[i].call(root, root.Bokeh);\n",
       "\t} catch(e) {\n",
       "\t  if (!reloading) {\n",
       "\t    throw e;\n",
       "\t  }\n",
       "\t}\n",
       "      }\n",
       "      // Cache old bokeh versions\n",
       "      if (Bokeh != undefined && !reloading) {\n",
       "\tvar NewBokeh = root.Bokeh;\n",
       "\tif (Bokeh.versions === undefined) {\n",
       "\t  Bokeh.versions = new Map();\n",
       "\t}\n",
       "\tif (NewBokeh.version !== Bokeh.version) {\n",
       "\t  Bokeh.versions.set(NewBokeh.version, NewBokeh)\n",
       "\t}\n",
       "\troot.Bokeh = Bokeh;\n",
       "      }} else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!root._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      root._bokeh_failed_load = true;\n",
       "    }\n",
       "    root._bokeh_is_initializing = false\n",
       "  }\n",
       "\n",
       "  function load_or_wait() {\n",
       "    // Implement a backoff loop that tries to ensure we do not load multiple\n",
       "    // versions of Bokeh and its dependencies at the same time.\n",
       "    // In recent versions we use the root._bokeh_is_initializing flag\n",
       "    // to determine whether there is an ongoing attempt to initialize\n",
       "    // bokeh, however for backward compatibility we also try to ensure\n",
       "    // that we do not start loading a newer (Panel>=1.0 and Bokeh>3) version\n",
       "    // before older versions are fully initialized.\n",
       "    if (root._bokeh_is_initializing && Date.now() > root._bokeh_timeout) {\n",
       "      root._bokeh_is_initializing = false;\n",
       "      root._bokeh_onload_callbacks = undefined;\n",
       "      console.log(\"Bokeh: BokehJS was loaded multiple times but one version failed to initialize.\");\n",
       "      load_or_wait();\n",
       "    } else if (root._bokeh_is_initializing || (typeof root._bokeh_is_initializing === \"undefined\" && root._bokeh_onload_callbacks !== undefined)) {\n",
       "      setTimeout(load_or_wait, 100);\n",
       "    } else {\n",
       "      root._bokeh_is_initializing = true\n",
       "      root._bokeh_onload_callbacks = []\n",
       "      var bokeh_loaded = Bokeh != null && (Bokeh.version === py_version || (Bokeh.versions !== undefined && Bokeh.versions.has(py_version)));\n",
       "      if (!reloading && !bokeh_loaded) {\n",
       "\troot.Bokeh = undefined;\n",
       "      }\n",
       "      load_libs(css_urls, js_urls, js_modules, js_exports, function() {\n",
       "\tconsole.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "\trun_inline_js();\n",
       "      });\n",
       "    }\n",
       "  }\n",
       "  // Give older versions of the autoload script a head-start to ensure\n",
       "  // they initialize before we start loading newer version.\n",
       "  setTimeout(load_or_wait, 100)\n",
       "}(window));"
      ],
      "application/vnd.holoviews_load.v0+json": "(function(root) {\n  function now() {\n    return new Date();\n  }\n\n  var force = true;\n  var py_version = '3.4.3'.replace('rc', '-rc.').replace('.dev', '-dev.');\n  var reloading = false;\n  var Bokeh = root.Bokeh;\n\n  if (typeof (root._bokeh_timeout) === \"undefined\" || force) {\n    root._bokeh_timeout = Date.now() + 5000;\n    root._bokeh_failed_load = false;\n  }\n\n  function run_callbacks() {\n    try {\n      root._bokeh_onload_callbacks.forEach(function(callback) {\n        if (callback != null)\n          callback();\n      });\n    } finally {\n      delete root._bokeh_onload_callbacks;\n    }\n    console.debug(\"Bokeh: all callbacks have finished\");\n  }\n\n  function load_libs(css_urls, js_urls, js_modules, js_exports, callback) {\n    if (css_urls == null) css_urls = [];\n    if (js_urls == null) js_urls = [];\n    if (js_modules == null) js_modules = [];\n    if (js_exports == null) js_exports = {};\n\n    root._bokeh_onload_callbacks.push(callback);\n\n    if (root._bokeh_is_loading > 0) {\n      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n      return null;\n    }\n    if (js_urls.length === 0 && js_modules.length === 0 && Object.keys(js_exports).length === 0) {\n      run_callbacks();\n      return null;\n    }\n    if (!reloading) {\n      console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n    }\n\n    function on_load() {\n      root._bokeh_is_loading--;\n      if (root._bokeh_is_loading === 0) {\n        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n        run_callbacks()\n      }\n    }\n    window._bokeh_on_load = on_load\n\n    function on_error() {\n      console.error(\"failed to load \" + url);\n    }\n\n    var skip = [];\n    if (window.requirejs) {\n      window.requirejs.config({'packages': {}, 'paths': {}, 'shim': {}});\n      root._bokeh_is_loading = css_urls.length + 0;\n    } else {\n      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length + Object.keys(js_exports).length;\n    }\n\n    var existing_stylesheets = []\n    var links = document.getElementsByTagName('link')\n    for (var i = 0; i < links.length; i++) {\n      var link = links[i]\n      if (link.href != null) {\n\texisting_stylesheets.push(link.href)\n      }\n    }\n    for (var i = 0; i < css_urls.length; i++) {\n      var url = css_urls[i];\n      if (existing_stylesheets.indexOf(url) !== -1) {\n\ton_load()\n\tcontinue;\n      }\n      const element = document.createElement(\"link\");\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.rel = \"stylesheet\";\n      element.type = \"text/css\";\n      element.href = url;\n      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n      document.body.appendChild(element);\n    }    var existing_scripts = []\n    var scripts = document.getElementsByTagName('script')\n    for (var i = 0; i < scripts.length; i++) {\n      var script = scripts[i]\n      if (script.src != null) {\n\texisting_scripts.push(script.src)\n      }\n    }\n    for (var i = 0; i < js_urls.length; i++) {\n      var url = js_urls[i];\n      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (var i = 0; i < js_modules.length; i++) {\n      var url = js_modules[i];\n      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (const name in js_exports) {\n      var url = js_exports[name];\n      if (skip.indexOf(url) >= 0 || root[name] != null) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onerror = on_error;\n      element.async = false;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      element.textContent = `\n      import ${name} from \"${url}\"\n      window.${name} = ${name}\n      window._bokeh_on_load()\n      `\n      document.head.appendChild(element);\n    }\n    if (!js_urls.length && !js_modules.length) {\n      on_load()\n    }\n  };\n\n  function inject_raw_css(css) {\n    const element = document.createElement(\"style\");\n    element.appendChild(document.createTextNode(css));\n    document.body.appendChild(element);\n  }\n\n  var js_urls = [\"https://cdn.bokeh.org/bokeh/release/bokeh-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-3.4.3.min.js\", \"https://cdn.holoviz.org/panel/1.4.5/dist/panel.min.js\"];\n  var js_modules = [];\n  var js_exports = {};\n  var css_urls = [];\n  var inline_js = [    function(Bokeh) {\n      Bokeh.set_log_level(\"info\");\n    },\nfunction(Bokeh) {} // ensure no trailing comma for IE\n  ];\n\n  function run_inline_js() {\n    if ((root.Bokeh !== undefined) || (force === true)) {\n      for (var i = 0; i < inline_js.length; i++) {\n\ttry {\n          inline_js[i].call(root, root.Bokeh);\n\t} catch(e) {\n\t  if (!reloading) {\n\t    throw e;\n\t  }\n\t}\n      }\n      // Cache old bokeh versions\n      if (Bokeh != undefined && !reloading) {\n\tvar NewBokeh = root.Bokeh;\n\tif (Bokeh.versions === undefined) {\n\t  Bokeh.versions = new Map();\n\t}\n\tif (NewBokeh.version !== Bokeh.version) {\n\t  Bokeh.versions.set(NewBokeh.version, NewBokeh)\n\t}\n\troot.Bokeh = Bokeh;\n      }} else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(run_inline_js, 100);\n    } else if (!root._bokeh_failed_load) {\n      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n      root._bokeh_failed_load = true;\n    }\n    root._bokeh_is_initializing = false\n  }\n\n  function load_or_wait() {\n    // Implement a backoff loop that tries to ensure we do not load multiple\n    // versions of Bokeh and its dependencies at the same time.\n    // In recent versions we use the root._bokeh_is_initializing flag\n    // to determine whether there is an ongoing attempt to initialize\n    // bokeh, however for backward compatibility we also try to ensure\n    // that we do not start loading a newer (Panel>=1.0 and Bokeh>3) version\n    // before older versions are fully initialized.\n    if (root._bokeh_is_initializing && Date.now() > root._bokeh_timeout) {\n      root._bokeh_is_initializing = false;\n      root._bokeh_onload_callbacks = undefined;\n      console.log(\"Bokeh: BokehJS was loaded multiple times but one version failed to initialize.\");\n      load_or_wait();\n    } else if (root._bokeh_is_initializing || (typeof root._bokeh_is_initializing === \"undefined\" && root._bokeh_onload_callbacks !== undefined)) {\n      setTimeout(load_or_wait, 100);\n    } else {\n      root._bokeh_is_initializing = true\n      root._bokeh_onload_callbacks = []\n      var bokeh_loaded = Bokeh != null && (Bokeh.version === py_version || (Bokeh.versions !== undefined && Bokeh.versions.has(py_version)));\n      if (!reloading && !bokeh_loaded) {\n\troot.Bokeh = undefined;\n      }\n      load_libs(css_urls, js_urls, js_modules, js_exports, function() {\n\tconsole.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n\trun_inline_js();\n      });\n    }\n  }\n  // Give older versions of the autoload script a head-start to ensure\n  // they initialize before we start loading newer version.\n  setTimeout(load_or_wait, 100)\n}(window));"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "if ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n",
       "  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n",
       "}\n",
       "\n",
       "\n",
       "    function JupyterCommManager() {\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n",
       "      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        comm_manager.register_target(comm_id, function(comm) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        });\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        });\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n",
       "          var messages = comm.messages[Symbol.asyncIterator]();\n",
       "          function processIteratorResult(result) {\n",
       "            var message = result.value;\n",
       "            console.log(message)\n",
       "            var content = {data: message.data, comm_id};\n",
       "            var buffers = []\n",
       "            for (var buffer of message.buffers || []) {\n",
       "              buffers.push(new DataView(buffer))\n",
       "            }\n",
       "            var metadata = message.metadata || {};\n",
       "            var msg = {content, buffers, metadata}\n",
       "            msg_handler(msg);\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "          return messages.next().then(processIteratorResult);\n",
       "        })\n",
       "      }\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n",
       "      if (comm_id in window.PyViz.comms) {\n",
       "        return window.PyViz.comms[comm_id];\n",
       "      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n",
       "        if (msg_handler) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        }\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n",
       "        comm.open();\n",
       "        if (msg_handler) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        }\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        var comm_promise = google.colab.kernel.comms.open(comm_id)\n",
       "        comm_promise.then((comm) => {\n",
       "          window.PyViz.comms[comm_id] = comm;\n",
       "          if (msg_handler) {\n",
       "            var messages = comm.messages[Symbol.asyncIterator]();\n",
       "            function processIteratorResult(result) {\n",
       "              var message = result.value;\n",
       "              var content = {data: message.data};\n",
       "              var metadata = message.metadata || {comm_id};\n",
       "              var msg = {content, metadata}\n",
       "              msg_handler(msg);\n",
       "              return messages.next().then(processIteratorResult);\n",
       "            }\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "        }) \n",
       "        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n",
       "          return comm_promise.then((comm) => {\n",
       "            comm.send(data, metadata, buffers, disposeOnDone);\n",
       "          });\n",
       "        };\n",
       "        var comm = {\n",
       "          send: sendClosure\n",
       "        };\n",
       "      }\n",
       "      window.PyViz.comms[comm_id] = comm;\n",
       "      return comm;\n",
       "    }\n",
       "    window.PyViz.comm_manager = new JupyterCommManager();\n",
       "    \n",
       "\n",
       "\n",
       "var JS_MIME_TYPE = 'application/javascript';\n",
       "var HTML_MIME_TYPE = 'text/html';\n",
       "var EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\n",
       "var CLASS_NAME = 'output';\n",
       "\n",
       "/**\n",
       " * Render data to the DOM node\n",
       " */\n",
       "function render(props, node) {\n",
       "  var div = document.createElement(\"div\");\n",
       "  var script = document.createElement(\"script\");\n",
       "  node.appendChild(div);\n",
       "  node.appendChild(script);\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when a new output is added\n",
       " */\n",
       "function handle_add_output(event, handle) {\n",
       "  var output_area = handle.output_area;\n",
       "  var output = handle.output;\n",
       "  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n",
       "    return\n",
       "  }\n",
       "  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n",
       "  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n",
       "  if (id !== undefined) {\n",
       "    var nchildren = toinsert.length;\n",
       "    var html_node = toinsert[nchildren-1].children[0];\n",
       "    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var scripts = [];\n",
       "    var nodelist = html_node.querySelectorAll(\"script\");\n",
       "    for (var i in nodelist) {\n",
       "      if (nodelist.hasOwnProperty(i)) {\n",
       "        scripts.push(nodelist[i])\n",
       "      }\n",
       "    }\n",
       "\n",
       "    scripts.forEach( function (oldScript) {\n",
       "      var newScript = document.createElement(\"script\");\n",
       "      var attrs = [];\n",
       "      var nodemap = oldScript.attributes;\n",
       "      for (var j in nodemap) {\n",
       "        if (nodemap.hasOwnProperty(j)) {\n",
       "          attrs.push(nodemap[j])\n",
       "        }\n",
       "      }\n",
       "      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n",
       "      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n",
       "      oldScript.parentNode.replaceChild(newScript, oldScript);\n",
       "    });\n",
       "    if (JS_MIME_TYPE in output.data) {\n",
       "      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n",
       "    }\n",
       "    output_area._hv_plot_id = id;\n",
       "    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n",
       "      window.PyViz.plot_index[id] = Bokeh.index[id];\n",
       "    } else {\n",
       "      window.PyViz.plot_index[id] = null;\n",
       "    }\n",
       "  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n",
       "    var bk_div = document.createElement(\"div\");\n",
       "    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var script_attrs = bk_div.children[0].attributes;\n",
       "    for (var i = 0; i < script_attrs.length; i++) {\n",
       "      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n",
       "    }\n",
       "    // store reference to server id on output_area\n",
       "    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when an output is cleared or removed\n",
       " */\n",
       "function handle_clear_output(event, handle) {\n",
       "  var id = handle.cell.output_area._hv_plot_id;\n",
       "  var server_id = handle.cell.output_area._bokeh_server_id;\n",
       "  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n",
       "  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n",
       "  if (server_id !== null) {\n",
       "    comm.send({event_type: 'server_delete', 'id': server_id});\n",
       "    return;\n",
       "  } else if (comm !== null) {\n",
       "    comm.send({event_type: 'delete', 'id': id});\n",
       "  }\n",
       "  delete PyViz.plot_index[id];\n",
       "  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n",
       "    var doc = window.Bokeh.index[id].model.document\n",
       "    doc.clear();\n",
       "    const i = window.Bokeh.documents.indexOf(doc);\n",
       "    if (i > -1) {\n",
       "      window.Bokeh.documents.splice(i, 1);\n",
       "    }\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle kernel restart event\n",
       " */\n",
       "function handle_kernel_cleanup(event, handle) {\n",
       "  delete PyViz.comms[\"hv-extension-comm\"];\n",
       "  window.PyViz.plot_index = {}\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle update_display_data messages\n",
       " */\n",
       "function handle_update_output(event, handle) {\n",
       "  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n",
       "  handle_add_output(event, handle)\n",
       "}\n",
       "\n",
       "function register_renderer(events, OutputArea) {\n",
       "  function append_mime(data, metadata, element) {\n",
       "    // create a DOM node to render to\n",
       "    var toinsert = this.create_output_subarea(\n",
       "    metadata,\n",
       "    CLASS_NAME,\n",
       "    EXEC_MIME_TYPE\n",
       "    );\n",
       "    this.keyboard_manager.register_events(toinsert);\n",
       "    // Render to node\n",
       "    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n",
       "    render(props, toinsert[0]);\n",
       "    element.append(toinsert);\n",
       "    return toinsert\n",
       "  }\n",
       "\n",
       "  events.on('output_added.OutputArea', handle_add_output);\n",
       "  events.on('output_updated.OutputArea', handle_update_output);\n",
       "  events.on('clear_output.CodeCell', handle_clear_output);\n",
       "  events.on('delete.Cell', handle_clear_output);\n",
       "  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n",
       "\n",
       "  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n",
       "    safe: true,\n",
       "    index: 0\n",
       "  });\n",
       "}\n",
       "\n",
       "if (window.Jupyter !== undefined) {\n",
       "  try {\n",
       "    var events = require('base/js/events');\n",
       "    var OutputArea = require('notebook/js/outputarea').OutputArea;\n",
       "    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n",
       "      register_renderer(events, OutputArea);\n",
       "    }\n",
       "  } catch(err) {\n",
       "  }\n",
       "}\n"
      ],
      "application/vnd.holoviews_load.v0+json": "\nif ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n}\n\n\n    function JupyterCommManager() {\n    }\n\n    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        comm_manager.register_target(comm_id, function(comm) {\n          comm.on_msg(msg_handler);\n        });\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n          comm.onMsg = msg_handler;\n        });\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n          var messages = comm.messages[Symbol.asyncIterator]();\n          function processIteratorResult(result) {\n            var message = result.value;\n            console.log(message)\n            var content = {data: message.data, comm_id};\n            var buffers = []\n            for (var buffer of message.buffers || []) {\n              buffers.push(new DataView(buffer))\n            }\n            var metadata = message.metadata || {};\n            var msg = {content, buffers, metadata}\n            msg_handler(msg);\n            return messages.next().then(processIteratorResult);\n          }\n          return messages.next().then(processIteratorResult);\n        })\n      }\n    }\n\n    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n      if (comm_id in window.PyViz.comms) {\n        return window.PyViz.comms[comm_id];\n      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n        if (msg_handler) {\n          comm.on_msg(msg_handler);\n        }\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n        comm.open();\n        if (msg_handler) {\n          comm.onMsg = msg_handler;\n        }\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        var comm_promise = google.colab.kernel.comms.open(comm_id)\n        comm_promise.then((comm) => {\n          window.PyViz.comms[comm_id] = comm;\n          if (msg_handler) {\n            var messages = comm.messages[Symbol.asyncIterator]();\n            function processIteratorResult(result) {\n              var message = result.value;\n              var content = {data: message.data};\n              var metadata = message.metadata || {comm_id};\n              var msg = {content, metadata}\n              msg_handler(msg);\n              return messages.next().then(processIteratorResult);\n            }\n            return messages.next().then(processIteratorResult);\n          }\n        }) \n        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n          return comm_promise.then((comm) => {\n            comm.send(data, metadata, buffers, disposeOnDone);\n          });\n        };\n        var comm = {\n          send: sendClosure\n        };\n      }\n      window.PyViz.comms[comm_id] = comm;\n      return comm;\n    }\n    window.PyViz.comm_manager = new JupyterCommManager();\n    \n\n\nvar JS_MIME_TYPE = 'application/javascript';\nvar HTML_MIME_TYPE = 'text/html';\nvar EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\nvar CLASS_NAME = 'output';\n\n/**\n * Render data to the DOM node\n */\nfunction render(props, node) {\n  var div = document.createElement(\"div\");\n  var script = document.createElement(\"script\");\n  node.appendChild(div);\n  node.appendChild(script);\n}\n\n/**\n * Handle when a new output is added\n */\nfunction handle_add_output(event, handle) {\n  var output_area = handle.output_area;\n  var output = handle.output;\n  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n    return\n  }\n  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n  if (id !== undefined) {\n    var nchildren = toinsert.length;\n    var html_node = toinsert[nchildren-1].children[0];\n    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n    var scripts = [];\n    var nodelist = html_node.querySelectorAll(\"script\");\n    for (var i in nodelist) {\n      if (nodelist.hasOwnProperty(i)) {\n        scripts.push(nodelist[i])\n      }\n    }\n\n    scripts.forEach( function (oldScript) {\n      var newScript = document.createElement(\"script\");\n      var attrs = [];\n      var nodemap = oldScript.attributes;\n      for (var j in nodemap) {\n        if (nodemap.hasOwnProperty(j)) {\n          attrs.push(nodemap[j])\n        }\n      }\n      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n      oldScript.parentNode.replaceChild(newScript, oldScript);\n    });\n    if (JS_MIME_TYPE in output.data) {\n      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n    }\n    output_area._hv_plot_id = id;\n    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n      window.PyViz.plot_index[id] = Bokeh.index[id];\n    } else {\n      window.PyViz.plot_index[id] = null;\n    }\n  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n    var bk_div = document.createElement(\"div\");\n    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n    var script_attrs = bk_div.children[0].attributes;\n    for (var i = 0; i < script_attrs.length; i++) {\n      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n    }\n    // store reference to server id on output_area\n    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n  }\n}\n\n/**\n * Handle when an output is cleared or removed\n */\nfunction handle_clear_output(event, handle) {\n  var id = handle.cell.output_area._hv_plot_id;\n  var server_id = handle.cell.output_area._bokeh_server_id;\n  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n  if (server_id !== null) {\n    comm.send({event_type: 'server_delete', 'id': server_id});\n    return;\n  } else if (comm !== null) {\n    comm.send({event_type: 'delete', 'id': id});\n  }\n  delete PyViz.plot_index[id];\n  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n    var doc = window.Bokeh.index[id].model.document\n    doc.clear();\n    const i = window.Bokeh.documents.indexOf(doc);\n    if (i > -1) {\n      window.Bokeh.documents.splice(i, 1);\n    }\n  }\n}\n\n/**\n * Handle kernel restart event\n */\nfunction handle_kernel_cleanup(event, handle) {\n  delete PyViz.comms[\"hv-extension-comm\"];\n  window.PyViz.plot_index = {}\n}\n\n/**\n * Handle update_display_data messages\n */\nfunction handle_update_output(event, handle) {\n  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n  handle_add_output(event, handle)\n}\n\nfunction register_renderer(events, OutputArea) {\n  function append_mime(data, metadata, element) {\n    // create a DOM node to render to\n    var toinsert = this.create_output_subarea(\n    metadata,\n    CLASS_NAME,\n    EXEC_MIME_TYPE\n    );\n    this.keyboard_manager.register_events(toinsert);\n    // Render to node\n    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n    render(props, toinsert[0]);\n    element.append(toinsert);\n    return toinsert\n  }\n\n  events.on('output_added.OutputArea', handle_add_output);\n  events.on('output_updated.OutputArea', handle_update_output);\n  events.on('clear_output.CodeCell', handle_clear_output);\n  events.on('delete.Cell', handle_clear_output);\n  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n\n  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n    safe: true,\n    index: 0\n  });\n}\n\nif (window.Jupyter !== undefined) {\n  try {\n    var events = require('base/js/events');\n    var OutputArea = require('notebook/js/outputarea').OutputArea;\n    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n      register_renderer(events, OutputArea);\n    }\n  } catch(err) {\n  }\n}\n"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>*[data-root-id],\n",
       "*[data-root-id] > * {\n",
       "  box-sizing: border-box;\n",
       "  font-family: var(--jp-ui-font-family);\n",
       "  font-size: var(--jp-ui-font-size1);\n",
       "  color: var(--vscode-editor-foreground, var(--jp-ui-font-color1));\n",
       "}\n",
       "\n",
       "/* Override VSCode background color */\n",
       ".cell-output-ipywidget-background:has(\n",
       "    > .cell-output-ipywidget-background > .lm-Widget > *[data-root-id]\n",
       "  ),\n",
       ".cell-output-ipywidget-background:has(> .lm-Widget > *[data-root-id]) {\n",
       "  background-color: transparent !important;\n",
       "}\n",
       "</style>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='p1002'>\n",
       "  <div id=\"d62748fd-73c1-432c-874d-7a83f31cabb7\" data-root-id=\"p1002\" style=\"display: contents;\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  var docs_json = {\"a55d0675-ce64-4ef7-8db3-9d83f50c5e4c\":{\"version\":\"3.4.3\",\"title\":\"Bokeh Application\",\"roots\":[{\"type\":\"object\",\"name\":\"panel.models.browser.BrowserInfo\",\"id\":\"p1002\"},{\"type\":\"object\",\"name\":\"panel.models.comm_manager.CommManager\",\"id\":\"p1003\",\"attributes\":{\"plot_id\":\"p1002\",\"comm_id\":\"7e188f97e1cd4740809ffe7b3133d11f\",\"client_comm_id\":\"20df12a617fc499e85365e40511fdcf0\"}}],\"defs\":[{\"type\":\"model\",\"name\":\"ReactiveHTML1\"},{\"type\":\"model\",\"name\":\"FlexBox1\",\"properties\":[{\"name\":\"align_content\",\"kind\":\"Any\",\"default\":\"flex-start\"},{\"name\":\"align_items\",\"kind\":\"Any\",\"default\":\"flex-start\"},{\"name\":\"flex_direction\",\"kind\":\"Any\",\"default\":\"row\"},{\"name\":\"flex_wrap\",\"kind\":\"Any\",\"default\":\"wrap\"},{\"name\":\"gap\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"justify_content\",\"kind\":\"Any\",\"default\":\"flex-start\"}]},{\"type\":\"model\",\"name\":\"FloatPanel1\",\"properties\":[{\"name\":\"config\",\"kind\":\"Any\",\"default\":{\"type\":\"map\"}},{\"name\":\"contained\",\"kind\":\"Any\",\"default\":true},{\"name\":\"position\",\"kind\":\"Any\",\"default\":\"right-top\"},{\"name\":\"offsetx\",\"kind\":\"Any\",\"default\":null},{\"name\":\"offsety\",\"kind\":\"Any\",\"default\":null},{\"name\":\"theme\",\"kind\":\"Any\",\"default\":\"primary\"},{\"name\":\"status\",\"kind\":\"Any\",\"default\":\"normalized\"}]},{\"type\":\"model\",\"name\":\"GridStack1\",\"properties\":[{\"name\":\"mode\",\"kind\":\"Any\",\"default\":\"warn\"},{\"name\":\"ncols\",\"kind\":\"Any\",\"default\":null},{\"name\":\"nrows\",\"kind\":\"Any\",\"default\":null},{\"name\":\"allow_resize\",\"kind\":\"Any\",\"default\":true},{\"name\":\"allow_drag\",\"kind\":\"Any\",\"default\":true},{\"name\":\"state\",\"kind\":\"Any\",\"default\":[]}]},{\"type\":\"model\",\"name\":\"drag1\",\"properties\":[{\"name\":\"slider_width\",\"kind\":\"Any\",\"default\":5},{\"name\":\"slider_color\",\"kind\":\"Any\",\"default\":\"black\"},{\"name\":\"value\",\"kind\":\"Any\",\"default\":50}]},{\"type\":\"model\",\"name\":\"click1\",\"properties\":[{\"name\":\"terminal_output\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"debug_name\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"clears\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"FastWrapper1\",\"properties\":[{\"name\":\"object\",\"kind\":\"Any\",\"default\":null},{\"name\":\"style\",\"kind\":\"Any\",\"default\":null}]},{\"type\":\"model\",\"name\":\"NotificationAreaBase1\",\"properties\":[{\"name\":\"js_events\",\"kind\":\"Any\",\"default\":{\"type\":\"map\"}},{\"name\":\"position\",\"kind\":\"Any\",\"default\":\"bottom-right\"},{\"name\":\"_clear\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"NotificationArea1\",\"properties\":[{\"name\":\"js_events\",\"kind\":\"Any\",\"default\":{\"type\":\"map\"}},{\"name\":\"notifications\",\"kind\":\"Any\",\"default\":[]},{\"name\":\"position\",\"kind\":\"Any\",\"default\":\"bottom-right\"},{\"name\":\"_clear\",\"kind\":\"Any\",\"default\":0},{\"name\":\"types\",\"kind\":\"Any\",\"default\":[{\"type\":\"map\",\"entries\":[[\"type\",\"warning\"],[\"background\",\"#ffc107\"],[\"icon\",{\"type\":\"map\",\"entries\":[[\"className\",\"fas fa-exclamation-triangle\"],[\"tagName\",\"i\"],[\"color\",\"white\"]]}]]},{\"type\":\"map\",\"entries\":[[\"type\",\"info\"],[\"background\",\"#007bff\"],[\"icon\",{\"type\":\"map\",\"entries\":[[\"className\",\"fas fa-info-circle\"],[\"tagName\",\"i\"],[\"color\",\"white\"]]}]]}]}]},{\"type\":\"model\",\"name\":\"Notification\",\"properties\":[{\"name\":\"background\",\"kind\":\"Any\",\"default\":null},{\"name\":\"duration\",\"kind\":\"Any\",\"default\":3000},{\"name\":\"icon\",\"kind\":\"Any\",\"default\":null},{\"name\":\"message\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"notification_type\",\"kind\":\"Any\",\"default\":null},{\"name\":\"_destroyed\",\"kind\":\"Any\",\"default\":false}]},{\"type\":\"model\",\"name\":\"TemplateActions1\",\"properties\":[{\"name\":\"open_modal\",\"kind\":\"Any\",\"default\":0},{\"name\":\"close_modal\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"BootstrapTemplateActions1\",\"properties\":[{\"name\":\"open_modal\",\"kind\":\"Any\",\"default\":0},{\"name\":\"close_modal\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"TemplateEditor1\",\"properties\":[{\"name\":\"layout\",\"kind\":\"Any\",\"default\":[]}]},{\"type\":\"model\",\"name\":\"MaterialTemplateActions1\",\"properties\":[{\"name\":\"open_modal\",\"kind\":\"Any\",\"default\":0},{\"name\":\"close_modal\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"copy_to_clipboard1\",\"properties\":[{\"name\":\"fill\",\"kind\":\"Any\",\"default\":\"none\"},{\"name\":\"value\",\"kind\":\"Any\",\"default\":null}]}]}};\n",
       "  var render_items = [{\"docid\":\"a55d0675-ce64-4ef7-8db3-9d83f50c5e4c\",\"roots\":{\"p1002\":\"d62748fd-73c1-432c-874d-7a83f31cabb7\"},\"root_ids\":[\"p1002\"]}];\n",
       "  var docs = Object.values(docs_json)\n",
       "  if (!docs) {\n",
       "    return\n",
       "  }\n",
       "  const py_version = docs[0].version.replace('rc', '-rc.').replace('.dev', '-dev.')\n",
       "  async function embed_document(root) {\n",
       "    var Bokeh = get_bokeh(root)\n",
       "    await Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && id_el.children[0].hasAttribute('data-root-id')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t  for (const child of root_el.children) {\n",
       "            // Ensure JupyterLab does not capture keyboard shortcuts\n",
       "            // see: https://jupyterlab.readthedocs.io/en/4.1.x/extension/notebook.html#keyboard-interaction-model\n",
       "\t    child.setAttribute('data-lm-suppress-shortcuts', 'true')\n",
       "\t  }\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  function get_bokeh(root) {\n",
       "    if (root.Bokeh === undefined) {\n",
       "      return null\n",
       "    } else if (root.Bokeh.version !== py_version) {\n",
       "      if (root.Bokeh.versions === undefined || !root.Bokeh.versions.has(py_version)) {\n",
       "\treturn null\n",
       "      }\n",
       "      return root.Bokeh.versions.get(py_version);\n",
       "    } else if (root.Bokeh.version === py_version) {\n",
       "      return root.Bokeh\n",
       "    }\n",
       "    return null\n",
       "  }\n",
       "  function is_loaded(root) {\n",
       "    var Bokeh = get_bokeh(root)\n",
       "    return (Bokeh != null && Bokeh.Panel !== undefined)\n",
       "  }\n",
       "  if (is_loaded(root)) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (is_loaded(root)) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "\t  var Bokeh = get_bokeh(root)\n",
       "\t  if (Bokeh == null || Bokeh.Panel == null) {\n",
       "            console.warn(\"Panel: ERROR: Unable to run Panel code because Bokeh or Panel library is missing\");\n",
       "\t  } else {\n",
       "\t    console.warn(\"Panel: WARNING: Attempting to render but not all required libraries could be resolved.\")\n",
       "\t    embed_document(root)\n",
       "\t  }\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ]
     },
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "p1002"
      }
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "(function(root) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  var force = true;\n",
       "  var py_version = '3.4.3'.replace('rc', '-rc.').replace('.dev', '-dev.');\n",
       "  var reloading = true;\n",
       "  var Bokeh = root.Bokeh;\n",
       "\n",
       "  if (typeof (root._bokeh_timeout) === \"undefined\" || force) {\n",
       "    root._bokeh_timeout = Date.now() + 5000;\n",
       "    root._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  function run_callbacks() {\n",
       "    try {\n",
       "      root._bokeh_onload_callbacks.forEach(function(callback) {\n",
       "        if (callback != null)\n",
       "          callback();\n",
       "      });\n",
       "    } finally {\n",
       "      delete root._bokeh_onload_callbacks;\n",
       "    }\n",
       "    console.debug(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(css_urls, js_urls, js_modules, js_exports, callback) {\n",
       "    if (css_urls == null) css_urls = [];\n",
       "    if (js_urls == null) js_urls = [];\n",
       "    if (js_modules == null) js_modules = [];\n",
       "    if (js_exports == null) js_exports = {};\n",
       "\n",
       "    root._bokeh_onload_callbacks.push(callback);\n",
       "\n",
       "    if (root._bokeh_is_loading > 0) {\n",
       "      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls.length === 0 && js_modules.length === 0 && Object.keys(js_exports).length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    if (!reloading) {\n",
       "      console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "    }\n",
       "\n",
       "    function on_load() {\n",
       "      root._bokeh_is_loading--;\n",
       "      if (root._bokeh_is_loading === 0) {\n",
       "        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n",
       "        run_callbacks()\n",
       "      }\n",
       "    }\n",
       "    window._bokeh_on_load = on_load\n",
       "\n",
       "    function on_error() {\n",
       "      console.error(\"failed to load \" + url);\n",
       "    }\n",
       "\n",
       "    var skip = [];\n",
       "    if (window.requirejs) {\n",
       "      window.requirejs.config({'packages': {}, 'paths': {}, 'shim': {}});\n",
       "      root._bokeh_is_loading = css_urls.length + 0;\n",
       "    } else {\n",
       "      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length + Object.keys(js_exports).length;\n",
       "    }\n",
       "\n",
       "    var existing_stylesheets = []\n",
       "    var links = document.getElementsByTagName('link')\n",
       "    for (var i = 0; i < links.length; i++) {\n",
       "      var link = links[i]\n",
       "      if (link.href != null) {\n",
       "\texisting_stylesheets.push(link.href)\n",
       "      }\n",
       "    }\n",
       "    for (var i = 0; i < css_urls.length; i++) {\n",
       "      var url = css_urls[i];\n",
       "      if (existing_stylesheets.indexOf(url) !== -1) {\n",
       "\ton_load()\n",
       "\tcontinue;\n",
       "      }\n",
       "      const element = document.createElement(\"link\");\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.rel = \"stylesheet\";\n",
       "      element.type = \"text/css\";\n",
       "      element.href = url;\n",
       "      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n",
       "      document.body.appendChild(element);\n",
       "    }    var existing_scripts = []\n",
       "    var scripts = document.getElementsByTagName('script')\n",
       "    for (var i = 0; i < scripts.length; i++) {\n",
       "      var script = scripts[i]\n",
       "      if (script.src != null) {\n",
       "\texisting_scripts.push(script.src)\n",
       "      }\n",
       "    }\n",
       "    for (var i = 0; i < js_urls.length; i++) {\n",
       "      var url = js_urls[i];\n",
       "      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (var i = 0; i < js_modules.length; i++) {\n",
       "      var url = js_modules[i];\n",
       "      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (const name in js_exports) {\n",
       "      var url = js_exports[name];\n",
       "      if (skip.indexOf(url) >= 0 || root[name] != null) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      element.textContent = `\n",
       "      import ${name} from \"${url}\"\n",
       "      window.${name} = ${name}\n",
       "      window._bokeh_on_load()\n",
       "      `\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    if (!js_urls.length && !js_modules.length) {\n",
       "      on_load()\n",
       "    }\n",
       "  };\n",
       "\n",
       "  function inject_raw_css(css) {\n",
       "    const element = document.createElement(\"style\");\n",
       "    element.appendChild(document.createTextNode(css));\n",
       "    document.body.appendChild(element);\n",
       "  }\n",
       "\n",
       "  var js_urls = [];\n",
       "  var js_modules = [];\n",
       "  var js_exports = {};\n",
       "  var css_urls = [];\n",
       "  var inline_js = [    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "function(Bokeh) {} // ensure no trailing comma for IE\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    if ((root.Bokeh !== undefined) || (force === true)) {\n",
       "      for (var i = 0; i < inline_js.length; i++) {\n",
       "\ttry {\n",
       "          inline_js[i].call(root, root.Bokeh);\n",
       "\t} catch(e) {\n",
       "\t  if (!reloading) {\n",
       "\t    throw e;\n",
       "\t  }\n",
       "\t}\n",
       "      }\n",
       "      // Cache old bokeh versions\n",
       "      if (Bokeh != undefined && !reloading) {\n",
       "\tvar NewBokeh = root.Bokeh;\n",
       "\tif (Bokeh.versions === undefined) {\n",
       "\t  Bokeh.versions = new Map();\n",
       "\t}\n",
       "\tif (NewBokeh.version !== Bokeh.version) {\n",
       "\t  Bokeh.versions.set(NewBokeh.version, NewBokeh)\n",
       "\t}\n",
       "\troot.Bokeh = Bokeh;\n",
       "      }} else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!root._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      root._bokeh_failed_load = true;\n",
       "    }\n",
       "    root._bokeh_is_initializing = false\n",
       "  }\n",
       "\n",
       "  function load_or_wait() {\n",
       "    // Implement a backoff loop that tries to ensure we do not load multiple\n",
       "    // versions of Bokeh and its dependencies at the same time.\n",
       "    // In recent versions we use the root._bokeh_is_initializing flag\n",
       "    // to determine whether there is an ongoing attempt to initialize\n",
       "    // bokeh, however for backward compatibility we also try to ensure\n",
       "    // that we do not start loading a newer (Panel>=1.0 and Bokeh>3) version\n",
       "    // before older versions are fully initialized.\n",
       "    if (root._bokeh_is_initializing && Date.now() > root._bokeh_timeout) {\n",
       "      root._bokeh_is_initializing = false;\n",
       "      root._bokeh_onload_callbacks = undefined;\n",
       "      console.log(\"Bokeh: BokehJS was loaded multiple times but one version failed to initialize.\");\n",
       "      load_or_wait();\n",
       "    } else if (root._bokeh_is_initializing || (typeof root._bokeh_is_initializing === \"undefined\" && root._bokeh_onload_callbacks !== undefined)) {\n",
       "      setTimeout(load_or_wait, 100);\n",
       "    } else {\n",
       "      root._bokeh_is_initializing = true\n",
       "      root._bokeh_onload_callbacks = []\n",
       "      var bokeh_loaded = Bokeh != null && (Bokeh.version === py_version || (Bokeh.versions !== undefined && Bokeh.versions.has(py_version)));\n",
       "      if (!reloading && !bokeh_loaded) {\n",
       "\troot.Bokeh = undefined;\n",
       "      }\n",
       "      load_libs(css_urls, js_urls, js_modules, js_exports, function() {\n",
       "\tconsole.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "\trun_inline_js();\n",
       "      });\n",
       "    }\n",
       "  }\n",
       "  // Give older versions of the autoload script a head-start to ensure\n",
       "  // they initialize before we start loading newer version.\n",
       "  setTimeout(load_or_wait, 100)\n",
       "}(window));"
      ],
      "application/vnd.holoviews_load.v0+json": "(function(root) {\n  function now() {\n    return new Date();\n  }\n\n  var force = true;\n  var py_version = '3.4.3'.replace('rc', '-rc.').replace('.dev', '-dev.');\n  var reloading = true;\n  var Bokeh = root.Bokeh;\n\n  if (typeof (root._bokeh_timeout) === \"undefined\" || force) {\n    root._bokeh_timeout = Date.now() + 5000;\n    root._bokeh_failed_load = false;\n  }\n\n  function run_callbacks() {\n    try {\n      root._bokeh_onload_callbacks.forEach(function(callback) {\n        if (callback != null)\n          callback();\n      });\n    } finally {\n      delete root._bokeh_onload_callbacks;\n    }\n    console.debug(\"Bokeh: all callbacks have finished\");\n  }\n\n  function load_libs(css_urls, js_urls, js_modules, js_exports, callback) {\n    if (css_urls == null) css_urls = [];\n    if (js_urls == null) js_urls = [];\n    if (js_modules == null) js_modules = [];\n    if (js_exports == null) js_exports = {};\n\n    root._bokeh_onload_callbacks.push(callback);\n\n    if (root._bokeh_is_loading > 0) {\n      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n      return null;\n    }\n    if (js_urls.length === 0 && js_modules.length === 0 && Object.keys(js_exports).length === 0) {\n      run_callbacks();\n      return null;\n    }\n    if (!reloading) {\n      console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n    }\n\n    function on_load() {\n      root._bokeh_is_loading--;\n      if (root._bokeh_is_loading === 0) {\n        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n        run_callbacks()\n      }\n    }\n    window._bokeh_on_load = on_load\n\n    function on_error() {\n      console.error(\"failed to load \" + url);\n    }\n\n    var skip = [];\n    if (window.requirejs) {\n      window.requirejs.config({'packages': {}, 'paths': {}, 'shim': {}});\n      root._bokeh_is_loading = css_urls.length + 0;\n    } else {\n      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length + Object.keys(js_exports).length;\n    }\n\n    var existing_stylesheets = []\n    var links = document.getElementsByTagName('link')\n    for (var i = 0; i < links.length; i++) {\n      var link = links[i]\n      if (link.href != null) {\n\texisting_stylesheets.push(link.href)\n      }\n    }\n    for (var i = 0; i < css_urls.length; i++) {\n      var url = css_urls[i];\n      if (existing_stylesheets.indexOf(url) !== -1) {\n\ton_load()\n\tcontinue;\n      }\n      const element = document.createElement(\"link\");\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.rel = \"stylesheet\";\n      element.type = \"text/css\";\n      element.href = url;\n      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n      document.body.appendChild(element);\n    }    var existing_scripts = []\n    var scripts = document.getElementsByTagName('script')\n    for (var i = 0; i < scripts.length; i++) {\n      var script = scripts[i]\n      if (script.src != null) {\n\texisting_scripts.push(script.src)\n      }\n    }\n    for (var i = 0; i < js_urls.length; i++) {\n      var url = js_urls[i];\n      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (var i = 0; i < js_modules.length; i++) {\n      var url = js_modules[i];\n      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (const name in js_exports) {\n      var url = js_exports[name];\n      if (skip.indexOf(url) >= 0 || root[name] != null) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onerror = on_error;\n      element.async = false;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      element.textContent = `\n      import ${name} from \"${url}\"\n      window.${name} = ${name}\n      window._bokeh_on_load()\n      `\n      document.head.appendChild(element);\n    }\n    if (!js_urls.length && !js_modules.length) {\n      on_load()\n    }\n  };\n\n  function inject_raw_css(css) {\n    const element = document.createElement(\"style\");\n    element.appendChild(document.createTextNode(css));\n    document.body.appendChild(element);\n  }\n\n  var js_urls = [];\n  var js_modules = [];\n  var js_exports = {};\n  var css_urls = [];\n  var inline_js = [    function(Bokeh) {\n      Bokeh.set_log_level(\"info\");\n    },\nfunction(Bokeh) {} // ensure no trailing comma for IE\n  ];\n\n  function run_inline_js() {\n    if ((root.Bokeh !== undefined) || (force === true)) {\n      for (var i = 0; i < inline_js.length; i++) {\n\ttry {\n          inline_js[i].call(root, root.Bokeh);\n\t} catch(e) {\n\t  if (!reloading) {\n\t    throw e;\n\t  }\n\t}\n      }\n      // Cache old bokeh versions\n      if (Bokeh != undefined && !reloading) {\n\tvar NewBokeh = root.Bokeh;\n\tif (Bokeh.versions === undefined) {\n\t  Bokeh.versions = new Map();\n\t}\n\tif (NewBokeh.version !== Bokeh.version) {\n\t  Bokeh.versions.set(NewBokeh.version, NewBokeh)\n\t}\n\troot.Bokeh = Bokeh;\n      }} else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(run_inline_js, 100);\n    } else if (!root._bokeh_failed_load) {\n      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n      root._bokeh_failed_load = true;\n    }\n    root._bokeh_is_initializing = false\n  }\n\n  function load_or_wait() {\n    // Implement a backoff loop that tries to ensure we do not load multiple\n    // versions of Bokeh and its dependencies at the same time.\n    // In recent versions we use the root._bokeh_is_initializing flag\n    // to determine whether there is an ongoing attempt to initialize\n    // bokeh, however for backward compatibility we also try to ensure\n    // that we do not start loading a newer (Panel>=1.0 and Bokeh>3) version\n    // before older versions are fully initialized.\n    if (root._bokeh_is_initializing && Date.now() > root._bokeh_timeout) {\n      root._bokeh_is_initializing = false;\n      root._bokeh_onload_callbacks = undefined;\n      console.log(\"Bokeh: BokehJS was loaded multiple times but one version failed to initialize.\");\n      load_or_wait();\n    } else if (root._bokeh_is_initializing || (typeof root._bokeh_is_initializing === \"undefined\" && root._bokeh_onload_callbacks !== undefined)) {\n      setTimeout(load_or_wait, 100);\n    } else {\n      root._bokeh_is_initializing = true\n      root._bokeh_onload_callbacks = []\n      var bokeh_loaded = Bokeh != null && (Bokeh.version === py_version || (Bokeh.versions !== undefined && Bokeh.versions.has(py_version)));\n      if (!reloading && !bokeh_loaded) {\n\troot.Bokeh = undefined;\n      }\n      load_libs(css_urls, js_urls, js_modules, js_exports, function() {\n\tconsole.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n\trun_inline_js();\n      });\n    }\n  }\n  // Give older versions of the autoload script a head-start to ensure\n  // they initialize before we start loading newer version.\n  setTimeout(load_or_wait, 100)\n}(window));"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "if ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n",
       "  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n",
       "}\n",
       "\n",
       "\n",
       "    function JupyterCommManager() {\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n",
       "      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        comm_manager.register_target(comm_id, function(comm) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        });\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        });\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n",
       "          var messages = comm.messages[Symbol.asyncIterator]();\n",
       "          function processIteratorResult(result) {\n",
       "            var message = result.value;\n",
       "            console.log(message)\n",
       "            var content = {data: message.data, comm_id};\n",
       "            var buffers = []\n",
       "            for (var buffer of message.buffers || []) {\n",
       "              buffers.push(new DataView(buffer))\n",
       "            }\n",
       "            var metadata = message.metadata || {};\n",
       "            var msg = {content, buffers, metadata}\n",
       "            msg_handler(msg);\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "          return messages.next().then(processIteratorResult);\n",
       "        })\n",
       "      }\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n",
       "      if (comm_id in window.PyViz.comms) {\n",
       "        return window.PyViz.comms[comm_id];\n",
       "      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n",
       "        if (msg_handler) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        }\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n",
       "        comm.open();\n",
       "        if (msg_handler) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        }\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        var comm_promise = google.colab.kernel.comms.open(comm_id)\n",
       "        comm_promise.then((comm) => {\n",
       "          window.PyViz.comms[comm_id] = comm;\n",
       "          if (msg_handler) {\n",
       "            var messages = comm.messages[Symbol.asyncIterator]();\n",
       "            function processIteratorResult(result) {\n",
       "              var message = result.value;\n",
       "              var content = {data: message.data};\n",
       "              var metadata = message.metadata || {comm_id};\n",
       "              var msg = {content, metadata}\n",
       "              msg_handler(msg);\n",
       "              return messages.next().then(processIteratorResult);\n",
       "            }\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "        }) \n",
       "        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n",
       "          return comm_promise.then((comm) => {\n",
       "            comm.send(data, metadata, buffers, disposeOnDone);\n",
       "          });\n",
       "        };\n",
       "        var comm = {\n",
       "          send: sendClosure\n",
       "        };\n",
       "      }\n",
       "      window.PyViz.comms[comm_id] = comm;\n",
       "      return comm;\n",
       "    }\n",
       "    window.PyViz.comm_manager = new JupyterCommManager();\n",
       "    \n",
       "\n",
       "\n",
       "var JS_MIME_TYPE = 'application/javascript';\n",
       "var HTML_MIME_TYPE = 'text/html';\n",
       "var EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\n",
       "var CLASS_NAME = 'output';\n",
       "\n",
       "/**\n",
       " * Render data to the DOM node\n",
       " */\n",
       "function render(props, node) {\n",
       "  var div = document.createElement(\"div\");\n",
       "  var script = document.createElement(\"script\");\n",
       "  node.appendChild(div);\n",
       "  node.appendChild(script);\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when a new output is added\n",
       " */\n",
       "function handle_add_output(event, handle) {\n",
       "  var output_area = handle.output_area;\n",
       "  var output = handle.output;\n",
       "  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n",
       "    return\n",
       "  }\n",
       "  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n",
       "  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n",
       "  if (id !== undefined) {\n",
       "    var nchildren = toinsert.length;\n",
       "    var html_node = toinsert[nchildren-1].children[0];\n",
       "    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var scripts = [];\n",
       "    var nodelist = html_node.querySelectorAll(\"script\");\n",
       "    for (var i in nodelist) {\n",
       "      if (nodelist.hasOwnProperty(i)) {\n",
       "        scripts.push(nodelist[i])\n",
       "      }\n",
       "    }\n",
       "\n",
       "    scripts.forEach( function (oldScript) {\n",
       "      var newScript = document.createElement(\"script\");\n",
       "      var attrs = [];\n",
       "      var nodemap = oldScript.attributes;\n",
       "      for (var j in nodemap) {\n",
       "        if (nodemap.hasOwnProperty(j)) {\n",
       "          attrs.push(nodemap[j])\n",
       "        }\n",
       "      }\n",
       "      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n",
       "      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n",
       "      oldScript.parentNode.replaceChild(newScript, oldScript);\n",
       "    });\n",
       "    if (JS_MIME_TYPE in output.data) {\n",
       "      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n",
       "    }\n",
       "    output_area._hv_plot_id = id;\n",
       "    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n",
       "      window.PyViz.plot_index[id] = Bokeh.index[id];\n",
       "    } else {\n",
       "      window.PyViz.plot_index[id] = null;\n",
       "    }\n",
       "  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n",
       "    var bk_div = document.createElement(\"div\");\n",
       "    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var script_attrs = bk_div.children[0].attributes;\n",
       "    for (var i = 0; i < script_attrs.length; i++) {\n",
       "      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n",
       "    }\n",
       "    // store reference to server id on output_area\n",
       "    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when an output is cleared or removed\n",
       " */\n",
       "function handle_clear_output(event, handle) {\n",
       "  var id = handle.cell.output_area._hv_plot_id;\n",
       "  var server_id = handle.cell.output_area._bokeh_server_id;\n",
       "  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n",
       "  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n",
       "  if (server_id !== null) {\n",
       "    comm.send({event_type: 'server_delete', 'id': server_id});\n",
       "    return;\n",
       "  } else if (comm !== null) {\n",
       "    comm.send({event_type: 'delete', 'id': id});\n",
       "  }\n",
       "  delete PyViz.plot_index[id];\n",
       "  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n",
       "    var doc = window.Bokeh.index[id].model.document\n",
       "    doc.clear();\n",
       "    const i = window.Bokeh.documents.indexOf(doc);\n",
       "    if (i > -1) {\n",
       "      window.Bokeh.documents.splice(i, 1);\n",
       "    }\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle kernel restart event\n",
       " */\n",
       "function handle_kernel_cleanup(event, handle) {\n",
       "  delete PyViz.comms[\"hv-extension-comm\"];\n",
       "  window.PyViz.plot_index = {}\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle update_display_data messages\n",
       " */\n",
       "function handle_update_output(event, handle) {\n",
       "  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n",
       "  handle_add_output(event, handle)\n",
       "}\n",
       "\n",
       "function register_renderer(events, OutputArea) {\n",
       "  function append_mime(data, metadata, element) {\n",
       "    // create a DOM node to render to\n",
       "    var toinsert = this.create_output_subarea(\n",
       "    metadata,\n",
       "    CLASS_NAME,\n",
       "    EXEC_MIME_TYPE\n",
       "    );\n",
       "    this.keyboard_manager.register_events(toinsert);\n",
       "    // Render to node\n",
       "    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n",
       "    render(props, toinsert[0]);\n",
       "    element.append(toinsert);\n",
       "    return toinsert\n",
       "  }\n",
       "\n",
       "  events.on('output_added.OutputArea', handle_add_output);\n",
       "  events.on('output_updated.OutputArea', handle_update_output);\n",
       "  events.on('clear_output.CodeCell', handle_clear_output);\n",
       "  events.on('delete.Cell', handle_clear_output);\n",
       "  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n",
       "\n",
       "  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n",
       "    safe: true,\n",
       "    index: 0\n",
       "  });\n",
       "}\n",
       "\n",
       "if (window.Jupyter !== undefined) {\n",
       "  try {\n",
       "    var events = require('base/js/events');\n",
       "    var OutputArea = require('notebook/js/outputarea').OutputArea;\n",
       "    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n",
       "      register_renderer(events, OutputArea);\n",
       "    }\n",
       "  } catch(err) {\n",
       "  }\n",
       "}\n"
      ],
      "application/vnd.holoviews_load.v0+json": "\nif ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n}\n\n\n    function JupyterCommManager() {\n    }\n\n    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        comm_manager.register_target(comm_id, function(comm) {\n          comm.on_msg(msg_handler);\n        });\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n          comm.onMsg = msg_handler;\n        });\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n          var messages = comm.messages[Symbol.asyncIterator]();\n          function processIteratorResult(result) {\n            var message = result.value;\n            console.log(message)\n            var content = {data: message.data, comm_id};\n            var buffers = []\n            for (var buffer of message.buffers || []) {\n              buffers.push(new DataView(buffer))\n            }\n            var metadata = message.metadata || {};\n            var msg = {content, buffers, metadata}\n            msg_handler(msg);\n            return messages.next().then(processIteratorResult);\n          }\n          return messages.next().then(processIteratorResult);\n        })\n      }\n    }\n\n    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n      if (comm_id in window.PyViz.comms) {\n        return window.PyViz.comms[comm_id];\n      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n        if (msg_handler) {\n          comm.on_msg(msg_handler);\n        }\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n        comm.open();\n        if (msg_handler) {\n          comm.onMsg = msg_handler;\n        }\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        var comm_promise = google.colab.kernel.comms.open(comm_id)\n        comm_promise.then((comm) => {\n          window.PyViz.comms[comm_id] = comm;\n          if (msg_handler) {\n            var messages = comm.messages[Symbol.asyncIterator]();\n            function processIteratorResult(result) {\n              var message = result.value;\n              var content = {data: message.data};\n              var metadata = message.metadata || {comm_id};\n              var msg = {content, metadata}\n              msg_handler(msg);\n              return messages.next().then(processIteratorResult);\n            }\n            return messages.next().then(processIteratorResult);\n          }\n        }) \n        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n          return comm_promise.then((comm) => {\n            comm.send(data, metadata, buffers, disposeOnDone);\n          });\n        };\n        var comm = {\n          send: sendClosure\n        };\n      }\n      window.PyViz.comms[comm_id] = comm;\n      return comm;\n    }\n    window.PyViz.comm_manager = new JupyterCommManager();\n    \n\n\nvar JS_MIME_TYPE = 'application/javascript';\nvar HTML_MIME_TYPE = 'text/html';\nvar EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\nvar CLASS_NAME = 'output';\n\n/**\n * Render data to the DOM node\n */\nfunction render(props, node) {\n  var div = document.createElement(\"div\");\n  var script = document.createElement(\"script\");\n  node.appendChild(div);\n  node.appendChild(script);\n}\n\n/**\n * Handle when a new output is added\n */\nfunction handle_add_output(event, handle) {\n  var output_area = handle.output_area;\n  var output = handle.output;\n  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n    return\n  }\n  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n  if (id !== undefined) {\n    var nchildren = toinsert.length;\n    var html_node = toinsert[nchildren-1].children[0];\n    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n    var scripts = [];\n    var nodelist = html_node.querySelectorAll(\"script\");\n    for (var i in nodelist) {\n      if (nodelist.hasOwnProperty(i)) {\n        scripts.push(nodelist[i])\n      }\n    }\n\n    scripts.forEach( function (oldScript) {\n      var newScript = document.createElement(\"script\");\n      var attrs = [];\n      var nodemap = oldScript.attributes;\n      for (var j in nodemap) {\n        if (nodemap.hasOwnProperty(j)) {\n          attrs.push(nodemap[j])\n        }\n      }\n      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n      oldScript.parentNode.replaceChild(newScript, oldScript);\n    });\n    if (JS_MIME_TYPE in output.data) {\n      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n    }\n    output_area._hv_plot_id = id;\n    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n      window.PyViz.plot_index[id] = Bokeh.index[id];\n    } else {\n      window.PyViz.plot_index[id] = null;\n    }\n  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n    var bk_div = document.createElement(\"div\");\n    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n    var script_attrs = bk_div.children[0].attributes;\n    for (var i = 0; i < script_attrs.length; i++) {\n      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n    }\n    // store reference to server id on output_area\n    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n  }\n}\n\n/**\n * Handle when an output is cleared or removed\n */\nfunction handle_clear_output(event, handle) {\n  var id = handle.cell.output_area._hv_plot_id;\n  var server_id = handle.cell.output_area._bokeh_server_id;\n  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n  if (server_id !== null) {\n    comm.send({event_type: 'server_delete', 'id': server_id});\n    return;\n  } else if (comm !== null) {\n    comm.send({event_type: 'delete', 'id': id});\n  }\n  delete PyViz.plot_index[id];\n  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n    var doc = window.Bokeh.index[id].model.document\n    doc.clear();\n    const i = window.Bokeh.documents.indexOf(doc);\n    if (i > -1) {\n      window.Bokeh.documents.splice(i, 1);\n    }\n  }\n}\n\n/**\n * Handle kernel restart event\n */\nfunction handle_kernel_cleanup(event, handle) {\n  delete PyViz.comms[\"hv-extension-comm\"];\n  window.PyViz.plot_index = {}\n}\n\n/**\n * Handle update_display_data messages\n */\nfunction handle_update_output(event, handle) {\n  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n  handle_add_output(event, handle)\n}\n\nfunction register_renderer(events, OutputArea) {\n  function append_mime(data, metadata, element) {\n    // create a DOM node to render to\n    var toinsert = this.create_output_subarea(\n    metadata,\n    CLASS_NAME,\n    EXEC_MIME_TYPE\n    );\n    this.keyboard_manager.register_events(toinsert);\n    // Render to node\n    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n    render(props, toinsert[0]);\n    element.append(toinsert);\n    return toinsert\n  }\n\n  events.on('output_added.OutputArea', handle_add_output);\n  events.on('output_updated.OutputArea', handle_update_output);\n  events.on('clear_output.CodeCell', handle_clear_output);\n  events.on('delete.Cell', handle_clear_output);\n  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n\n  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n    safe: true,\n    index: 0\n  });\n}\n\nif (window.Jupyter !== undefined) {\n  try {\n    var events = require('base/js/events');\n    var OutputArea = require('notebook/js/outputarea').OutputArea;\n    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n      register_renderer(events, OutputArea);\n    }\n  } catch(err) {\n  }\n}\n"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>*[data-root-id],\n",
       "*[data-root-id] > * {\n",
       "  box-sizing: border-box;\n",
       "  font-family: var(--jp-ui-font-family);\n",
       "  font-size: var(--jp-ui-font-size1);\n",
       "  color: var(--vscode-editor-foreground, var(--jp-ui-font-color1));\n",
       "}\n",
       "\n",
       "/* Override VSCode background color */\n",
       ".cell-output-ipywidget-background:has(\n",
       "    > .cell-output-ipywidget-background > .lm-Widget > *[data-root-id]\n",
       "  ),\n",
       ".cell-output-ipywidget-background:has(> .lm-Widget > *[data-root-id]) {\n",
       "  background-color: transparent !important;\n",
       "}\n",
       "</style>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n",
      "--------Version info---------\n",
      "Polars:              1.9.0\n",
      "Index type:          UInt32\n",
      "Platform:            macOS-12.7.6-x86_64-i386-64bit\n",
      "Python:              3.11.5 (main, Jan 16 2024, 17:25:53) [Clang 14.0.0 (clang-1400.0.29.202)]\n",
      "\n",
      "----Optional dependencies----\n",
      "adbc_driver_manager  1.1.0\n",
      "altair               5.4.0\n",
      "cloudpickle          3.0.0\n",
      "connectorx           0.3.3\n",
      "deltalake            0.19.1\n",
      "fastexcel            0.11.6\n",
      "fsspec               2023.12.2\n",
      "gevent               24.2.1\n",
      "great_tables         0.10.0\n",
      "matplotlib           3.9.2\n",
      "nest_asyncio         1.6.0\n",
      "numpy                2.0.2\n",
      "openpyxl             3.1.5\n",
      "pandas               2.2.2\n",
      "pyarrow              17.0.0\n",
      "pydantic             2.8.2\n",
      "pyiceberg            0.6.1\n",
      "sqlalchemy           2.0.32\n",
      "torch                <not installed>\n",
      "xlsx2csv             0.8.3\n",
      "xlsxwriter           3.2.0\n"
     ]
    }
   ],
   "source": [
    "import polars as pl\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib_inline.backend_inline\n",
    "import chardet\n",
    "import pprint as pp\n",
    "import hvplot.polars\n",
    "hvplot.extension('matplotlib')\n",
    "\n",
    "matplotlib_inline.backend_inline.set_matplotlib_formats(\"retina\")\n",
    "pd.options.mode.copy_on_write = True\n",
    "print(pd.options.mode.copy_on_write)\n",
    "pl.Config.set_verbose(True)\n",
    "pl.show_versions()\n",
    "\n",
    "def HR():\n",
    "    print(\"-\"*40)\n",
    "\n",
    "@pl.Config(tbl_cols=-1, ascii_tables=True)\n",
    "def tight_layout(df: pl.DataFrame, n=5) -> None:\n",
    "    with pl.Config(tbl_cols=-1, fmt_str_lengths=n):\n",
    "        print(df)\n",
    "\n",
    "def detect_encoding(filename: str) -> str:\n",
    "    \"\"\"Return the most probable character encoding for a file.\"\"\"\n",
    "    with open(filename, \"rb\") as f:\n",
    "        raw_data = f.read()\n",
    "        print(raw_data)\n",
    "        result = chardet.detect(raw_data)\n",
    "        return result[\"encoding\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb7cb9a1-667b-418b-8eb3-e56b6a5b1b54",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.1 Introduction\n",
    "[back to Top](#Top)\n",
    "\n",
    "One of the most common tasks in data analysis is aggregation and grouping operations on datasets.\n",
    "\n",
    "This is especially true for large datasets, where summarizing and aggregating data is often necessary to gain insights and make sense of information."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e001580e-32c5-4542-ae3d-3285c0523937",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.2 Loading the Data\n",
    "[back to Top](#Top)\n",
    "\n",
    "Use the Fuel Economy dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "6acffdc8-da7f-4b7f-af67-36b838c02c12",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "avg line length: 434.78027\n",
      "std. dev. line length: 23.885818\n",
      "initial row estimate: 47850\n",
      "no. of chunks: 4 processed by: 4 threads.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(48202, 84)\n",
      "29.34MB\n",
      "<class 'polars.dataframe.frame.DataFrame'>\n",
      "(48202, 15)\n",
      "2.94MB\n"
     ]
    }
   ],
   "source": [
    "path = 'data/vehicles.csv'\n",
    "raw = pl.read_csv(path, null_values=['NA'])\n",
    "print(raw.shape)\n",
    "print(f\"{raw.estimated_size(unit='mb'):.2f}MB\")\n",
    "\n",
    "@pl.StringCache()\n",
    "def tweak_auto(df):\n",
    "    cols = ['year', 'make', 'model', 'displ', 'cylinders', 'trany', \n",
    "           'drive', 'VClass', 'fuelType', 'barrels08', 'city08', \n",
    "           'highway08', 'createdOn']\n",
    "    return (\n",
    "        df\n",
    "        .select(pl.col(cols))\n",
    "        .with_columns( \n",
    "            pl.col('year').cast(pl.Int16),\n",
    "            pl.col(['cylinders', 'highway08', 'city08']).cast(pl.UInt8),\n",
    "            pl.col(['displ', 'barrels08']).cast(pl.Float32),\n",
    "            pl.col(['make', 'model', 'VClass', 'drive', 'fuelType']).cast(pl.Categorical),\n",
    "            pl.col('createdOn').str.to_datetime('%a %b %d %H:%M:%S %Z %Y'),\n",
    "            pl.col('trany')                    \n",
    "               .str.contains('Automatic')\n",
    "               .fill_null('Automatic').alias('is_automatic'),\n",
    "            pl.col('trany')\n",
    "                .str.extract(r'(\\d+)')\n",
    "                .cast(pl.UInt8)\n",
    "                .fill_null(6).alias('num_gears'),\n",
    "        )\n",
    "    )\n",
    "\n",
    "autos = tweak_auto(raw)\n",
    "print(type(autos))\n",
    "print(autos.shape)\n",
    "print(f\"{autos.estimated_size(unit='mb'):.2f}MB\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0ff94516-3b18-47f7-8145-787465ab7f24",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'year': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'make': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'model': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'displ': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'cylinders': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'trany': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'drive': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'VClass': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'fuelType': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'barrels08': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'city08': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'highway08': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'createdOn': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'is_automatic': {'SORTED_ASC': False, 'SORTED_DESC': False},\n",
       " 'num_gears': {'SORTED_ASC': False, 'SORTED_DESC': False}}"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autos.flags"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4cdf9614-81c2-420c-826c-09374294c5ca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'year': False,\n",
       " 'make': False,\n",
       " 'model': False,\n",
       " 'displ': False,\n",
       " 'cylinders': False,\n",
       " 'trany': False,\n",
       " 'drive': False,\n",
       " 'VClass': False,\n",
       " 'fuelType': False,\n",
       " 'barrels08': False,\n",
       " 'city08': False,\n",
       " 'highway08': False,\n",
       " 'createdOn': False,\n",
       " 'is_automatic': False,\n",
       " 'num_gears': False}"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "{col:autos[col].is_sorted() for col in autos.columns}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23f93330-2446-4435-9952-a3e36eae355a",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.3 Aggregations\n",
    "[back to Top](#Top)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e4b5f1fa-714a-46dd-80b5-d42b22fb088b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['abs', 'add', 'agg_groups', 'alias', 'all', 'and_', 'any', 'append', 'approx_n_unique', 'arccos', 'arccosh', 'arcsin', 'arcsinh', 'arctan', 'arctanh', 'arg_max', 'arg_min', 'arg_sort', 'arg_true', 'arg_unique', 'arr', 'backward_fill', 'bin', 'bitwise_and', 'bitwise_count_ones', 'bitwise_count_zeros', 'bitwise_leading_ones', 'bitwise_leading_zeros', 'bitwise_or', 'bitwise_trailing_ones', 'bitwise_trailing_zeros', 'bitwise_xor', 'bottom_k', 'bottom_k_by', 'cast', 'cat', 'cbrt', 'ceil', 'clip', 'cos', 'cosh', 'cot', 'count', 'cum_count', 'cum_max', 'cum_min', 'cum_prod', 'cum_sum', 'cumulative_eval', 'cut', 'degrees', 'deserialize', 'diff', 'dot', 'drop_nans', 'drop_nulls', 'dt', 'entropy', 'eq', 'eq_missing', 'ewm_mean', 'ewm_mean_by', 'ewm_std', 'ewm_var', 'exclude', 'exp', 'explode', 'extend_constant', 'fill_nan', 'fill_null', 'filter', 'first', 'flatten', 'floor', 'floordiv', 'forward_fill', 'from_json', 'gather', 'gather_every', 'ge', 'get', 'gt', 'has_nulls', 'hash', 'head', 'hist', 'implode', 'inspect', 'interpolate', 'interpolate_by', 'is_between', 'is_duplicated', 'is_finite', 'is_first_distinct', 'is_in', 'is_infinite', 'is_last_distinct', 'is_nan', 'is_not_nan', 'is_not_null', 'is_null', 'is_unique', 'kurtosis', 'last', 'le', 'len', 'limit', 'list', 'log', 'log10', 'log1p', 'lower_bound', 'lt', 'map_batches', 'map_elements', 'max', 'mean', 'median', 'meta', 'min', 'mod', 'mode', 'mul', 'n_unique', 'name', 'nan_max', 'nan_min', 'ne', 'ne_missing', 'neg', 'not_', 'null_count', 'or_', 'over', 'pct_change', 'peak_max', 'peak_min', 'pipe', 'pow', 'product', 'qcut', 'quantile', 'radians', 'rank', 'rechunk', 'register_plugin', 'reinterpret', 'repeat_by', 'replace', 'replace_strict', 'reshape', 'reverse', 'rle', 'rle_id', 'rolling', 'rolling_map', 'rolling_max', 'rolling_max_by', 'rolling_mean', 'rolling_mean_by', 'rolling_median', 'rolling_median_by', 'rolling_min', 'rolling_min_by', 'rolling_quantile', 'rolling_quantile_by', 'rolling_skew', 'rolling_std', 'rolling_std_by', 'rolling_sum', 'rolling_sum_by', 'rolling_var', 'rolling_var_by', 'round', 'round_sig_figs', 'sample', 'search_sorted', 'set_sorted', 'shift', 'shrink_dtype', 'shuffle', 'sign', 'sin', 'sinh', 'skew', 'slice', 'sort', 'sort_by', 'sqrt', 'std', 'str', 'struct', 'sub', 'sum', 'tail', 'tan', 'tanh', 'to_physical', 'top_k', 'top_k_by', 'truediv', 'unique', 'unique_counts', 'upper_bound', 'value_counts', 'var', 'where', 'xor']\n"
     ]
    }
   ],
   "source": [
    "col = pl.col('make')\n",
    "print(sorted([att for att in dir(col) if not att.startswith('_')]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e5eeb8b-8aa2-40ab-bccf-ac1549642fb8",
   "metadata": {},
   "source": [
    "---\n",
    "Experiment using some aggregation functions on the autos `comb08` column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c3471453-c520-4ddc-b885-9c15d6024377",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['VClass',\n",
       " 'barrels08',\n",
       " 'city08',\n",
       " 'createdOn',\n",
       " 'cylinders',\n",
       " 'displ',\n",
       " 'drive',\n",
       " 'fuelType',\n",
       " 'highway08',\n",
       " 'is_automatic',\n",
       " 'make',\n",
       " 'model',\n",
       " 'num_gears',\n",
       " 'trany',\n",
       " 'year']"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(autos.columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa4625c3-f603-4ee6-a533-7e0432ebe3b4",
   "metadata": {},
   "source": [
    "* Using `.select()` creates a *new* dataframe with new columns.\n",
    "* Using `.with_columns()` appends a new columns to the original dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c51dd2d7-77fc-4441-a30b-60b131af3710",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>mean_city</th><th>std_city</th><th>var_city</th><th>q99_city</th></tr><tr><td>f64</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>19.929235</td><td>12.658018</td><td>160.225423</td><td>95.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 4)\n",
       "+-----------+-----------+------------+----------+\n",
       "| mean_city | std_city  | var_city   | q99_city |\n",
       "| ---       | ---       | ---        | ---      |\n",
       "| f64       | f64       | f64        | f64      |\n",
       "+===============================================+\n",
       "| 19.929235 | 12.658018 | 160.225423 | 95.0     |\n",
       "+-----------+-----------+------------+----------+"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "city = pl.col('city08')\n",
    "\n",
    "test = (\n",
    "    autos\n",
    "    .select(\n",
    "        mean_city=city.mean(),\n",
    "        std_city=city.std(),\n",
    "        var_city=city.var(),\n",
    "        q99_city=city.quantile(.99)\n",
    "    )\n",
    ")\n",
    "test"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb5a3353-8fed-4a10-a01b-2255ee87e58e",
   "metadata": {},
   "source": [
    "If we add a new aggregated column to an existing dataframe, the column will be appended to the right of the existing columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "cfdc97e7-913e-4043-a00e-d84e1825d65f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (2, 16)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th><th>mean_city</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td><td>f64</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>19.929235</td></tr><tr><td>1985</td><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>4.9</td><td>12</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>9</td><td>14</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>19.929235</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (2, 16)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | hig | cre | is_ | num | mean_c |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | hwa | ate | aut | _ge | ity    |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | y08 | dOn | oma | ars | ---    |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | --- | --- | tic | --- | f64    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     | u8  | dat | --- | u8  |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | eti | str |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | me[ |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | μs] |     |     |        |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25  | 201 | fal | 5   | 19.929 |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |     | 3-0 | se  |     | 235    |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |     | 1-0 |     |     |        |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |     | 1   |     |     |        |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 198 | Fer | Tes | 4.9 | 12  | Man | Rea | Two | Reg | 27. | 9   | 14  | 201 | fal | 5   | 19.929 |\n",
       "| 5   | rar | tar |     |     | ual | r-W | Sea | ula | 046 |     |     | 3-0 | se  |     | 235    |\n",
       "|     | i   | oss |     |     | 5-s | hee | ter | r   | 364 |     |     | 1-0 |     |     |        |\n",
       "|     |     | a   |     |     | pd  | l   | s   |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .with_columns(mean_city=city.mean())\n",
    "    .head(2)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "67ef310a-d082-4682-a64e-4ce52c668dae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>mean_city</th></tr><tr><td>f64</td></tr></thead><tbody><tr><td>19.929235</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 1)\n",
       "+-----------+\n",
       "| mean_city |\n",
       "| ---       |\n",
       "| f64       |\n",
       "+===========+\n",
       "| 19.929235 |\n",
       "+-----------+"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# contrast to .select()\n",
    "(\n",
    "    autos\n",
    "    .select(mean_city=city.mean())\n",
    "    .head(2)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8a1cae0-f0c9-4606-8c43-1217987a729d",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.4 Lists\n",
    "[back to Top](#Top)\n",
    "\n",
    "* Polars can have typed lists in a column.\n",
    "* Polars stores the lists as typed arrays.\n",
    "* This allows Polars to **perform operations on the lists in a vectorized manner**.\n",
    "* Polars list operations are methods found on the column `.list` attribute\n",
    "\n",
    "*Example:*\n",
    "* We have columns of student test scores and want to drop the two lowest scores for each student.\n",
    "* There is not easy way to do this with columns.\n",
    "* However, we can combine the test columns into a single column of lists.\n",
    "* Then, we use list operations to sort the list and remove the two lowest values.\n",
    "* Finally we sort the list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "6de345b1-dd4e-4906-ba7d-e3ef86a8d5f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: (3, 5)\n",
      "+-------+-------+-------+-------+-------+\n",
      "| name  | test1 | test2 | test3 | test4 |\n",
      "| ---   | ---   | ---   | ---   | ---   |\n",
      "| str   | i64   | i64   | i64   | i64   |\n",
      "+=======================================+\n",
      "| Tom   | 99    | 92    | 91    | 94    |\n",
      "| Sally | 98    | null  | 93    | 92    |\n",
      "| Jose  | 95    | 99    | 95    | 99    |\n",
      "+-------+-------+-------+-------+-------+\n"
     ]
    }
   ],
   "source": [
    "tests = pl.DataFrame({\n",
    "    'name':['Tom', 'Sally', 'Jose'],\n",
    "    'test1':[99, 98, 95],\n",
    "    'test2':[92, None, 99],\n",
    "    'test3':[91, 93, 95],\n",
    "    'test4':[94, 92, 99]\n",
    "})\n",
    "print(tests)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f9d00b9-9302-49d9-ba8f-7af5f3464408",
   "metadata": {},
   "source": [
    "Next, create a dataframe with a column for each step\n",
    "\n",
    "**Since each of the columns depends on the previous column, we need to put them in their own `.with_column` call, otherwise Polars would try to evaluate the columns in parallel.**\n",
    "\n",
    "* First column: Combine the test scores into a list with `pl.concat_list()` and name the column `scores`.\n",
    "* Next column: Use the `.list` attribute, which has `.sort()`, to sort values in each cell in the column.\n",
    "* Next column: Use `.slice()` to remove the two lowest values.\n",
    "* Final column: Sum up the list to get the total score for each student.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e9b9a8e6-d865-47cb-a9f8-94f258e1186f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import polars.selectors as cs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5149b36a-792e-4b95-b7d3-dbbdfc4d08dc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>scores</th></tr><tr><td>list[i64]</td></tr></thead><tbody><tr><td>[99, 92, … 94]</td></tr><tr><td>[98, null, … 92]</td></tr><tr><td>[95, 99, … 99]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 1)\n",
       "+------------------+\n",
       "| scores           |\n",
       "| ---              |\n",
       "| list[i64]        |\n",
       "+==================+\n",
       "| [99, 92, … 94]   |\n",
       "| [98, null, … 92] |\n",
       "| [95, 99, … 99]   |\n",
       "+------------------+"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# step 1\n",
    "(\n",
    "    tests\n",
    "    .select(scores=pl.concat_list(cs.matches(r'test\\d+')))\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bdb7f569-bacf-4b90-8776-cc4e782a0a88",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>scores</th><th>sorted_scores</th></tr><tr><td>list[i64]</td><td>list[i64]</td></tr></thead><tbody><tr><td>[99, 92, … 94]</td><td>[91, 92, … 99]</td></tr><tr><td>[98, null, … 92]</td><td>[null, 92, … 98]</td></tr><tr><td>[95, 99, … 99]</td><td>[95, 95, … 99]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 2)\n",
       "+------------------+------------------+\n",
       "| scores           | sorted_scores    |\n",
       "| ---              | ---              |\n",
       "| list[i64]        | list[i64]        |\n",
       "+=====================================+\n",
       "| [99, 92, … 94]   | [91, 92, … 99]   |\n",
       "| [98, null, … 92] | [null, 92, … 98] |\n",
       "| [95, 99, … 99]   | [95, 95, … 99]   |\n",
       "+------------------+------------------+"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# step 2\n",
    "(\n",
    "    tests\n",
    "    .select(scores=pl.concat_list(cs.matches(r'test\\d+')))\n",
    "    .with_columns(sorted_scores=pl.col('scores').list.sort()) # 1\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "9f213eb8-4695-4363-bdf9-f0e0aed7ff4c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>scores</th><th>sorted_scores</th><th>slice_scores</th></tr><tr><td>list[i64]</td><td>list[i64]</td><td>list[i64]</td></tr></thead><tbody><tr><td>[99, 92, … 94]</td><td>[91, 92, … 99]</td><td>[94, 99]</td></tr><tr><td>[98, null, … 92]</td><td>[null, 92, … 98]</td><td>[93, 98]</td></tr><tr><td>[95, 99, … 99]</td><td>[95, 95, … 99]</td><td>[99, 99]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 3)\n",
       "+------------------+------------------+--------------+\n",
       "| scores           | sorted_scores    | slice_scores |\n",
       "| ---              | ---              | ---          |\n",
       "| list[i64]        | list[i64]        | list[i64]    |\n",
       "+====================================================+\n",
       "| [99, 92, … 94]   | [91, 92, … 99]   | [94, 99]     |\n",
       "| [98, null, … 92] | [null, 92, … 98] | [93, 98]     |\n",
       "| [95, 99, … 99]   | [95, 95, … 99]   | [99, 99]     |\n",
       "+------------------+------------------+--------------+"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# step 3\n",
    "(\n",
    "    tests\n",
    "    .select(scores=pl.concat_list(cs.matches(r'test\\d+')))\n",
    "    .with_columns(sorted_scores=pl.col('scores').list.sort()) # 1\n",
    "    .with_columns(slice_scores=pl.col('sorted_scores').list.slice(2,4)) #2\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7425433e-5784-41c6-939b-b832d4058766",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>scores</th><th>sorted_scores</th><th>slice_scores</th><th>sum_scores</th></tr><tr><td>list[i64]</td><td>list[i64]</td><td>list[i64]</td><td>i64</td></tr></thead><tbody><tr><td>[99, 92, … 94]</td><td>[91, 92, … 99]</td><td>[94, 99]</td><td>193</td></tr><tr><td>[98, null, … 92]</td><td>[null, 92, … 98]</td><td>[93, 98]</td><td>191</td></tr><tr><td>[95, 99, … 99]</td><td>[95, 95, … 99]</td><td>[99, 99]</td><td>198</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 4)\n",
       "+------------------+------------------+--------------+------------+\n",
       "| scores           | sorted_scores    | slice_scores | sum_scores |\n",
       "| ---              | ---              | ---          | ---        |\n",
       "| list[i64]        | list[i64]        | list[i64]    | i64        |\n",
       "+=================================================================+\n",
       "| [99, 92, … 94]   | [91, 92, … 99]   | [94, 99]     | 193        |\n",
       "| [98, null, … 92] | [null, 92, … 98] | [93, 98]     | 191        |\n",
       "| [95, 99, … 99]   | [95, 95, … 99]   | [99, 99]     | 198        |\n",
       "+------------------+------------------+--------------+------------+"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# step 4\n",
    "(\n",
    "    tests\n",
    "    .select(scores=pl.concat_list(cs.matches(r'test\\d+')))\n",
    "    .with_columns(sorted_scores=pl.col('scores').list.sort()) # 1\n",
    "    .with_columns(slice_scores=pl.col('sorted_scores').list.slice(2,4)) #2\n",
    "    .with_columns(sum_scores=pl.col('slice_scores').list.sum())\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "541c7bf1-b722-4fda-a3de-e89296260e2c",
   "metadata": {},
   "source": [
    "---\n",
    "## 4.4.1 Using `.with_columns_seq`\n",
    "[back to Top](#Top)\n",
    "\n",
    "In Polars, `with_columns_seq` and `with_columns` are both methods used to add columns to a DataFrame or LazyFrame, but they have some key differences:\n",
    "\n",
    "Execution Order\n",
    "\n",
    "**with_columns:**\n",
    "- Executes expressions in parallel\n",
    "- Each expression is unaware of other expressions in the same context\n",
    "\n",
    "**with_columns_seq:**\n",
    "- Executes expressions sequentially\n",
    "- Designed for cases where the work per expression is cheap\n",
    "\n",
    "Behavior with New Columns\n",
    "\n",
    "**with_columns:**\n",
    "- Cannot reference newly created columns within the same method call\n",
    "\n",
    "**with_columns_seq:**\n",
    "- Still does not allow referencing newly created columns within the same call\n",
    "- Constructs the new DataFrame after all expressions have been computed[3]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c2e975ac-c423-4445-92d5-97e92c42fbe2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>scores</th><th>sorted_scores</th><th>slice_scores</th><th>sum_scores</th></tr><tr><td>list[i64]</td><td>list[i64]</td><td>list[i64]</td><td>i64</td></tr></thead><tbody><tr><td>[99, 92, … 94]</td><td>[91, 92, … 99]</td><td>[94, 99]</td><td>193</td></tr><tr><td>[98, null, … 92]</td><td>[null, 92, … 98]</td><td>[93, 98]</td><td>191</td></tr><tr><td>[95, 99, … 99]</td><td>[95, 95, … 99]</td><td>[99, 99]</td><td>198</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 4)\n",
       "+------------------+------------------+--------------+------------+\n",
       "| scores           | sorted_scores    | slice_scores | sum_scores |\n",
       "| ---              | ---              | ---          | ---        |\n",
       "| list[i64]        | list[i64]        | list[i64]    | i64        |\n",
       "+=================================================================+\n",
       "| [99, 92, … 94]   | [91, 92, … 99]   | [94, 99]     | 193        |\n",
       "| [98, null, … 92] | [null, 92, … 98] | [93, 98]     | 191        |\n",
       "| [95, 99, … 99]   | [95, 95, … 99]   | [99, 99]     | 198        |\n",
       "+------------------+------------------+--------------+------------+"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Explicit sequential operation\n",
    "(\n",
    "    tests\n",
    "    .select(scores=pl.concat_list(cs.matches(r'test\\d+')))\n",
    "    .with_columns_seq(sorted_scores=pl.col('scores').list.sort()) # 1\n",
    "    .with_columns_seq(slice_scores=pl.col('sorted_scores').list.slice(2,4)) #2\n",
    "    .with_columns_seq(sum_scores=pl.col('slice_scores').list.sum())\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47392ca1-941a-4505-a559-232fdd9d6968",
   "metadata": {},
   "source": [
    "---\n",
    "Selecting Columns\n",
    "\n",
    "We can `.exclude` to rewrite the previous example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "1c4e7a9b-e4e9-43f5-a56b-d887cef9eb83",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>test1</th><th>test2</th><th>test3</th><th>test4</th></tr><tr><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>99</td><td>92</td><td>91</td><td>94</td></tr><tr><td>98</td><td>null</td><td>93</td><td>92</td></tr><tr><td>95</td><td>99</td><td>95</td><td>99</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 4)\n",
       "+-------+-------+-------+-------+\n",
       "| test1 | test2 | test3 | test4 |\n",
       "| ---   | ---   | ---   | ---   |\n",
       "| i64   | i64   | i64   | i64   |\n",
       "+===============================+\n",
       "| 99    | 92    | 91    | 94    |\n",
       "| 98    | null  | 93    | 92    |\n",
       "| 95    | 99    | 95    | 99    |\n",
       "+-------+-------+-------+-------+"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    tests\n",
    "    .select(pl.all().exclude('name'))\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d97d7ce7-8426-4cc6-b82c-94ff06ec399f",
   "metadata": {},
   "source": [
    "We can also use *set difference* to start with all of the columns and remove the *name* column.\n",
    "\n",
    "For this, we cannot use the expression syntax directly. We have to use `cs.all()` to use set operations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "8fcb6b8a-527e-49ed-b4f3-9e6e74c05f9b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "arithmetic on string and numeric not allowed, try an explicit cast first\n",
      "\n",
      "Resolved plan until failure:\n",
      "\n",
      "\t---> FAILED HERE RESOLVING 'select' <---\n",
      " SELECT [[(col(\"name\")) - (col(\"name\"))], [(col(\"test1\")) - (col(\"name\"))], [(col(\"test2\")) - (col(\"name\"))], [(col(\"test3\")) - (col(\"name\"))], [(col(\"test4\")) - (col(\"name\"))]] FROM\n",
      "  DF [\"name\", \"test1\", \"test2\", \"test3\"]; PROJECT */5 COLUMNS; SELECTION: None\n"
     ]
    }
   ],
   "source": [
    "try:\n",
    "    print(tests.select(pl.all() - pl.col('name')))\n",
    "except Exception as e:\n",
    "    print(e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "70516766-a462-4a4d-9eb8-f1fa15310da2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Using cs.all for set operations\n",
    "import polars.selectors as cs\n",
    "\n",
    "# tests.select(cs.all() - pl.col('name'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "94858e95-a0fc-482f-b063-1c04097aff45",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 5)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>name</th><th>test1</th><th>test2</th><th>test3</th><th>test4</th></tr><tr><td>str</td><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>&quot;Tom&quot;</td><td>99</td><td>92</td><td>91</td><td>94</td></tr><tr><td>&quot;Sally&quot;</td><td>98</td><td>null</td><td>93</td><td>92</td></tr><tr><td>&quot;Jose&quot;</td><td>95</td><td>99</td><td>95</td><td>99</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 5)\n",
       "+-------+-------+-------+-------+-------+\n",
       "| name  | test1 | test2 | test3 | test4 |\n",
       "| ---   | ---   | ---   | ---   | ---   |\n",
       "| str   | i64   | i64   | i64   | i64   |\n",
       "+=======================================+\n",
       "| Tom   | 99    | 92    | 91    | 94    |\n",
       "| Sally | 98    | null  | 93    | 92    |\n",
       "| Jose  | 95    | 99    | 95    | 99    |\n",
       "+-------+-------+-------+-------+-------+"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check\n",
    "tests.select(pl.all())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "1a3ecf9c-ed2a-4b6b-8521-d771a314b0ad",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 5)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>name</th><th>test1</th><th>test2</th><th>test3</th><th>test4</th></tr><tr><td>str</td><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>&quot;Tom&quot;</td><td>99</td><td>92</td><td>91</td><td>94</td></tr><tr><td>&quot;Sally&quot;</td><td>98</td><td>null</td><td>93</td><td>92</td></tr><tr><td>&quot;Jose&quot;</td><td>95</td><td>99</td><td>95</td><td>99</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 5)\n",
       "+-------+-------+-------+-------+-------+\n",
       "| name  | test1 | test2 | test3 | test4 |\n",
       "| ---   | ---   | ---   | ---   | ---   |\n",
       "| str   | i64   | i64   | i64   | i64   |\n",
       "+=======================================+\n",
       "| Tom   | 99    | 92    | 91    | 94    |\n",
       "| Sally | 98    | null  | 93    | 92    |\n",
       "| Jose  | 95    | 99    | 95    | 99    |\n",
       "+-------+-------+-------+-------+-------+"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tests.select(cs.all())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee92050f-cb3f-4ade-b1ff-7c062f1ad0ed",
   "metadata": {},
   "source": [
    "We can also use `cs.starts_with` to filter the columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "f36d7008-926e-4a08-bf6a-37c89a5b61fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>test1</th><th>test2</th><th>test3</th><th>test4</th></tr><tr><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>99</td><td>92</td><td>91</td><td>94</td></tr><tr><td>98</td><td>null</td><td>93</td><td>92</td></tr><tr><td>95</td><td>99</td><td>95</td><td>99</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 4)\n",
       "+-------+-------+-------+-------+\n",
       "| test1 | test2 | test3 | test4 |\n",
       "| ---   | ---   | ---   | ---   |\n",
       "| i64   | i64   | i64   | i64   |\n",
       "+===============================+\n",
       "| 99    | 92    | 91    | 94    |\n",
       "| 98    | null  | 93    | 92    |\n",
       "| 95    | 99    | 95    | 99    |\n",
       "+-------+-------+-------+-------+"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tests.select(cs.starts_with('test'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6481651-a551-476b-b7e4-4863966a97dc",
   "metadata": {},
   "source": [
    "If we want to calculate horizontal aggregations, we can also use the list functionality.\n",
    "\n",
    "Here, calculate the maximum test score for each student."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "8d70f9da-3c95-428a-b1c5-6525aac5ec1e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['name', 'test1', 'test2', 'test3', 'test4']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tests.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "de6ddb97-45fd-4a42-b3f8-cfd22c9aa2b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>scores</th><th>max</th></tr><tr><td>list[i64]</td><td>i64</td></tr></thead><tbody><tr><td>[99, 92, … 94]</td><td>99</td></tr><tr><td>[98, null, … 92]</td><td>98</td></tr><tr><td>[95, 99, … 99]</td><td>99</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 2)\n",
       "+------------------+-----+\n",
       "| scores           | max |\n",
       "| ---              | --- |\n",
       "| list[i64]        | i64 |\n",
       "+========================+\n",
       "| [99, 92, … 94]   | 99  |\n",
       "| [98, null, … 92] | 98  |\n",
       "| [95, 99, … 99]   | 99  |\n",
       "+------------------+-----+"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    tests\n",
    "    .select(scores=pl.concat_list(cs.starts_with('test')))\n",
    "    .with_columns(max=pl.col('scores').list.max())\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa3c151f-afe3-43cd-a896-7c57526bdd0e",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.5 GroupBy Operations\n",
    "[back to Top](#Top)\n",
    "\n",
    "Grouping is a general term for a class of operations that involves:\n",
    "\n",
    "* Splitting up a dataset based on some criteria.\n",
    "* Applying a function to each group independently.\n",
    "* Combining the results into a data structure.\n",
    "\n",
    "In this example, we do a basic group-by operation on the *make* column. \n",
    "\n",
    "* Group the data by the *make* column, using `.group_by()`\n",
    "* Calculate each group's average *city08* value, using `.agg()` on the result of the groupby."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "8caab552-4faa-4d23-8da3-c38cfcc0fda7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<polars.dataframe.group_by.GroupBy at 0x12436fdd0>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14116e5b-fa12-429f-b7b1-d9736d2e3f6b",
   "metadata": {},
   "source": [
    "* Use a column expression to create the aggregation.\n",
    "* Use `.mean()` on the *comb08* column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "55b594b8-40e8-4e25-8e8f-475a0695d6b2",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Lamborghini&quot;</td><td>10.662577</td></tr><tr><td>&quot;STI&quot;</td><td>15.0</td></tr><tr><td>&quot;Jaguar&quot;</td><td>17.97037</td></tr><tr><td>&quot;Autokraft Limited&quot;</td><td>14.0</td></tr><tr><td>&quot;Ruf Automobile Gmbh&quot;</td><td>11.0</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Audi&quot;</td><td>20.760091</td></tr><tr><td>&quot;CODA Automotive&quot;</td><td>77.0</td></tr><tr><td>&quot;Superior Coaches Div E.p. Dutt…</td><td>10.0</td></tr><tr><td>&quot;Karma&quot;</td><td>23.0</td></tr><tr><td>&quot;Jeep&quot;</td><td>17.319111</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+---------------------------------+-----------+\n",
       "| make                            | mean_city |\n",
       "| ---                             | ---       |\n",
       "| cat                             | f64       |\n",
       "+=============================================+\n",
       "| Lamborghini                     | 10.662577 |\n",
       "| STI                             | 15.0      |\n",
       "| Jaguar                          | 17.97037  |\n",
       "| Autokraft Limited               | 14.0      |\n",
       "| Ruf Automobile Gmbh             | 11.0      |\n",
       "| …                               | …         |\n",
       "| Audi                            | 20.760091 |\n",
       "| CODA Automotive                 | 77.0      |\n",
       "| Superior Coaches Div E.p. Dutt… | 10.0      |\n",
       "| Karma                           | 23.0      |\n",
       "| Jeep                            | 17.319111 |\n",
       "+---------------------------------+-----------+"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .agg(mean_city=pl.col('city08').mean())\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2c57932-f968-42b7-9a59-d36ef748f5e8",
   "metadata": {},
   "source": [
    "* A `maintain_order` parameter can be set to True when calling `.group_by()`.\n",
    "* This maintains the order of the data.\n",
    "* Here, the output will be in the order that the *make* values were in the original data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "28b5bd2c-53c8-4fbd-91e3-869b034e4426",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td></tr><tr><td>&quot;Ferrari&quot;</td><td>11.842294</td></tr><tr><td>&quot;Dodge&quot;</td><td>15.461967</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Toyota&quot;</td><td>22.4</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;General Motors&quot;</td><td>15.0</td></tr><tr><td>&quot;Consulier Industries Inc&quot;</td><td>18.0</td></tr><tr><td>&quot;Goldacre&quot;</td><td>16.0</td></tr><tr><td>&quot;Isis Imports Ltd&quot;</td><td>16.0</td></tr><tr><td>&quot;PAS Inc - GMC&quot;</td><td>14.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+--------------------------+-----------+\n",
       "| make                     | mean_city |\n",
       "| ---                      | ---       |\n",
       "| cat                      | f64       |\n",
       "+======================================+\n",
       "| Alfa Romeo               | 19.402062 |\n",
       "| Ferrari                  | 11.842294 |\n",
       "| Dodge                    | 15.461967 |\n",
       "| Subaru                   | 21.130306 |\n",
       "| Toyota                   | 22.4      |\n",
       "| …                        | …         |\n",
       "| General Motors           | 15.0      |\n",
       "| Consulier Industries Inc | 18.0      |\n",
       "| Goldacre                 | 16.0      |\n",
       "| Isis Imports Ltd         | 16.0      |\n",
       "| PAS Inc - GMC            | 14.0      |\n",
       "+--------------------------+-----------+"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by('make', maintain_order=True)\n",
    "    .agg(mean_city=pl.col('city08').mean())\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e25fd8e-4d42-4a71-9e76-8be2e77d0dd7",
   "metadata": {},
   "source": [
    "We can use `.sort()` to sort the data by the *make* column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0fe56abe-bf36-436b-89f5-b75988fb1220",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td></tr><tr><td>&quot;Ferrari&quot;</td><td>11.842294</td></tr><tr><td>&quot;Dodge&quot;</td><td>15.461967</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Toyota&quot;</td><td>22.4</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Quantum Technologies&quot;</td><td>19.0</td></tr><tr><td>&quot;London Taxi&quot;</td><td>21.0</td></tr><tr><td>&quot;Excalibur Autos&quot;</td><td>12.0</td></tr><tr><td>&quot;Mahindra&quot;</td><td>19.0</td></tr><tr><td>&quot;Azure Dynamics&quot;</td><td>62.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+----------------------+-----------+\n",
       "| make                 | mean_city |\n",
       "| ---                  | ---       |\n",
       "| cat                  | f64       |\n",
       "+==================================+\n",
       "| Alfa Romeo           | 19.402062 |\n",
       "| Ferrari              | 11.842294 |\n",
       "| Dodge                | 15.461967 |\n",
       "| Subaru               | 21.130306 |\n",
       "| Toyota               | 22.4      |\n",
       "| …                    | …         |\n",
       "| Quantum Technologies | 19.0      |\n",
       "| London Taxi          | 21.0      |\n",
       "| Excalibur Autos      | 12.0      |\n",
       "| Mahindra             | 19.0      |\n",
       "| Azure Dynamics       | 62.0      |\n",
       "+----------------------+-----------+"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .agg(mean_city=pl.col('city08').mean())\n",
    "    .sort('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9770c79-9125-4d5c-8ee1-73d64aa85dd4",
   "metadata": {},
   "source": [
    "The make column is not alphabetically sorted, becuase the make column is a categorical columnn, and the categories are sorted by the order in which they appear in the original dataframe.\n",
    "\n",
    "We can specify the ordering when casting to pl.Categorical. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "f12761b7-769d-42bc-b045-930609268bd7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1985</td><td>&quot;AM General&quot;</td><td>&quot;Post Office DJ5 2WD&quot;</td><td>2.5</td><td>4</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle 2WD&quot;</td><td>&quot;Regular&quot;</td><td>18.594376</td><td>16</td><td>17</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1985</td><td>&quot;AM General&quot;</td><td>&quot;Post Office DJ8 2WD&quot;</td><td>4.2</td><td>6</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle 2WD&quot;</td><td>&quot;Regular&quot;</td><td>22.885386</td><td>13</td><td>13</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1984</td><td>&quot;AM General&quot;</td><td>&quot;FJ8c Post Office&quot;</td><td>4.2</td><td>6</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;2-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle 2WD&quot;</td><td>&quot;Regular&quot;</td><td>22.885386</td><td>13</td><td>13</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1984</td><td>&quot;AM General&quot;</td><td>&quot;DJ Po Vehicle 2WD&quot;</td><td>2.5</td><td>4</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;2-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle 2WD&quot;</td><td>&quot;Regular&quot;</td><td>17.500587</td><td>18</td><td>17</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1984</td><td>&quot;AM General&quot;</td><td>&quot;FJ8c Post Office&quot;</td><td>4.2</td><td>6</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;2-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle 2WD&quot;</td><td>&quot;Regular&quot;</td><td>22.885386</td><td>13</td><td>13</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>2017</td><td>&quot;smart&quot;</td><td>&quot;fortwo electric drive converti…</td><td>null</td><td>null</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Electricity&quot;</td><td>0.0792</td><td>112</td><td>91</td><td>2017-10-11 00:00:00</td><td>&quot;true&quot;</td><td>1</td></tr><tr><td>2018</td><td>&quot;smart&quot;</td><td>&quot;fortwo electric drive coupe&quot;</td><td>null</td><td>null</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Electricity&quot;</td><td>0.0744</td><td>124</td><td>94</td><td>2017-12-05 00:00:00</td><td>&quot;true&quot;</td><td>1</td></tr><tr><td>2018</td><td>&quot;smart&quot;</td><td>&quot;fortwo electric drive converti…</td><td>null</td><td>null</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Electricity&quot;</td><td>0.0792</td><td>112</td><td>91</td><td>2017-12-05 00:00:00</td><td>&quot;true&quot;</td><td>1</td></tr><tr><td>2019</td><td>&quot;smart&quot;</td><td>&quot;EQ fortwo (coupe)&quot;</td><td>null</td><td>null</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Electricity&quot;</td><td>0.0744</td><td>124</td><td>94</td><td>2019-01-29 00:00:00</td><td>&quot;true&quot;</td><td>1</td></tr><tr><td>2019</td><td>&quot;smart&quot;</td><td>&quot;EQ fortwo (convertible)&quot;</td><td>null</td><td>null</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Electricity&quot;</td><td>0.0792</td><td>112</td><td>91</td><td>2019-01-29 00:00:00</td><td>&quot;true&quot;</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | AM  | Pos | 2.5 | 4   | Aut | Rea | Spe | Reg | 18. | 16  | 17     | 2013- | true  | 3     |\n",
       "| 5   | Gen | t   |     |     | oma | r-W | cia | ula | 594 |     |        | 01-01 |       |       |\n",
       "|     | era | Off |     |     | tic | hee | l   | r   | 376 |     |        | 00:00 |       |       |\n",
       "|     | l   | ice |     |     | 3-s | l   | Pur |     |     |     |        | :00   |       |       |\n",
       "|     |     | DJ5 |     |     | pd  | Dri | pos |     |     |     |        |       |       |       |\n",
       "|     |     | 2WD |     |     |     | ve  | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 2WD |     |     |     |        |       |       |       |\n",
       "| 198 | AM  | Pos | 4.2 | 6   | Aut | Rea | Spe | Reg | 22. | 13  | 13     | 2013- | true  | 3     |\n",
       "| 5   | Gen | t   |     |     | oma | r-W | cia | ula | 885 |     |        | 01-01 |       |       |\n",
       "|     | era | Off |     |     | tic | hee | l   | r   | 386 |     |        | 00:00 |       |       |\n",
       "|     | l   | ice |     |     | 3-s | l   | Pur |     |     |     |        | :00   |       |       |\n",
       "|     |     | DJ8 |     |     | pd  | Dri | pos |     |     |     |        |       |       |       |\n",
       "|     |     | 2WD |     |     |     | ve  | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 2WD |     |     |     |        |       |       |       |\n",
       "| 198 | AM  | FJ8 | 4.2 | 6   | Aut | 2-W | Spe | Reg | 22. | 13  | 13     | 2013- | true  | 3     |\n",
       "| 4   | Gen | c   |     |     | oma | hee | cia | ula | 885 |     |        | 01-01 |       |       |\n",
       "|     | era | Pos |     |     | tic | l   | l   | r   | 386 |     |        | 00:00 |       |       |\n",
       "|     | l   | t   |     |     | 3-s | Dri | Pur |     |     |     |        | :00   |       |       |\n",
       "|     |     | Off |     |     | pd  | ve  | pos |     |     |     |        |       |       |       |\n",
       "|     |     | ice |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 2WD |     |     |     |        |       |       |       |\n",
       "| 198 | AM  | DJ  | 2.5 | 4   | Aut | 2-W | Spe | Reg | 17. | 18  | 17     | 2013- | true  | 3     |\n",
       "| 4   | Gen | Po  |     |     | oma | hee | cia | ula | 500 |     |        | 01-01 |       |       |\n",
       "|     | era | Veh |     |     | tic | l   | l   | r   | 587 |     |        | 00:00 |       |       |\n",
       "|     | l   | icl |     |     | 3-s | Dri | Pur |     |     |     |        | :00   |       |       |\n",
       "|     |     | e   |     |     | pd  | ve  | pos |     |     |     |        |       |       |       |\n",
       "|     |     | 2WD |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 2WD |     |     |     |        |       |       |       |\n",
       "| 198 | AM  | FJ8 | 4.2 | 6   | Aut | 2-W | Spe | Reg | 22. | 13  | 13     | 2013- | true  | 3     |\n",
       "| 4   | Gen | c   |     |     | oma | hee | cia | ula | 885 |     |        | 01-01 |       |       |\n",
       "|     | era | Pos |     |     | tic | l   | l   | r   | 386 |     |        | 00:00 |       |       |\n",
       "|     | l   | t   |     |     | 3-s | Dri | Pur |     |     |     |        | :00   |       |       |\n",
       "|     |     | Off |     |     | pd  | ve  | pos |     |     |     |        |       |       |       |\n",
       "|     |     | ice |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 2WD |     |     |     |        |       |       |       |\n",
       "| …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …      | …     | …     | …     |\n",
       "| 201 | sma | for | nul | nul | Aut | Rea | Two | Ele | 0.0 | 112 | 91     | 2017- | true  | 1     |\n",
       "| 7   | rt  | two | l   | l   | oma | r-W | Sea | ctr | 792 |     |        | 10-11 |       |       |\n",
       "|     |     | ele |     |     | tic | hee | ter | ici |     |     |        | 00:00 |       |       |\n",
       "|     |     | ctr |     |     | (A1 | l   | s   | ty  |     |     |        | :00   |       |       |\n",
       "|     |     | ic  |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | dri |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "|     |     | ve  |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | con |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | ver |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | ti… |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "| 201 | sma | for | nul | nul | Aut | Rea | Two | Ele | 0.0 | 124 | 94     | 2017- | true  | 1     |\n",
       "| 8   | rt  | two | l   | l   | oma | r-W | Sea | ctr | 744 |     |        | 12-05 |       |       |\n",
       "|     |     | ele |     |     | tic | hee | ter | ici |     |     |        | 00:00 |       |       |\n",
       "|     |     | ctr |     |     | (A1 | l   | s   | ty  |     |     |        | :00   |       |       |\n",
       "|     |     | ic  |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | dri |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "|     |     | ve  |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | cou |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | pe  |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "| 201 | sma | for | nul | nul | Aut | Rea | Two | Ele | 0.0 | 112 | 91     | 2017- | true  | 1     |\n",
       "| 8   | rt  | two | l   | l   | oma | r-W | Sea | ctr | 792 |     |        | 12-05 |       |       |\n",
       "|     |     | ele |     |     | tic | hee | ter | ici |     |     |        | 00:00 |       |       |\n",
       "|     |     | ctr |     |     | (A1 | l   | s   | ty  |     |     |        | :00   |       |       |\n",
       "|     |     | ic  |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | dri |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "|     |     | ve  |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | con |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | ver |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | ti… |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "| 201 | sma | EQ  | nul | nul | Aut | Rea | Two | Ele | 0.0 | 124 | 94     | 2019- | true  | 1     |\n",
       "| 9   | rt  | for | l   | l   | oma | r-W | Sea | ctr | 744 |     |        | 01-29 |       |       |\n",
       "|     |     | two |     |     | tic | hee | ter | ici |     |     |        | 00:00 |       |       |\n",
       "|     |     | (co |     |     | (A1 | l   | s   | ty  |     |     |        | :00   |       |       |\n",
       "|     |     | upe |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | )   |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 201 | sma | EQ  | nul | nul | Aut | Rea | Two | Ele | 0.0 | 112 | 91     | 2019- | true  | 1     |\n",
       "| 9   | rt  | for | l   | l   | oma | r-W | Sea | ctr | 792 |     |        | 01-29 |       |       |\n",
       "|     |     | two |     |     | tic | hee | ter | ici |     |     |        | 00:00 |       |       |\n",
       "|     |     | (co |     |     | (A1 | l   | s   | ty  |     |     |        | :00   |       |       |\n",
       "|     |     | nve |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | rti |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "|     |     | ble |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "|     |     | )   |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test\n",
    "(\n",
    "    autos\n",
    "    .with_columns(pl.col('make').cast(pl.Categorical('lexical')))\n",
    "    .sort('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ebd9e38f-b076-4f27-803b-08f7481070de",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;AM General&quot;</td><td>15.166667</td></tr><tr><td>&quot;ASC Incorporated&quot;</td><td>14.0</td></tr><tr><td>&quot;Acura&quot;</td><td>20.011792</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td></tr><tr><td>&quot;American Motors Corporation&quot;</td><td>16.296296</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Volkswagen&quot;</td><td>23.39819</td></tr><tr><td>&quot;Volvo&quot;</td><td>20.320775</td></tr><tr><td>&quot;Wallace Environmental&quot;</td><td>12.4375</td></tr><tr><td>&quot;Yugo&quot;</td><td>23.0</td></tr><tr><td>&quot;smart&quot;</td><td>68.421053</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+-----------------------------+-----------+\n",
       "| make                        | mean_city |\n",
       "| ---                         | ---       |\n",
       "| cat                         | f64       |\n",
       "+=========================================+\n",
       "| AM General                  | 15.166667 |\n",
       "| ASC Incorporated            | 14.0      |\n",
       "| Acura                       | 20.011792 |\n",
       "| Alfa Romeo                  | 19.402062 |\n",
       "| American Motors Corporation | 16.296296 |\n",
       "| …                           | …         |\n",
       "| Volkswagen                  | 23.39819  |\n",
       "| Volvo                       | 20.320775 |\n",
       "| Wallace Environmental       | 12.4375   |\n",
       "| Yugo                        | 23.0      |\n",
       "| smart                       | 68.421053 |\n",
       "+-----------------------------+-----------+"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .with_columns(pl.col('make').cast(pl.Categorical('lexical')))\n",
    "    .group_by('make')\n",
    "    .agg(mean_city=pl.col('city08').mean())\n",
    "    .sort('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aca06ecc-ad38-4f2e-a30e-32e53cd63c36",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.6 Multiple Aggregations\n",
    "[back to Top](#Top)\n",
    "\n",
    "We can pass multiple expressions to `.agg()` to perform multiple aggregations. \n",
    "\n",
    "Here, we calculate the average and median values for both *city08* and *highway08* columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "453d1ff2-52b2-45d5-8590-c0ae825cfa32",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city</th><th>median_city</th><th>median_highway</th></tr><tr><td>cat</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>&quot;AM General&quot;</td><td>15.166667</td><td>14.5</td><td>15.0</td></tr><tr><td>&quot;ASC Incorporated&quot;</td><td>14.0</td><td>14.0</td><td>21.0</td></tr><tr><td>&quot;Acura&quot;</td><td>20.011792</td><td>20.0</td><td>27.0</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td><td>19.0</td><td>25.0</td></tr><tr><td>&quot;American Motors Corporation&quot;</td><td>16.296296</td><td>16.0</td><td>20.0</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Volkswagen&quot;</td><td>23.39819</td><td>21.0</td><td>28.0</td></tr><tr><td>&quot;Volvo&quot;</td><td>20.320775</td><td>18.0</td><td>26.0</td></tr><tr><td>&quot;Wallace Environmental&quot;</td><td>12.4375</td><td>11.0</td><td>15.0</td></tr><tr><td>&quot;Yugo&quot;</td><td>23.0</td><td>23.0</td><td>28.5</td></tr><tr><td>&quot;smart&quot;</td><td>68.421053</td><td>34.0</td><td>41.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 4)\n",
       "+-----------------------------+-----------+-------------+----------------+\n",
       "| make                        | mean_city | median_city | median_highway |\n",
       "| ---                         | ---       | ---         | ---            |\n",
       "| cat                         | f64       | f64         | f64            |\n",
       "+========================================================================+\n",
       "| AM General                  | 15.166667 | 14.5        | 15.0           |\n",
       "| ASC Incorporated            | 14.0      | 14.0        | 21.0           |\n",
       "| Acura                       | 20.011792 | 20.0        | 27.0           |\n",
       "| Alfa Romeo                  | 19.402062 | 19.0        | 25.0           |\n",
       "| American Motors Corporation | 16.296296 | 16.0        | 20.0           |\n",
       "| …                           | …         | …           | …              |\n",
       "| Volkswagen                  | 23.39819  | 21.0        | 28.0           |\n",
       "| Volvo                       | 20.320775 | 18.0        | 26.0           |\n",
       "| Wallace Environmental       | 12.4375   | 11.0        | 15.0           |\n",
       "| Yugo                        | 23.0      | 23.0        | 28.5           |\n",
       "| smart                       | 68.421053 | 34.0        | 41.0           |\n",
       "+-----------------------------+-----------+-------------+----------------+"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Basic operation, using whole dataset with 15 columns\n",
    "(\n",
    "    autos\n",
    "    .with_columns(pl.col('make').cast(pl.Categorical('lexical')))\n",
    "    .group_by('make')\n",
    "    .agg(\n",
    "        mean_city=pl.col('city08').mean(),\n",
    "        median_city=pl.col('city08').median(),\n",
    "        median_highway=pl.col('highway08').median()\n",
    "    )\n",
    "    .sort('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "3b3445c7-bd46-4e26-aa72-ef373ba88cae",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city</th><th>median_city</th><th>median_highway</th></tr><tr><td>cat</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>&quot;AM General&quot;</td><td>15.166667</td><td>14.5</td><td>15.0</td></tr><tr><td>&quot;ASC Incorporated&quot;</td><td>14.0</td><td>14.0</td><td>21.0</td></tr><tr><td>&quot;Acura&quot;</td><td>20.011792</td><td>20.0</td><td>27.0</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td><td>19.0</td><td>25.0</td></tr><tr><td>&quot;American Motors Corporation&quot;</td><td>16.296296</td><td>16.0</td><td>20.0</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Volkswagen&quot;</td><td>23.39819</td><td>21.0</td><td>28.0</td></tr><tr><td>&quot;Volvo&quot;</td><td>20.320775</td><td>18.0</td><td>26.0</td></tr><tr><td>&quot;Wallace Environmental&quot;</td><td>12.4375</td><td>11.0</td><td>15.0</td></tr><tr><td>&quot;Yugo&quot;</td><td>23.0</td><td>23.0</td><td>28.5</td></tr><tr><td>&quot;smart&quot;</td><td>68.421053</td><td>34.0</td><td>41.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 4)\n",
       "+-----------------------------+-----------+-------------+----------------+\n",
       "| make                        | mean_city | median_city | median_highway |\n",
       "| ---                         | ---       | ---         | ---            |\n",
       "| cat                         | f64       | f64         | f64            |\n",
       "+========================================================================+\n",
       "| AM General                  | 15.166667 | 14.5        | 15.0           |\n",
       "| ASC Incorporated            | 14.0      | 14.0        | 21.0           |\n",
       "| Acura                       | 20.011792 | 20.0        | 27.0           |\n",
       "| Alfa Romeo                  | 19.402062 | 19.0        | 25.0           |\n",
       "| American Motors Corporation | 16.296296 | 16.0        | 20.0           |\n",
       "| …                           | …         | …           | …              |\n",
       "| Volkswagen                  | 23.39819  | 21.0        | 28.0           |\n",
       "| Volvo                       | 20.320775 | 18.0        | 26.0           |\n",
       "| Wallace Environmental       | 12.4375   | 11.0        | 15.0           |\n",
       "| Yugo                        | 23.0      | 23.0        | 28.5           |\n",
       "| smart                       | 68.421053 | 34.0        | 41.0           |\n",
       "+-----------------------------+-----------+-------------+----------------+"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# variation using .select instead of .with_columns\n",
    "# and only 3 columns\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        pl.col('make').cast(pl.Categorical('lexical')),\n",
    "        pl.col('city08'),\n",
    "        pl.col('highway08')\n",
    "    )\n",
    "    .group_by('make')\n",
    "    .agg(\n",
    "        mean_city=pl.col('city08').mean(),\n",
    "        median_city=pl.col('city08').median(),\n",
    "        median_highway=pl.col('highway08').median()\n",
    "    )\n",
    "    .sort('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "1abe1661-7141-482e-a3c2-e76875ed5cd5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>city08_mean</th><th>highway08_mean</th><th>city08_min</th><th>highway08_min</th><th>city08_max</th><th>highway08_max</th></tr><tr><td>cat</td><td>f64</td><td>f64</td><td>u8</td><td>u8</td><td>u8</td><td>u8</td></tr></thead><tbody><tr><td>&quot;AM General&quot;</td><td>15.166667</td><td>15.0</td><td>13</td><td>13</td><td>18</td><td>17</td></tr><tr><td>&quot;ASC Incorporated&quot;</td><td>14.0</td><td>21.0</td><td>14</td><td>21</td><td>14</td><td>21</td></tr><tr><td>&quot;Acura&quot;</td><td>20.011792</td><td>26.816038</td><td>13</td><td>17</td><td>96</td><td>83</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td><td>26.546392</td><td>11</td><td>16</td><td>29</td><td>34</td></tr><tr><td>&quot;American Motors Corporation&quot;</td><td>16.296296</td><td>20.481481</td><td>15</td><td>17</td><td>19</td><td>23</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Volkswagen&quot;</td><td>23.39819</td><td>30.463801</td><td>11</td><td>14</td><td>126</td><td>111</td></tr><tr><td>&quot;Volvo&quot;</td><td>20.320775</td><td>26.926803</td><td>13</td><td>19</td><td>127</td><td>104</td></tr><tr><td>&quot;Wallace Environmental&quot;</td><td>12.4375</td><td>16.0</td><td>8</td><td>13</td><td>17</td><td>22</td></tr><tr><td>&quot;Yugo&quot;</td><td>23.0</td><td>28.25</td><td>21</td><td>27</td><td>24</td><td>29</td></tr><tr><td>&quot;smart&quot;</td><td>68.421053</td><td>60.973684</td><td>31</td><td>38</td><td>124</td><td>94</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 7)\n",
       "+--------------+-------------+--------------+------------+--------------+------------+-------------+\n",
       "| make         | city08_mean | highway08_me | city08_min | highway08_mi | city08_max | highway08_m |\n",
       "| ---          | ---         | an           | ---        | n            | ---        | ax          |\n",
       "| cat          | f64         | ---          | u8         | ---          | u8         | ---         |\n",
       "|              |             | f64          |            | u8           |            | u8          |\n",
       "+==================================================================================================+\n",
       "| AM General   | 15.166667   | 15.0         | 13         | 13           | 18         | 17          |\n",
       "| ASC          | 14.0        | 21.0         | 14         | 21           | 14         | 21          |\n",
       "| Incorporated |             |              |            |              |            |             |\n",
       "| Acura        | 20.011792   | 26.816038    | 13         | 17           | 96         | 83          |\n",
       "| Alfa Romeo   | 19.402062   | 26.546392    | 11         | 16           | 29         | 34          |\n",
       "| American     | 16.296296   | 20.481481    | 15         | 17           | 19         | 23          |\n",
       "| Motors       |             |              |            |              |            |             |\n",
       "| Corporation  |             |              |            |              |            |             |\n",
       "| …            | …           | …            | …          | …            | …          | …           |\n",
       "| Volkswagen   | 23.39819    | 30.463801    | 11         | 14           | 126        | 111         |\n",
       "| Volvo        | 20.320775   | 26.926803    | 13         | 19           | 127        | 104         |\n",
       "| Wallace Envi | 12.4375     | 16.0         | 8          | 13           | 17         | 22          |\n",
       "| ronmental    |             |              |            |              |            |             |\n",
       "| Yugo         | 23.0        | 28.25        | 21         | 27           | 24         | 29          |\n",
       "| smart        | 68.421053   | 60.973684    | 31         | 38           | 124        | 94          |\n",
       "+--------------+-------------+--------------+------------+--------------+------------+-------------+"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# variation using .select instead of .with_columns\n",
    "# and pl.all() with name.suffix()\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        pl.col('make').cast(pl.Categorical('lexical')),\n",
    "        pl.col('city08'),\n",
    "        pl.col('highway08')\n",
    "    )\n",
    "    .group_by('make')\n",
    "    .agg(\n",
    "        pl.all().mean().name.suffix('_mean'),\n",
    "        pl.all().min().name.suffix('_min'),\n",
    "        pl.all().max().name.suffix('_max')\n",
    "    )\n",
    "    .sort('make')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "141cfd93-a7b6-41fe-b500-b8f6a43519d9",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.7 Grouping by Multiple Columns\n",
    "[back to Top](#Top)\n",
    "\n",
    "* We can also group by multiple columns.\n",
    "* Here, we group by the *make* and *year* columns.\n",
    "* We then calculate each group's average *city08* value.\n",
    "* We use `.cast` to convert *make* to an alphabetically ordered category (via the 'lexical' argument)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "c42414fb-55b3-4762-836e-83c6a94e590b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "estimated unique values: 2223\n",
      "estimated unique count: 2223 exceeded the boundary: 1000, running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (2_016, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>year</th><th>mean_city08</th></tr><tr><td>cat</td><td>i16</td><td>f64</td></tr></thead><tbody><tr><td>&quot;AM General&quot;</td><td>1984</td><td>15.5</td></tr><tr><td>&quot;AM General&quot;</td><td>1985</td><td>14.5</td></tr><tr><td>&quot;ASC Incorporated&quot;</td><td>1987</td><td>14.0</td></tr><tr><td>&quot;Acura&quot;</td><td>1986</td><td>20.0</td></tr><tr><td>&quot;Acura&quot;</td><td>1987</td><td>18.833333</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;smart&quot;</td><td>2015</td><td>78.0</td></tr><tr><td>&quot;smart&quot;</td><td>2016</td><td>77.0</td></tr><tr><td>&quot;smart&quot;</td><td>2017</td><td>60.666667</td></tr><tr><td>&quot;smart&quot;</td><td>2018</td><td>118.0</td></tr><tr><td>&quot;smart&quot;</td><td>2019</td><td>118.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (2_016, 3)\n",
       "+------------------+------+-------------+\n",
       "| make             | year | mean_city08 |\n",
       "| ---              | ---  | ---         |\n",
       "| cat              | i16  | f64         |\n",
       "+=======================================+\n",
       "| AM General       | 1984 | 15.5        |\n",
       "| AM General       | 1985 | 14.5        |\n",
       "| ASC Incorporated | 1987 | 14.0        |\n",
       "| Acura            | 1986 | 20.0        |\n",
       "| Acura            | 1987 | 18.833333   |\n",
       "| …                | …    | …           |\n",
       "| smart            | 2015 | 78.0        |\n",
       "| smart            | 2016 | 77.0        |\n",
       "| smart            | 2017 | 60.666667   |\n",
       "| smart            | 2018 | 118.0       |\n",
       "| smart            | 2019 | 118.0       |\n",
       "+------------------+------+-------------+"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .with_columns(pl.col('make').cast(pl.Categorical('lexical')))\n",
    "    .group_by(['make', 'year'])\n",
    "    .agg(mean_city08=pl.col('city08').mean())\n",
    "    .sort(['make', 'year'])\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf1776a8-8a6f-4680-b367-c11616b03d97",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.8 Pivoting with Multiple Aggregations\n",
    "[back to Top](#Top)\n",
    "\n",
    "\n",
    "* `.pivot()` can only be used with a single aggregation.\n",
    "* If we want to use multiple aggregations, we need to use `.group_by()` and then use `.pivot()` on the result of the groupby.\n",
    "\n",
    "  \n",
    "Here, we calculate the min and max *city08* values by year for the top-3 makes.\n",
    "\n",
    "* First, calculate the top-three makes.\n",
    "* This is akin to Pandas value counts.\n",
    "* Do this by grouping by the *make* column and aggregating with `.len()`.\n",
    "* We need to sort the data by the aggregated column before taking the top-3 rows.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "2f07b04b-90bb-4642-bad3-4848aafd28e6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>len</th></tr><tr><td>cat</td><td>u32</td></tr></thead><tbody><tr><td>&quot;Chevrolet&quot;</td><td>4472</td></tr><tr><td>&quot;Ford&quot;</td><td>3834</td></tr><tr><td>&quot;GMC&quot;</td><td>2823</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 2)\n",
       "+-----------+------+\n",
       "| make      | len  |\n",
       "| ---       | ---  |\n",
       "| cat       | u32  |\n",
       "+==================+\n",
       "| Chevrolet | 4472 |\n",
       "| Ford      | 3834 |\n",
       "| GMC       | 2823 |\n",
       "+-----------+------+"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# equivalent to Pandas .value_counts\n",
    "top3 = (\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .len()\n",
    "    .sort(by=['len'], descending=True)\n",
    "    .head(3)\n",
    ")\n",
    "top3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "edbc6f15-cd80-493f-af8d-df8fc303f75c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "make\n",
       "Chevrolet    4472\n",
       "Ford         3834\n",
       "GMC          2823\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pandas version with value_counts()\n",
    "autos_pandas = autos.to_pandas()\n",
    "autos_pandas['make'].value_counts(ascending=False).head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2553e93b-82a7-4fb3-bc88-570ec1adb4b9",
   "metadata": {},
   "source": [
    "---\n",
    "Now, we can do the groupby operation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "f60bccf3-d83a-4198-b893-89abc135a493",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n",
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (126, 5)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>src_city08</th><th>min_city08</th><th>max_city08</th></tr><tr><td>i16</td><td>cat</td><td>list[u8]</td><td>u8</td><td>u8</td></tr></thead><tbody><tr><td>1984</td><td>&quot;Chevrolet&quot;</td><td>[10, 10, … 33]</td><td>10</td><td>33</td></tr><tr><td>1984</td><td>&quot;Ford&quot;</td><td>[9, 9, … 35]</td><td>9</td><td>35</td></tr><tr><td>1984</td><td>&quot;GMC&quot;</td><td>[10, 10, … 26]</td><td>10</td><td>26</td></tr><tr><td>1985</td><td>&quot;Chevrolet&quot;</td><td>[9, 10, … 39]</td><td>9</td><td>39</td></tr><tr><td>1985</td><td>&quot;Ford&quot;</td><td>[9, 10, … 36]</td><td>9</td><td>36</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>2024</td><td>&quot;Ford&quot;</td><td>[10, 14, … 111]</td><td>10</td><td>111</td></tr><tr><td>2024</td><td>&quot;GMC&quot;</td><td>[14, 14, … 59]</td><td>14</td><td>59</td></tr><tr><td>2025</td><td>&quot;Chevrolet&quot;</td><td>[12, 12, … 102]</td><td>12</td><td>102</td></tr><tr><td>2025</td><td>&quot;Ford&quot;</td><td>[18, 18, … 42]</td><td>18</td><td>42</td></tr><tr><td>2025</td><td>&quot;GMC&quot;</td><td>[14, 14, … 59]</td><td>14</td><td>59</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (126, 5)\n",
       "+------+-----------+-----------------+------------+------------+\n",
       "| year | make      | src_city08      | min_city08 | max_city08 |\n",
       "| ---  | ---       | ---             | ---        | ---        |\n",
       "| i16  | cat       | list[u8]        | u8         | u8         |\n",
       "+==============================================================+\n",
       "| 1984 | Chevrolet | [10, 10, … 33]  | 10         | 33         |\n",
       "| 1984 | Ford      | [9, 9, … 35]    | 9          | 35         |\n",
       "| 1984 | GMC       | [10, 10, … 26]  | 10         | 26         |\n",
       "| 1985 | Chevrolet | [9, 10, … 39]   | 9          | 39         |\n",
       "| 1985 | Ford      | [9, 10, … 36]   | 9          | 36         |\n",
       "| …    | …         | …               | …          | …          |\n",
       "| 2024 | Ford      | [10, 14, … 111] | 10         | 111        |\n",
       "| 2024 | GMC       | [14, 14, … 59]  | 14         | 59         |\n",
       "| 2025 | Chevrolet | [12, 12, … 102] | 12         | 102        |\n",
       "| 2025 | Ford      | [18, 18, … 42]  | 18         | 42         |\n",
       "| 2025 | GMC       | [14, 14, … 59]  | 14         | 59         |\n",
       "+------+-----------+-----------------+------------+------------+"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .filter(pl.col('make').is_in(top3['make']))\n",
    "    .group_by(['year', 'make'])\n",
    "    .agg(\n",
    "        src_city08=pl.col('city08').sort(),\n",
    "        min_city08=pl.col('city08').min(),\n",
    "        max_city08=pl.col('city08').max(),\n",
    "    )\n",
    "    .sort(['year', 'make'])\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25040d4f-5054-44ae-88a4-72d05951c264",
   "metadata": {},
   "source": [
    "If we want the *make* column to be pulled into the columns, we can use `.pivot()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "bdfc2327-2d82-435a-968f-ae5d28c39c33",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'polars.dataframe.frame.DataFrame'>\n",
      "<class 'polars.series.series.Series'>\n"
     ]
    }
   ],
   "source": [
    "print(type(top3))\n",
    "print(type(top3['make']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "4577230b-42b2-4880-99e1-1e156976313e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n",
      "estimated unique values: 5\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (126, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>min_city08</th><th>max_city08</th></tr><tr><td>i16</td><td>cat</td><td>u8</td><td>u8</td></tr></thead><tbody><tr><td>1984</td><td>&quot;Chevrolet&quot;</td><td>10</td><td>33</td></tr><tr><td>1984</td><td>&quot;Ford&quot;</td><td>9</td><td>35</td></tr><tr><td>1984</td><td>&quot;GMC&quot;</td><td>10</td><td>26</td></tr><tr><td>1985</td><td>&quot;Chevrolet&quot;</td><td>9</td><td>39</td></tr><tr><td>1985</td><td>&quot;Ford&quot;</td><td>9</td><td>36</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>2024</td><td>&quot;Ford&quot;</td><td>10</td><td>111</td></tr><tr><td>2024</td><td>&quot;GMC&quot;</td><td>14</td><td>59</td></tr><tr><td>2025</td><td>&quot;Chevrolet&quot;</td><td>12</td><td>102</td></tr><tr><td>2025</td><td>&quot;Ford&quot;</td><td>18</td><td>42</td></tr><tr><td>2025</td><td>&quot;GMC&quot;</td><td>14</td><td>59</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (126, 4)\n",
       "+------+-----------+------------+------------+\n",
       "| year | make      | min_city08 | max_city08 |\n",
       "| ---  | ---       | ---        | ---        |\n",
       "| i16  | cat       | u8         | u8         |\n",
       "+============================================+\n",
       "| 1984 | Chevrolet | 10         | 33         |\n",
       "| 1984 | Ford      | 9          | 35         |\n",
       "| 1984 | GMC       | 10         | 26         |\n",
       "| 1985 | Chevrolet | 9          | 39         |\n",
       "| 1985 | Ford      | 9          | 36         |\n",
       "| …    | …         | …          | …          |\n",
       "| 2024 | Ford      | 10         | 111        |\n",
       "| 2024 | GMC       | 14         | 59         |\n",
       "| 2025 | Chevrolet | 12         | 102        |\n",
       "| 2025 | Ford      | 18         | 42         |\n",
       "| 2025 | GMC       | 14         | 59         |\n",
       "+------+-----------+------------+------------+"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .filter(pl.col('make').is_in(top3['make']))\n",
    "    .group_by(['year', 'make'])\n",
    "    .agg(\n",
    "        min_city08=pl.col('city08').min(),\n",
    "        max_city08=pl.col('city08').max(),\n",
    "    )\n",
    "    .sort(['year', 'make'])\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "912f0bc5-7b16-4947-b696-7258b0f46314",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n",
      "estimated unique values: 5\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (42, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>min_city08_Chevrolet</th><th>min_city08_Ford</th><th>min_city08_GMC</th><th>max_city08_Chevrolet</th><th>max_city08_Ford</th><th>max_city08_GMC</th></tr><tr><td>i16</td><td>u8</td><td>u8</td><td>u8</td><td>u8</td><td>u8</td><td>u8</td></tr></thead><tbody><tr><td>1984</td><td>10</td><td>9</td><td>10</td><td>33</td><td>35</td><td>26</td></tr><tr><td>1985</td><td>9</td><td>9</td><td>9</td><td>39</td><td>36</td><td>26</td></tr><tr><td>1986</td><td>9</td><td>10</td><td>9</td><td>44</td><td>34</td><td>21</td></tr><tr><td>1987</td><td>11</td><td>10</td><td>11</td><td>44</td><td>31</td><td>20</td></tr><tr><td>1988</td><td>11</td><td>10</td><td>11</td><td>44</td><td>33</td><td>21</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>2021</td><td>13</td><td>12</td><td>13</td><td>127</td><td>108</td><td>25</td></tr><tr><td>2022</td><td>13</td><td>12</td><td>13</td><td>131</td><td>110</td><td>25</td></tr><tr><td>2023</td><td>12</td><td>10</td><td>14</td><td>131</td><td>110</td><td>24</td></tr><tr><td>2024</td><td>12</td><td>10</td><td>14</td><td>117</td><td>111</td><td>59</td></tr><tr><td>2025</td><td>12</td><td>18</td><td>14</td><td>102</td><td>42</td><td>59</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (42, 7)\n",
       "+------+---------------+---------------+--------------+--------------+--------------+--------------+\n",
       "| year | min_city08_Ch | min_city08_Fo | min_city08_G | max_city08_C | max_city08_F | max_city08_G |\n",
       "| ---  | evrolet       | rd            | MC           | hevrolet     | ord          | MC           |\n",
       "| i16  | ---           | ---           | ---          | ---          | ---          | ---          |\n",
       "|      | u8            | u8            | u8           | u8           | u8           | u8           |\n",
       "+==================================================================================================+\n",
       "| 1984 | 10            | 9             | 10           | 33           | 35           | 26           |\n",
       "| 1985 | 9             | 9             | 9            | 39           | 36           | 26           |\n",
       "| 1986 | 9             | 10            | 9            | 44           | 34           | 21           |\n",
       "| 1987 | 11            | 10            | 11           | 44           | 31           | 20           |\n",
       "| 1988 | 11            | 10            | 11           | 44           | 33           | 21           |\n",
       "| …    | …             | …             | …            | …            | …            | …            |\n",
       "| 2021 | 13            | 12            | 13           | 127          | 108          | 25           |\n",
       "| 2022 | 13            | 12            | 13           | 131          | 110          | 25           |\n",
       "| 2023 | 12            | 10            | 14           | 131          | 110          | 24           |\n",
       "| 2024 | 12            | 10            | 14           | 117          | 111          | 59           |\n",
       "| 2025 | 12            | 18            | 14           | 102          | 42           | 59           |\n",
       "+------+---------------+---------------+--------------+--------------+--------------+--------------+"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .filter(pl.col('make').is_in(top3['make']))\n",
    "    .group_by(['year', 'make'])\n",
    "    .agg(\n",
    "        min_city08=pl.col('city08').min(),\n",
    "        max_city08=pl.col('city08').max(),\n",
    "    )\n",
    "    .sort(['year', 'make'])\n",
    "    .pivot(\n",
    "        index='year',\n",
    "        on='make',\n",
    "        values=['min_city08', 'max_city08']\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "43f21112-7920-4754-a96f-8090f3133ac7",
   "metadata": {},
   "source": [
    "---\n",
    "Using Pandas for plotting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "b3202273-96fc-4eb9-aa86-f66eb0b2a335",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n",
      "estimated unique values: 5\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 432,
       "width": 552
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "(\n",
    "    autos\n",
    "    .filter(pl.col('make').is_in(top3['make']))\n",
    "    .group_by(['year', 'make'])\n",
    "    .agg(\n",
    "        min_city08=pl.col('city08').min(),\n",
    "        max_city08=pl.col('city08').max(),\n",
    "    )\n",
    "    .sort(['year', 'make'])\n",
    "    .pivot(\n",
    "        index='year',\n",
    "        on='make',\n",
    "        values=['min_city08', 'max_city08']\n",
    "    )\n",
    "    .to_pandas()\n",
    "    .set_index('year')\n",
    "    .plot(\n",
    "        ax=ax,\n",
    "        color=['#55c667', '#2f8738', '#2c4279', '#141f38', '#d4e116', '#afba12']\n",
    "    )\n",
    ")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7598d62-df36-4084-b15a-43e533511ec8",
   "metadata": {},
   "source": [
    "---\n",
    "Using hvplot to plot in Polars."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "dafb2671-f823-4579-ad4c-3cde967ce05c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "(function(root) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  var force = true;\n",
       "  var py_version = '3.4.3'.replace('rc', '-rc.').replace('.dev', '-dev.');\n",
       "  var reloading = false;\n",
       "  var Bokeh = root.Bokeh;\n",
       "\n",
       "  if (typeof (root._bokeh_timeout) === \"undefined\" || force) {\n",
       "    root._bokeh_timeout = Date.now() + 5000;\n",
       "    root._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  function run_callbacks() {\n",
       "    try {\n",
       "      root._bokeh_onload_callbacks.forEach(function(callback) {\n",
       "        if (callback != null)\n",
       "          callback();\n",
       "      });\n",
       "    } finally {\n",
       "      delete root._bokeh_onload_callbacks;\n",
       "    }\n",
       "    console.debug(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(css_urls, js_urls, js_modules, js_exports, callback) {\n",
       "    if (css_urls == null) css_urls = [];\n",
       "    if (js_urls == null) js_urls = [];\n",
       "    if (js_modules == null) js_modules = [];\n",
       "    if (js_exports == null) js_exports = {};\n",
       "\n",
       "    root._bokeh_onload_callbacks.push(callback);\n",
       "\n",
       "    if (root._bokeh_is_loading > 0) {\n",
       "      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls.length === 0 && js_modules.length === 0 && Object.keys(js_exports).length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    if (!reloading) {\n",
       "      console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "    }\n",
       "\n",
       "    function on_load() {\n",
       "      root._bokeh_is_loading--;\n",
       "      if (root._bokeh_is_loading === 0) {\n",
       "        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n",
       "        run_callbacks()\n",
       "      }\n",
       "    }\n",
       "    window._bokeh_on_load = on_load\n",
       "\n",
       "    function on_error() {\n",
       "      console.error(\"failed to load \" + url);\n",
       "    }\n",
       "\n",
       "    var skip = [];\n",
       "    if (window.requirejs) {\n",
       "      window.requirejs.config({'packages': {}, 'paths': {}, 'shim': {}});\n",
       "      root._bokeh_is_loading = css_urls.length + 0;\n",
       "    } else {\n",
       "      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length + Object.keys(js_exports).length;\n",
       "    }\n",
       "\n",
       "    var existing_stylesheets = []\n",
       "    var links = document.getElementsByTagName('link')\n",
       "    for (var i = 0; i < links.length; i++) {\n",
       "      var link = links[i]\n",
       "      if (link.href != null) {\n",
       "\texisting_stylesheets.push(link.href)\n",
       "      }\n",
       "    }\n",
       "    for (var i = 0; i < css_urls.length; i++) {\n",
       "      var url = css_urls[i];\n",
       "      if (existing_stylesheets.indexOf(url) !== -1) {\n",
       "\ton_load()\n",
       "\tcontinue;\n",
       "      }\n",
       "      const element = document.createElement(\"link\");\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.rel = \"stylesheet\";\n",
       "      element.type = \"text/css\";\n",
       "      element.href = url;\n",
       "      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n",
       "      document.body.appendChild(element);\n",
       "    }    var existing_scripts = []\n",
       "    var scripts = document.getElementsByTagName('script')\n",
       "    for (var i = 0; i < scripts.length; i++) {\n",
       "      var script = scripts[i]\n",
       "      if (script.src != null) {\n",
       "\texisting_scripts.push(script.src)\n",
       "      }\n",
       "    }\n",
       "    for (var i = 0; i < js_urls.length; i++) {\n",
       "      var url = js_urls[i];\n",
       "      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (var i = 0; i < js_modules.length; i++) {\n",
       "      var url = js_modules[i];\n",
       "      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (const name in js_exports) {\n",
       "      var url = js_exports[name];\n",
       "      if (skip.indexOf(url) >= 0 || root[name] != null) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      element.textContent = `\n",
       "      import ${name} from \"${url}\"\n",
       "      window.${name} = ${name}\n",
       "      window._bokeh_on_load()\n",
       "      `\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    if (!js_urls.length && !js_modules.length) {\n",
       "      on_load()\n",
       "    }\n",
       "  };\n",
       "\n",
       "  function inject_raw_css(css) {\n",
       "    const element = document.createElement(\"style\");\n",
       "    element.appendChild(document.createTextNode(css));\n",
       "    document.body.appendChild(element);\n",
       "  }\n",
       "\n",
       "  var js_urls = [\"https://cdn.bokeh.org/bokeh/release/bokeh-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-3.4.3.min.js\", \"https://cdn.holoviz.org/panel/1.4.5/dist/panel.min.js\"];\n",
       "  var js_modules = [];\n",
       "  var js_exports = {};\n",
       "  var css_urls = [];\n",
       "  var inline_js = [    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "function(Bokeh) {} // ensure no trailing comma for IE\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    if ((root.Bokeh !== undefined) || (force === true)) {\n",
       "      for (var i = 0; i < inline_js.length; i++) {\n",
       "\ttry {\n",
       "          inline_js[i].call(root, root.Bokeh);\n",
       "\t} catch(e) {\n",
       "\t  if (!reloading) {\n",
       "\t    throw e;\n",
       "\t  }\n",
       "\t}\n",
       "      }\n",
       "      // Cache old bokeh versions\n",
       "      if (Bokeh != undefined && !reloading) {\n",
       "\tvar NewBokeh = root.Bokeh;\n",
       "\tif (Bokeh.versions === undefined) {\n",
       "\t  Bokeh.versions = new Map();\n",
       "\t}\n",
       "\tif (NewBokeh.version !== Bokeh.version) {\n",
       "\t  Bokeh.versions.set(NewBokeh.version, NewBokeh)\n",
       "\t}\n",
       "\troot.Bokeh = Bokeh;\n",
       "      }} else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!root._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      root._bokeh_failed_load = true;\n",
       "    }\n",
       "    root._bokeh_is_initializing = false\n",
       "  }\n",
       "\n",
       "  function load_or_wait() {\n",
       "    // Implement a backoff loop that tries to ensure we do not load multiple\n",
       "    // versions of Bokeh and its dependencies at the same time.\n",
       "    // In recent versions we use the root._bokeh_is_initializing flag\n",
       "    // to determine whether there is an ongoing attempt to initialize\n",
       "    // bokeh, however for backward compatibility we also try to ensure\n",
       "    // that we do not start loading a newer (Panel>=1.0 and Bokeh>3) version\n",
       "    // before older versions are fully initialized.\n",
       "    if (root._bokeh_is_initializing && Date.now() > root._bokeh_timeout) {\n",
       "      root._bokeh_is_initializing = false;\n",
       "      root._bokeh_onload_callbacks = undefined;\n",
       "      console.log(\"Bokeh: BokehJS was loaded multiple times but one version failed to initialize.\");\n",
       "      load_or_wait();\n",
       "    } else if (root._bokeh_is_initializing || (typeof root._bokeh_is_initializing === \"undefined\" && root._bokeh_onload_callbacks !== undefined)) {\n",
       "      setTimeout(load_or_wait, 100);\n",
       "    } else {\n",
       "      root._bokeh_is_initializing = true\n",
       "      root._bokeh_onload_callbacks = []\n",
       "      var bokeh_loaded = Bokeh != null && (Bokeh.version === py_version || (Bokeh.versions !== undefined && Bokeh.versions.has(py_version)));\n",
       "      if (!reloading && !bokeh_loaded) {\n",
       "\troot.Bokeh = undefined;\n",
       "      }\n",
       "      load_libs(css_urls, js_urls, js_modules, js_exports, function() {\n",
       "\tconsole.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "\trun_inline_js();\n",
       "      });\n",
       "    }\n",
       "  }\n",
       "  // Give older versions of the autoload script a head-start to ensure\n",
       "  // they initialize before we start loading newer version.\n",
       "  setTimeout(load_or_wait, 100)\n",
       "}(window));"
      ],
      "application/vnd.holoviews_load.v0+json": "(function(root) {\n  function now() {\n    return new Date();\n  }\n\n  var force = true;\n  var py_version = '3.4.3'.replace('rc', '-rc.').replace('.dev', '-dev.');\n  var reloading = false;\n  var Bokeh = root.Bokeh;\n\n  if (typeof (root._bokeh_timeout) === \"undefined\" || force) {\n    root._bokeh_timeout = Date.now() + 5000;\n    root._bokeh_failed_load = false;\n  }\n\n  function run_callbacks() {\n    try {\n      root._bokeh_onload_callbacks.forEach(function(callback) {\n        if (callback != null)\n          callback();\n      });\n    } finally {\n      delete root._bokeh_onload_callbacks;\n    }\n    console.debug(\"Bokeh: all callbacks have finished\");\n  }\n\n  function load_libs(css_urls, js_urls, js_modules, js_exports, callback) {\n    if (css_urls == null) css_urls = [];\n    if (js_urls == null) js_urls = [];\n    if (js_modules == null) js_modules = [];\n    if (js_exports == null) js_exports = {};\n\n    root._bokeh_onload_callbacks.push(callback);\n\n    if (root._bokeh_is_loading > 0) {\n      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n      return null;\n    }\n    if (js_urls.length === 0 && js_modules.length === 0 && Object.keys(js_exports).length === 0) {\n      run_callbacks();\n      return null;\n    }\n    if (!reloading) {\n      console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n    }\n\n    function on_load() {\n      root._bokeh_is_loading--;\n      if (root._bokeh_is_loading === 0) {\n        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n        run_callbacks()\n      }\n    }\n    window._bokeh_on_load = on_load\n\n    function on_error() {\n      console.error(\"failed to load \" + url);\n    }\n\n    var skip = [];\n    if (window.requirejs) {\n      window.requirejs.config({'packages': {}, 'paths': {}, 'shim': {}});\n      root._bokeh_is_loading = css_urls.length + 0;\n    } else {\n      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length + Object.keys(js_exports).length;\n    }\n\n    var existing_stylesheets = []\n    var links = document.getElementsByTagName('link')\n    for (var i = 0; i < links.length; i++) {\n      var link = links[i]\n      if (link.href != null) {\n\texisting_stylesheets.push(link.href)\n      }\n    }\n    for (var i = 0; i < css_urls.length; i++) {\n      var url = css_urls[i];\n      if (existing_stylesheets.indexOf(url) !== -1) {\n\ton_load()\n\tcontinue;\n      }\n      const element = document.createElement(\"link\");\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.rel = \"stylesheet\";\n      element.type = \"text/css\";\n      element.href = url;\n      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n      document.body.appendChild(element);\n    }    var existing_scripts = []\n    var scripts = document.getElementsByTagName('script')\n    for (var i = 0; i < scripts.length; i++) {\n      var script = scripts[i]\n      if (script.src != null) {\n\texisting_scripts.push(script.src)\n      }\n    }\n    for (var i = 0; i < js_urls.length; i++) {\n      var url = js_urls[i];\n      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (var i = 0; i < js_modules.length; i++) {\n      var url = js_modules[i];\n      if (skip.indexOf(url) !== -1 || existing_scripts.indexOf(url) !== -1) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (const name in js_exports) {\n      var url = js_exports[name];\n      if (skip.indexOf(url) >= 0 || root[name] != null) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onerror = on_error;\n      element.async = false;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      element.textContent = `\n      import ${name} from \"${url}\"\n      window.${name} = ${name}\n      window._bokeh_on_load()\n      `\n      document.head.appendChild(element);\n    }\n    if (!js_urls.length && !js_modules.length) {\n      on_load()\n    }\n  };\n\n  function inject_raw_css(css) {\n    const element = document.createElement(\"style\");\n    element.appendChild(document.createTextNode(css));\n    document.body.appendChild(element);\n  }\n\n  var js_urls = [\"https://cdn.bokeh.org/bokeh/release/bokeh-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-3.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-3.4.3.min.js\", \"https://cdn.holoviz.org/panel/1.4.5/dist/panel.min.js\"];\n  var js_modules = [];\n  var js_exports = {};\n  var css_urls = [];\n  var inline_js = [    function(Bokeh) {\n      Bokeh.set_log_level(\"info\");\n    },\nfunction(Bokeh) {} // ensure no trailing comma for IE\n  ];\n\n  function run_inline_js() {\n    if ((root.Bokeh !== undefined) || (force === true)) {\n      for (var i = 0; i < inline_js.length; i++) {\n\ttry {\n          inline_js[i].call(root, root.Bokeh);\n\t} catch(e) {\n\t  if (!reloading) {\n\t    throw e;\n\t  }\n\t}\n      }\n      // Cache old bokeh versions\n      if (Bokeh != undefined && !reloading) {\n\tvar NewBokeh = root.Bokeh;\n\tif (Bokeh.versions === undefined) {\n\t  Bokeh.versions = new Map();\n\t}\n\tif (NewBokeh.version !== Bokeh.version) {\n\t  Bokeh.versions.set(NewBokeh.version, NewBokeh)\n\t}\n\troot.Bokeh = Bokeh;\n      }} else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(run_inline_js, 100);\n    } else if (!root._bokeh_failed_load) {\n      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n      root._bokeh_failed_load = true;\n    }\n    root._bokeh_is_initializing = false\n  }\n\n  function load_or_wait() {\n    // Implement a backoff loop that tries to ensure we do not load multiple\n    // versions of Bokeh and its dependencies at the same time.\n    // In recent versions we use the root._bokeh_is_initializing flag\n    // to determine whether there is an ongoing attempt to initialize\n    // bokeh, however for backward compatibility we also try to ensure\n    // that we do not start loading a newer (Panel>=1.0 and Bokeh>3) version\n    // before older versions are fully initialized.\n    if (root._bokeh_is_initializing && Date.now() > root._bokeh_timeout) {\n      root._bokeh_is_initializing = false;\n      root._bokeh_onload_callbacks = undefined;\n      console.log(\"Bokeh: BokehJS was loaded multiple times but one version failed to initialize.\");\n      load_or_wait();\n    } else if (root._bokeh_is_initializing || (typeof root._bokeh_is_initializing === \"undefined\" && root._bokeh_onload_callbacks !== undefined)) {\n      setTimeout(load_or_wait, 100);\n    } else {\n      root._bokeh_is_initializing = true\n      root._bokeh_onload_callbacks = []\n      var bokeh_loaded = Bokeh != null && (Bokeh.version === py_version || (Bokeh.versions !== undefined && Bokeh.versions.has(py_version)));\n      if (!reloading && !bokeh_loaded) {\n\troot.Bokeh = undefined;\n      }\n      load_libs(css_urls, js_urls, js_modules, js_exports, function() {\n\tconsole.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n\trun_inline_js();\n      });\n    }\n  }\n  // Give older versions of the autoload script a head-start to ensure\n  // they initialize before we start loading newer version.\n  setTimeout(load_or_wait, 100)\n}(window));"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "if ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n",
       "  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n",
       "}\n",
       "\n",
       "\n",
       "    function JupyterCommManager() {\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n",
       "      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        comm_manager.register_target(comm_id, function(comm) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        });\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        });\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n",
       "          var messages = comm.messages[Symbol.asyncIterator]();\n",
       "          function processIteratorResult(result) {\n",
       "            var message = result.value;\n",
       "            console.log(message)\n",
       "            var content = {data: message.data, comm_id};\n",
       "            var buffers = []\n",
       "            for (var buffer of message.buffers || []) {\n",
       "              buffers.push(new DataView(buffer))\n",
       "            }\n",
       "            var metadata = message.metadata || {};\n",
       "            var msg = {content, buffers, metadata}\n",
       "            msg_handler(msg);\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "          return messages.next().then(processIteratorResult);\n",
       "        })\n",
       "      }\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n",
       "      if (comm_id in window.PyViz.comms) {\n",
       "        return window.PyViz.comms[comm_id];\n",
       "      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n",
       "        if (msg_handler) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        }\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n",
       "        comm.open();\n",
       "        if (msg_handler) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        }\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        var comm_promise = google.colab.kernel.comms.open(comm_id)\n",
       "        comm_promise.then((comm) => {\n",
       "          window.PyViz.comms[comm_id] = comm;\n",
       "          if (msg_handler) {\n",
       "            var messages = comm.messages[Symbol.asyncIterator]();\n",
       "            function processIteratorResult(result) {\n",
       "              var message = result.value;\n",
       "              var content = {data: message.data};\n",
       "              var metadata = message.metadata || {comm_id};\n",
       "              var msg = {content, metadata}\n",
       "              msg_handler(msg);\n",
       "              return messages.next().then(processIteratorResult);\n",
       "            }\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "        }) \n",
       "        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n",
       "          return comm_promise.then((comm) => {\n",
       "            comm.send(data, metadata, buffers, disposeOnDone);\n",
       "          });\n",
       "        };\n",
       "        var comm = {\n",
       "          send: sendClosure\n",
       "        };\n",
       "      }\n",
       "      window.PyViz.comms[comm_id] = comm;\n",
       "      return comm;\n",
       "    }\n",
       "    window.PyViz.comm_manager = new JupyterCommManager();\n",
       "    \n",
       "\n",
       "\n",
       "var JS_MIME_TYPE = 'application/javascript';\n",
       "var HTML_MIME_TYPE = 'text/html';\n",
       "var EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\n",
       "var CLASS_NAME = 'output';\n",
       "\n",
       "/**\n",
       " * Render data to the DOM node\n",
       " */\n",
       "function render(props, node) {\n",
       "  var div = document.createElement(\"div\");\n",
       "  var script = document.createElement(\"script\");\n",
       "  node.appendChild(div);\n",
       "  node.appendChild(script);\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when a new output is added\n",
       " */\n",
       "function handle_add_output(event, handle) {\n",
       "  var output_area = handle.output_area;\n",
       "  var output = handle.output;\n",
       "  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n",
       "    return\n",
       "  }\n",
       "  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n",
       "  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n",
       "  if (id !== undefined) {\n",
       "    var nchildren = toinsert.length;\n",
       "    var html_node = toinsert[nchildren-1].children[0];\n",
       "    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var scripts = [];\n",
       "    var nodelist = html_node.querySelectorAll(\"script\");\n",
       "    for (var i in nodelist) {\n",
       "      if (nodelist.hasOwnProperty(i)) {\n",
       "        scripts.push(nodelist[i])\n",
       "      }\n",
       "    }\n",
       "\n",
       "    scripts.forEach( function (oldScript) {\n",
       "      var newScript = document.createElement(\"script\");\n",
       "      var attrs = [];\n",
       "      var nodemap = oldScript.attributes;\n",
       "      for (var j in nodemap) {\n",
       "        if (nodemap.hasOwnProperty(j)) {\n",
       "          attrs.push(nodemap[j])\n",
       "        }\n",
       "      }\n",
       "      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n",
       "      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n",
       "      oldScript.parentNode.replaceChild(newScript, oldScript);\n",
       "    });\n",
       "    if (JS_MIME_TYPE in output.data) {\n",
       "      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n",
       "    }\n",
       "    output_area._hv_plot_id = id;\n",
       "    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n",
       "      window.PyViz.plot_index[id] = Bokeh.index[id];\n",
       "    } else {\n",
       "      window.PyViz.plot_index[id] = null;\n",
       "    }\n",
       "  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n",
       "    var bk_div = document.createElement(\"div\");\n",
       "    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var script_attrs = bk_div.children[0].attributes;\n",
       "    for (var i = 0; i < script_attrs.length; i++) {\n",
       "      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n",
       "    }\n",
       "    // store reference to server id on output_area\n",
       "    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when an output is cleared or removed\n",
       " */\n",
       "function handle_clear_output(event, handle) {\n",
       "  var id = handle.cell.output_area._hv_plot_id;\n",
       "  var server_id = handle.cell.output_area._bokeh_server_id;\n",
       "  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n",
       "  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n",
       "  if (server_id !== null) {\n",
       "    comm.send({event_type: 'server_delete', 'id': server_id});\n",
       "    return;\n",
       "  } else if (comm !== null) {\n",
       "    comm.send({event_type: 'delete', 'id': id});\n",
       "  }\n",
       "  delete PyViz.plot_index[id];\n",
       "  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n",
       "    var doc = window.Bokeh.index[id].model.document\n",
       "    doc.clear();\n",
       "    const i = window.Bokeh.documents.indexOf(doc);\n",
       "    if (i > -1) {\n",
       "      window.Bokeh.documents.splice(i, 1);\n",
       "    }\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle kernel restart event\n",
       " */\n",
       "function handle_kernel_cleanup(event, handle) {\n",
       "  delete PyViz.comms[\"hv-extension-comm\"];\n",
       "  window.PyViz.plot_index = {}\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle update_display_data messages\n",
       " */\n",
       "function handle_update_output(event, handle) {\n",
       "  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n",
       "  handle_add_output(event, handle)\n",
       "}\n",
       "\n",
       "function register_renderer(events, OutputArea) {\n",
       "  function append_mime(data, metadata, element) {\n",
       "    // create a DOM node to render to\n",
       "    var toinsert = this.create_output_subarea(\n",
       "    metadata,\n",
       "    CLASS_NAME,\n",
       "    EXEC_MIME_TYPE\n",
       "    );\n",
       "    this.keyboard_manager.register_events(toinsert);\n",
       "    // Render to node\n",
       "    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n",
       "    render(props, toinsert[0]);\n",
       "    element.append(toinsert);\n",
       "    return toinsert\n",
       "  }\n",
       "\n",
       "  events.on('output_added.OutputArea', handle_add_output);\n",
       "  events.on('output_updated.OutputArea', handle_update_output);\n",
       "  events.on('clear_output.CodeCell', handle_clear_output);\n",
       "  events.on('delete.Cell', handle_clear_output);\n",
       "  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n",
       "\n",
       "  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n",
       "    safe: true,\n",
       "    index: 0\n",
       "  });\n",
       "}\n",
       "\n",
       "if (window.Jupyter !== undefined) {\n",
       "  try {\n",
       "    var events = require('base/js/events');\n",
       "    var OutputArea = require('notebook/js/outputarea').OutputArea;\n",
       "    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n",
       "      register_renderer(events, OutputArea);\n",
       "    }\n",
       "  } catch(err) {\n",
       "  }\n",
       "}\n"
      ],
      "application/vnd.holoviews_load.v0+json": "\nif ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n}\n\n\n    function JupyterCommManager() {\n    }\n\n    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        comm_manager.register_target(comm_id, function(comm) {\n          comm.on_msg(msg_handler);\n        });\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n          comm.onMsg = msg_handler;\n        });\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n          var messages = comm.messages[Symbol.asyncIterator]();\n          function processIteratorResult(result) {\n            var message = result.value;\n            console.log(message)\n            var content = {data: message.data, comm_id};\n            var buffers = []\n            for (var buffer of message.buffers || []) {\n              buffers.push(new DataView(buffer))\n            }\n            var metadata = message.metadata || {};\n            var msg = {content, buffers, metadata}\n            msg_handler(msg);\n            return messages.next().then(processIteratorResult);\n          }\n          return messages.next().then(processIteratorResult);\n        })\n      }\n    }\n\n    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n      if (comm_id in window.PyViz.comms) {\n        return window.PyViz.comms[comm_id];\n      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n        if (msg_handler) {\n          comm.on_msg(msg_handler);\n        }\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n        comm.open();\n        if (msg_handler) {\n          comm.onMsg = msg_handler;\n        }\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        var comm_promise = google.colab.kernel.comms.open(comm_id)\n        comm_promise.then((comm) => {\n          window.PyViz.comms[comm_id] = comm;\n          if (msg_handler) {\n            var messages = comm.messages[Symbol.asyncIterator]();\n            function processIteratorResult(result) {\n              var message = result.value;\n              var content = {data: message.data};\n              var metadata = message.metadata || {comm_id};\n              var msg = {content, metadata}\n              msg_handler(msg);\n              return messages.next().then(processIteratorResult);\n            }\n            return messages.next().then(processIteratorResult);\n          }\n        }) \n        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n          return comm_promise.then((comm) => {\n            comm.send(data, metadata, buffers, disposeOnDone);\n          });\n        };\n        var comm = {\n          send: sendClosure\n        };\n      }\n      window.PyViz.comms[comm_id] = comm;\n      return comm;\n    }\n    window.PyViz.comm_manager = new JupyterCommManager();\n    \n\n\nvar JS_MIME_TYPE = 'application/javascript';\nvar HTML_MIME_TYPE = 'text/html';\nvar EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\nvar CLASS_NAME = 'output';\n\n/**\n * Render data to the DOM node\n */\nfunction render(props, node) {\n  var div = document.createElement(\"div\");\n  var script = document.createElement(\"script\");\n  node.appendChild(div);\n  node.appendChild(script);\n}\n\n/**\n * Handle when a new output is added\n */\nfunction handle_add_output(event, handle) {\n  var output_area = handle.output_area;\n  var output = handle.output;\n  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n    return\n  }\n  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n  if (id !== undefined) {\n    var nchildren = toinsert.length;\n    var html_node = toinsert[nchildren-1].children[0];\n    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n    var scripts = [];\n    var nodelist = html_node.querySelectorAll(\"script\");\n    for (var i in nodelist) {\n      if (nodelist.hasOwnProperty(i)) {\n        scripts.push(nodelist[i])\n      }\n    }\n\n    scripts.forEach( function (oldScript) {\n      var newScript = document.createElement(\"script\");\n      var attrs = [];\n      var nodemap = oldScript.attributes;\n      for (var j in nodemap) {\n        if (nodemap.hasOwnProperty(j)) {\n          attrs.push(nodemap[j])\n        }\n      }\n      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n      oldScript.parentNode.replaceChild(newScript, oldScript);\n    });\n    if (JS_MIME_TYPE in output.data) {\n      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n    }\n    output_area._hv_plot_id = id;\n    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n      window.PyViz.plot_index[id] = Bokeh.index[id];\n    } else {\n      window.PyViz.plot_index[id] = null;\n    }\n  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n    var bk_div = document.createElement(\"div\");\n    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n    var script_attrs = bk_div.children[0].attributes;\n    for (var i = 0; i < script_attrs.length; i++) {\n      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n    }\n    // store reference to server id on output_area\n    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n  }\n}\n\n/**\n * Handle when an output is cleared or removed\n */\nfunction handle_clear_output(event, handle) {\n  var id = handle.cell.output_area._hv_plot_id;\n  var server_id = handle.cell.output_area._bokeh_server_id;\n  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n  if (server_id !== null) {\n    comm.send({event_type: 'server_delete', 'id': server_id});\n    return;\n  } else if (comm !== null) {\n    comm.send({event_type: 'delete', 'id': id});\n  }\n  delete PyViz.plot_index[id];\n  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n    var doc = window.Bokeh.index[id].model.document\n    doc.clear();\n    const i = window.Bokeh.documents.indexOf(doc);\n    if (i > -1) {\n      window.Bokeh.documents.splice(i, 1);\n    }\n  }\n}\n\n/**\n * Handle kernel restart event\n */\nfunction handle_kernel_cleanup(event, handle) {\n  delete PyViz.comms[\"hv-extension-comm\"];\n  window.PyViz.plot_index = {}\n}\n\n/**\n * Handle update_display_data messages\n */\nfunction handle_update_output(event, handle) {\n  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n  handle_add_output(event, handle)\n}\n\nfunction register_renderer(events, OutputArea) {\n  function append_mime(data, metadata, element) {\n    // create a DOM node to render to\n    var toinsert = this.create_output_subarea(\n    metadata,\n    CLASS_NAME,\n    EXEC_MIME_TYPE\n    );\n    this.keyboard_manager.register_events(toinsert);\n    // Render to node\n    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n    render(props, toinsert[0]);\n    element.append(toinsert);\n    return toinsert\n  }\n\n  events.on('output_added.OutputArea', handle_add_output);\n  events.on('output_updated.OutputArea', handle_update_output);\n  events.on('clear_output.CodeCell', handle_clear_output);\n  events.on('delete.Cell', handle_clear_output);\n  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n\n  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n    safe: true,\n    index: 0\n  });\n}\n\nif (window.Jupyter !== undefined) {\n  try {\n    var events = require('base/js/events');\n    var OutputArea = require('notebook/js/outputarea').OutputArea;\n    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n      register_renderer(events, OutputArea);\n    }\n  } catch(err) {\n  }\n}\n"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>*[data-root-id],\n",
       "*[data-root-id] > * {\n",
       "  box-sizing: border-box;\n",
       "  font-family: var(--jp-ui-font-family);\n",
       "  font-size: var(--jp-ui-font-size1);\n",
       "  color: var(--vscode-editor-foreground, var(--jp-ui-font-color1));\n",
       "}\n",
       "\n",
       "/* Override VSCode background color */\n",
       ".cell-output-ipywidget-background:has(\n",
       "    > .cell-output-ipywidget-background > .lm-Widget > *[data-root-id]\n",
       "  ),\n",
       ".cell-output-ipywidget-background:has(> .lm-Widget > *[data-root-id]) {\n",
       "  background-color: transparent !important;\n",
       "}\n",
       "</style>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='p1004'>\n",
       "  <div id=\"c0daa9df-e1d2-4940-9cfa-01089530aca8\" data-root-id=\"p1004\" style=\"display: contents;\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  var docs_json = {\"9b794fe5-c715-4083-a082-e88d428059f7\":{\"version\":\"3.4.3\",\"title\":\"Bokeh Application\",\"roots\":[{\"type\":\"object\",\"name\":\"panel.models.browser.BrowserInfo\",\"id\":\"p1004\"},{\"type\":\"object\",\"name\":\"panel.models.comm_manager.CommManager\",\"id\":\"p1005\",\"attributes\":{\"plot_id\":\"p1004\",\"comm_id\":\"e4f47d8cfbe748efa0229777c8c4f617\",\"client_comm_id\":\"540ee158f3f845499d0a3d01f47818ca\"}}],\"defs\":[{\"type\":\"model\",\"name\":\"ReactiveHTML1\"},{\"type\":\"model\",\"name\":\"FlexBox1\",\"properties\":[{\"name\":\"align_content\",\"kind\":\"Any\",\"default\":\"flex-start\"},{\"name\":\"align_items\",\"kind\":\"Any\",\"default\":\"flex-start\"},{\"name\":\"flex_direction\",\"kind\":\"Any\",\"default\":\"row\"},{\"name\":\"flex_wrap\",\"kind\":\"Any\",\"default\":\"wrap\"},{\"name\":\"gap\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"justify_content\",\"kind\":\"Any\",\"default\":\"flex-start\"}]},{\"type\":\"model\",\"name\":\"FloatPanel1\",\"properties\":[{\"name\":\"config\",\"kind\":\"Any\",\"default\":{\"type\":\"map\"}},{\"name\":\"contained\",\"kind\":\"Any\",\"default\":true},{\"name\":\"position\",\"kind\":\"Any\",\"default\":\"right-top\"},{\"name\":\"offsetx\",\"kind\":\"Any\",\"default\":null},{\"name\":\"offsety\",\"kind\":\"Any\",\"default\":null},{\"name\":\"theme\",\"kind\":\"Any\",\"default\":\"primary\"},{\"name\":\"status\",\"kind\":\"Any\",\"default\":\"normalized\"}]},{\"type\":\"model\",\"name\":\"GridStack1\",\"properties\":[{\"name\":\"mode\",\"kind\":\"Any\",\"default\":\"warn\"},{\"name\":\"ncols\",\"kind\":\"Any\",\"default\":null},{\"name\":\"nrows\",\"kind\":\"Any\",\"default\":null},{\"name\":\"allow_resize\",\"kind\":\"Any\",\"default\":true},{\"name\":\"allow_drag\",\"kind\":\"Any\",\"default\":true},{\"name\":\"state\",\"kind\":\"Any\",\"default\":[]}]},{\"type\":\"model\",\"name\":\"drag1\",\"properties\":[{\"name\":\"slider_width\",\"kind\":\"Any\",\"default\":5},{\"name\":\"slider_color\",\"kind\":\"Any\",\"default\":\"black\"},{\"name\":\"value\",\"kind\":\"Any\",\"default\":50}]},{\"type\":\"model\",\"name\":\"click1\",\"properties\":[{\"name\":\"terminal_output\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"debug_name\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"clears\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"FastWrapper1\",\"properties\":[{\"name\":\"object\",\"kind\":\"Any\",\"default\":null},{\"name\":\"style\",\"kind\":\"Any\",\"default\":null}]},{\"type\":\"model\",\"name\":\"NotificationAreaBase1\",\"properties\":[{\"name\":\"js_events\",\"kind\":\"Any\",\"default\":{\"type\":\"map\"}},{\"name\":\"position\",\"kind\":\"Any\",\"default\":\"bottom-right\"},{\"name\":\"_clear\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"NotificationArea1\",\"properties\":[{\"name\":\"js_events\",\"kind\":\"Any\",\"default\":{\"type\":\"map\"}},{\"name\":\"notifications\",\"kind\":\"Any\",\"default\":[]},{\"name\":\"position\",\"kind\":\"Any\",\"default\":\"bottom-right\"},{\"name\":\"_clear\",\"kind\":\"Any\",\"default\":0},{\"name\":\"types\",\"kind\":\"Any\",\"default\":[{\"type\":\"map\",\"entries\":[[\"type\",\"warning\"],[\"background\",\"#ffc107\"],[\"icon\",{\"type\":\"map\",\"entries\":[[\"className\",\"fas fa-exclamation-triangle\"],[\"tagName\",\"i\"],[\"color\",\"white\"]]}]]},{\"type\":\"map\",\"entries\":[[\"type\",\"info\"],[\"background\",\"#007bff\"],[\"icon\",{\"type\":\"map\",\"entries\":[[\"className\",\"fas fa-info-circle\"],[\"tagName\",\"i\"],[\"color\",\"white\"]]}]]}]}]},{\"type\":\"model\",\"name\":\"Notification\",\"properties\":[{\"name\":\"background\",\"kind\":\"Any\",\"default\":null},{\"name\":\"duration\",\"kind\":\"Any\",\"default\":3000},{\"name\":\"icon\",\"kind\":\"Any\",\"default\":null},{\"name\":\"message\",\"kind\":\"Any\",\"default\":\"\"},{\"name\":\"notification_type\",\"kind\":\"Any\",\"default\":null},{\"name\":\"_destroyed\",\"kind\":\"Any\",\"default\":false}]},{\"type\":\"model\",\"name\":\"TemplateActions1\",\"properties\":[{\"name\":\"open_modal\",\"kind\":\"Any\",\"default\":0},{\"name\":\"close_modal\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"BootstrapTemplateActions1\",\"properties\":[{\"name\":\"open_modal\",\"kind\":\"Any\",\"default\":0},{\"name\":\"close_modal\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"TemplateEditor1\",\"properties\":[{\"name\":\"layout\",\"kind\":\"Any\",\"default\":[]}]},{\"type\":\"model\",\"name\":\"MaterialTemplateActions1\",\"properties\":[{\"name\":\"open_modal\",\"kind\":\"Any\",\"default\":0},{\"name\":\"close_modal\",\"kind\":\"Any\",\"default\":0}]},{\"type\":\"model\",\"name\":\"copy_to_clipboard1\",\"properties\":[{\"name\":\"fill\",\"kind\":\"Any\",\"default\":\"none\"},{\"name\":\"value\",\"kind\":\"Any\",\"default\":null}]}]}};\n",
       "  var render_items = [{\"docid\":\"9b794fe5-c715-4083-a082-e88d428059f7\",\"roots\":{\"p1004\":\"c0daa9df-e1d2-4940-9cfa-01089530aca8\"},\"root_ids\":[\"p1004\"]}];\n",
       "  var docs = Object.values(docs_json)\n",
       "  if (!docs) {\n",
       "    return\n",
       "  }\n",
       "  const py_version = docs[0].version.replace('rc', '-rc.').replace('.dev', '-dev.')\n",
       "  async function embed_document(root) {\n",
       "    var Bokeh = get_bokeh(root)\n",
       "    await Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && id_el.children[0].hasAttribute('data-root-id')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t  for (const child of root_el.children) {\n",
       "            // Ensure JupyterLab does not capture keyboard shortcuts\n",
       "            // see: https://jupyterlab.readthedocs.io/en/4.1.x/extension/notebook.html#keyboard-interaction-model\n",
       "\t    child.setAttribute('data-lm-suppress-shortcuts', 'true')\n",
       "\t  }\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  function get_bokeh(root) {\n",
       "    if (root.Bokeh === undefined) {\n",
       "      return null\n",
       "    } else if (root.Bokeh.version !== py_version) {\n",
       "      if (root.Bokeh.versions === undefined || !root.Bokeh.versions.has(py_version)) {\n",
       "\treturn null\n",
       "      }\n",
       "      return root.Bokeh.versions.get(py_version);\n",
       "    } else if (root.Bokeh.version === py_version) {\n",
       "      return root.Bokeh\n",
       "    }\n",
       "    return null\n",
       "  }\n",
       "  function is_loaded(root) {\n",
       "    var Bokeh = get_bokeh(root)\n",
       "    return (Bokeh != null && Bokeh.Panel !== undefined)\n",
       "  }\n",
       "  if (is_loaded(root)) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (is_loaded(root)) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "\t  var Bokeh = get_bokeh(root)\n",
       "\t  if (Bokeh == null || Bokeh.Panel == null) {\n",
       "            console.warn(\"Panel: ERROR: Unable to run Panel code because Bokeh or Panel library is missing\");\n",
       "\t  } else {\n",
       "\t    console.warn(\"Panel: WARNING: Attempting to render but not all required libraries could be resolved.\")\n",
       "\t    embed_document(root)\n",
       "\t  }\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ]
     },
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "p1004"
      }
     },
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n",
      "estimated unique values: 5\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'holoviews.core.overlay.NdOverlay'>\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<img src='data:image/png;base64,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' style='max-width:100%; margin: auto; display: block; '/>"
      ],
      "text/plain": [
       ":NdOverlay   [Variable]\n",
       "   :Curve   [year]   (value)"
      ]
     },
     "execution_count": 42,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {}
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import hvplot.polars\n",
    "hvplot.extension('matplotlib')\n",
    "\n",
    "res = (\n",
    "    autos\n",
    "    .filter(pl.col('make').is_in(top3['make']))\n",
    "    .group_by(['year', 'make'])\n",
    "    .agg(\n",
    "        min_city08=pl.col('city08').min(),\n",
    "        max_city08=pl.col('city08').max(),\n",
    "    )\n",
    "    .sort(['year', 'make'])\n",
    "    .pivot(\n",
    "        index='year',\n",
    "        on='make',\n",
    "        values=['min_city08', 'max_city08']\n",
    "    )\n",
    "    .hvplot(\n",
    "        x='year', # pseudo index\n",
    "        color=['#55c667', '#2f8738', '#2c4279', '#141f38', '#d4e116', '#afba12']\n",
    "    )\n",
    "    \n",
    ")\n",
    "\n",
    "print(type(res))\n",
    "res"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21fbc873-d55b-4da8-8909-183e36cd50f3",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.9 Aggregating over the Original Rows\n",
    "[back to Top](#Top)\n",
    "\n",
    "* In Pandas, `.transform()` aggregates data and returns the aggregated values in terms of the original rows.\n",
    "* There is no `.transform()` in Polars, but you can achieve the same result using the window function `.over()`\n",
    "\n",
    "In this example:\n",
    "\n",
    "* Calculate the *make* average *city08* value and return it to the original rows.\n",
    "* Use `pl.col` to refer to the column you want to aggregate.\n",
    "* Create the aggregate function, here `.mean()`.\n",
    "* Follow with `.over`\n",
    "* Pass in the column you want to group by (in this case, *make*)\n",
    "\n",
    "**Translated to Pandas, we want to groupby on make and aggregate the value city08 ????**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "3786e2cb-e14f-40e2-9ca2-c0421570203a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>city08</th></tr><tr><td>cat</td><td>u8</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>19</td></tr><tr><td>&quot;Ferrari&quot;</td><td>9</td></tr><tr><td>&quot;Dodge&quot;</td><td>23</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 2)\n",
       "+------------+--------+\n",
       "| make       | city08 |\n",
       "| ---        | ---    |\n",
       "| cat        | u8     |\n",
       "+=====================+\n",
       "| Alfa Romeo | 19     |\n",
       "| Ferrari    | 9      |\n",
       "| Dodge      | 23     |\n",
       "+------------+--------+"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check make and city08\n",
    "autos[['make', 'city08']].head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "0d2e80d7-759c-4fc6-a845-271f1fca16b9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>mean_city</th></tr><tr><td>f64</td></tr></thead><tbody><tr><td>19.929235</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 1)\n",
       "+-----------+\n",
       "| mean_city |\n",
       "| ---       |\n",
       "| f64       |\n",
       "+===========+\n",
       "| 19.929235 |\n",
       "+-----------+"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# testing with .select and single column only\n",
    "# This creates only a single statistic\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        pl.col('city08')\n",
    "        .mean()\n",
    "        .alias('mean_city')\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "e055d1d5-8644-4404-a945-bee84b6ad2d4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city08</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td></tr><tr><td>&quot;Ferrari&quot;</td><td>11.842294</td></tr><tr><td>&quot;Dodge&quot;</td><td>15.461967</td></tr><tr><td>&quot;Dodge&quot;</td><td>15.461967</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 2)\n",
       "+------------+-------------+\n",
       "| make       | mean_city08 |\n",
       "| ---        | ---         |\n",
       "| cat        | f64         |\n",
       "+==========================+\n",
       "| Alfa Romeo | 19.402062   |\n",
       "| Ferrari    | 11.842294   |\n",
       "| Dodge      | 15.461967   |\n",
       "| Dodge      | 15.461967   |\n",
       "| Subaru     | 21.130306   |\n",
       "| …          | …           |\n",
       "| Subaru     | 21.130306   |\n",
       "| Subaru     | 21.130306   |\n",
       "| Subaru     | 21.130306   |\n",
       "| Subaru     | 21.130306   |\n",
       "| Subaru     | 21.130306   |\n",
       "+------------+-------------+"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# testing with .select and single column only\n",
    "# This creates an aggregated data statistic\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        'make',\n",
    "        pl.col('city08')\n",
    "        .mean()\n",
    "        .over('make')\n",
    "        .alias('mean_city08'),\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "e4e9e132-2738-4865-9520-b339be442b28",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 16)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th><th>mean_city08</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td><td>f64</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>19.402062</td></tr><tr><td>1985</td><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>4.9</td><td>12</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>9</td><td>14</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>11.842294</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;Charger&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Subcompact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.018888</td><td>23</td><td>33</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>15.461967</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;B150/B250 Wagon 2WD&quot;</td><td>5.2</td><td>8</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Vans&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>10</td><td>12</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td><td>15.461967</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>15.658422</td><td>17</td><td>23</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>21.130306</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>13.523182</td><td>19</td><td>26</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td><td>21.130306</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>12.935218</td><td>20</td><td>28</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>21.130306</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td><td>21.130306</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>21.130306</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>16.528334</td><td>16</td><td>21</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td><td>21.130306</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 16)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | hig | cre | is_ | num | mean_c |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | hwa | ate | aut | _ge | ity08  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | y08 | dOn | oma | ars | ---    |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | --- | --- | tic | --- | f64    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     | u8  | dat | --- | u8  |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | eti | str |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | me[ |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | μs] |     |     |        |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25  | 201 | fal | 5   | 19.402 |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |     | 3-0 | se  |     | 062    |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |     | 1-0 |     |     |        |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |     | 1   |     |     |        |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 198 | Fer | Tes | 4.9 | 12  | Man | Rea | Two | Reg | 27. | 9   | 14  | 201 | fal | 5   | 11.842 |\n",
       "| 5   | rar | tar |     |     | ual | r-W | Sea | ula | 046 |     |     | 3-0 | se  |     | 294    |\n",
       "|     | i   | oss |     |     | 5-s | hee | ter | r   | 364 |     |     | 1-0 |     |     |        |\n",
       "|     |     | a   |     |     | pd  | l   | s   |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 198 | Dod | Cha | 2.2 | 4   | Man | Fro | Sub | Reg | 11. | 23  | 33  | 201 | fal | 5   | 15.461 |\n",
       "| 5   | ge  | rge |     |     | ual | nt- | com | ula | 018 |     |     | 3-0 | se  |     | 967    |\n",
       "|     |     | r   |     |     | 5-s | Whe | pac | r   | 888 |     |     | 1-0 |     |     |        |\n",
       "|     |     |     |     |     | pd  | el  | t   |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     |     | Dri | Car |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | ve  | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 198 | Dod | B15 | 5.2 | 8   | Aut | Rea | Van | Reg | 27. | 10  | 12  | 201 | tru | 3   | 15.461 |\n",
       "| 5   | ge  | 0/B |     |     | oma | r-W | s   | ula | 046 |     |     | 3-0 | e   |     | 967    |\n",
       "|     |     | 250 |     |     | tic | hee |     | r   | 364 |     |     | 1-0 |     |     |        |\n",
       "|     |     | Wag |     |     | 3-s | l   |     |     |     |     |     | 1   |     |     |        |\n",
       "|     |     | on  |     |     | pd  | Dri |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     | 2WD |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | 4-W | Com | Pre | 15. | 17  | 23  | 201 | fal | 5   | 21.130 |\n",
       "| 3   | aru | acy |     |     | ual | hee | pac | miu | 658 |     |     | 3-0 | se  |     | 306    |\n",
       "|     |     | AWD |     |     | 5-s | l   | t   | m   | 422 |     |     | 1-0 |     |     |        |\n",
       "|     |     | Tur |     |     | pd  | or  | Car |     |     |     |     | 1   |     |     |        |\n",
       "|     |     | bo  |     |     |     | All | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |        |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |        |\n",
       "| …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …      |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | Fro | Com | Reg | 13. | 19  | 26  | 201 | tru | 4   | 21.130 |\n",
       "| 3   | aru | acy |     |     | oma | nt- | pac | ula | 523 |     |     | 3-0 | e   |     | 306    |\n",
       "|     |     |     |     |     | tic | Whe | t   | r   | 182 |     |     | 1-0 |     |     |        |\n",
       "|     |     |     |     |     | 4-s | el  | Car |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | Fro | Com | Reg | 12. | 20  | 28  | 201 | fal | 5   | 21.130 |\n",
       "| 3   | aru | acy |     |     | ual | nt- | pac | ula | 935 |     |     | 3-0 | se  |     | 306    |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 218 |     |     | 1-0 |     |     |        |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |        |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | 4-W | Com | Reg | 14. | 18  | 24  | 201 | tru | 4   | 21.130 |\n",
       "| 3   | aru | acy |     |     | oma | hee | pac | ula | 167 |     |     | 3-0 | e   |     | 306    |\n",
       "|     |     | AWD |     |     | tic | l   | t   | r   | 143 |     |     | 1-0 |     |     |        |\n",
       "|     |     |     |     |     | 4-s | or  | Car |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     | pd  | All | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |        |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |        |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | 4-W | Com | Reg | 14. | 18  | 24  | 201 | fal | 5   | 21.130 |\n",
       "| 3   | aru | acy |     |     | ual | hee | pac | ula | 167 |     |     | 3-0 | se  |     | 306    |\n",
       "|     |     | AWD |     |     | 5-s | l   | t   | r   | 143 |     |     | 1-0 |     |     |        |\n",
       "|     |     |     |     |     | pd  | or  | Car |     |     |     |     | 1   |     |     |        |\n",
       "|     |     |     |     |     |     | All | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |        |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |        |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | 4-W | Com | Pre | 16. | 16  | 21  | 201 | tru | 4   | 21.130 |\n",
       "| 3   | aru | acy |     |     | oma | hee | pac | miu | 528 |     |     | 3-0 | e   |     | 306    |\n",
       "|     |     | AWD |     |     | tic | l   | t   | m   | 334 |     |     | 1-0 |     |     |        |\n",
       "|     |     | Tur |     |     | 4-s | or  | Car |     |     |     |     | 1   |     |     |        |\n",
       "|     |     | bo  |     |     | pd  | All | s   |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |        |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |        |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |        |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |        |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# .with_columns version to embed in full dataset\n",
    "(\n",
    "    autos\n",
    "    .with_columns(\n",
    "        pl.col('city08')\n",
    "        .mean()\n",
    "        .over('make')\n",
    "        .alias('mean_city08')\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a94da2c-809a-43e6-bd1f-c0e71d05a7ae",
   "metadata": {},
   "source": [
    "--- \n",
    "We can validate the results for specifically Alfa Romeo, Ferrari and Dodge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "9e164650-1d72-4612-aa68-3b83aea48428",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>count_city08</th></tr><tr><td>cat</td><td>u32</td></tr></thead><tbody><tr><td>&quot;Azure Dynamics&quot;</td><td>1</td></tr><tr><td>&quot;Mahindra&quot;</td><td>1</td></tr><tr><td>&quot;Excalibur Autos&quot;</td><td>1</td></tr><tr><td>&quot;London Taxi&quot;</td><td>1</td></tr><tr><td>&quot;Quantum Technologies&quot;</td><td>2</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Toyota&quot;</td><td>2470</td></tr><tr><td>&quot;Subaru&quot;</td><td>1013</td></tr><tr><td>&quot;Dodge&quot;</td><td>2695</td></tr><tr><td>&quot;Ferrari&quot;</td><td>279</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>97</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+----------------------+--------------+\n",
       "| make                 | count_city08 |\n",
       "| ---                  | ---          |\n",
       "| cat                  | u32          |\n",
       "+=====================================+\n",
       "| Azure Dynamics       | 1            |\n",
       "| Mahindra             | 1            |\n",
       "| Excalibur Autos      | 1            |\n",
       "| London Taxi          | 1            |\n",
       "| Quantum Technologies | 2            |\n",
       "| …                    | …            |\n",
       "| Toyota               | 2470         |\n",
       "| Subaru               | 1013         |\n",
       "| Dodge                | 2695         |\n",
       "| Ferrari              | 279          |\n",
       "| Alfa Romeo           | 97           |\n",
       "+----------------------+--------------+"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Quick test, equivalent to Pandas value_counts, without filter\n",
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .agg(count_city08=pl.col('city08').len()) # equivalent to .len()\n",
    "    .sort(by='make', descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "bfe42e8f-7353-47c3-8d7c-b36fc3d41b5d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city08</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Azure Dynamics&quot;</td><td>62.0</td></tr><tr><td>&quot;Mahindra&quot;</td><td>19.0</td></tr><tr><td>&quot;Excalibur Autos&quot;</td><td>12.0</td></tr><tr><td>&quot;London Taxi&quot;</td><td>21.0</td></tr><tr><td>&quot;Quantum Technologies&quot;</td><td>19.0</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Toyota&quot;</td><td>22.4</td></tr><tr><td>&quot;Subaru&quot;</td><td>21.130306</td></tr><tr><td>&quot;Dodge&quot;</td><td>15.461967</td></tr><tr><td>&quot;Ferrari&quot;</td><td>11.842294</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+----------------------+-------------+\n",
       "| make                 | mean_city08 |\n",
       "| ---                  | ---         |\n",
       "| cat                  | f64         |\n",
       "+====================================+\n",
       "| Azure Dynamics       | 62.0        |\n",
       "| Mahindra             | 19.0        |\n",
       "| Excalibur Autos      | 12.0        |\n",
       "| London Taxi          | 21.0        |\n",
       "| Quantum Technologies | 19.0        |\n",
       "| …                    | …           |\n",
       "| Toyota               | 22.4        |\n",
       "| Subaru               | 21.130306   |\n",
       "| Dodge                | 15.461967   |\n",
       "| Ferrari              | 11.842294   |\n",
       "| Alfa Romeo           | 19.402062   |\n",
       "+----------------------+-------------+"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Without filter, showing results for all makes\n",
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .agg(mean_city08=pl.col('city08').mean())\n",
    "    .sort(by='make', descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "6343ce5e-e11f-416c-9987-95ac604c1e3d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n",
      "dataframe filtered\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>mean_city08</th></tr><tr><td>cat</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Dodge&quot;</td><td>15.461967</td></tr><tr><td>&quot;Ferrari&quot;</td><td>11.842294</td></tr><tr><td>&quot;Alfa Romeo&quot;</td><td>19.402062</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 2)\n",
       "+------------+-------------+\n",
       "| make       | mean_city08 |\n",
       "| ---        | ---         |\n",
       "| cat        | f64         |\n",
       "+==========================+\n",
       "| Dodge      | 15.461967   |\n",
       "| Ferrari    | 11.842294   |\n",
       "| Alfa Romeo | 19.402062   |\n",
       "+------------+-------------+"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# With filter\n",
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .agg(mean_city08=pl.col('city08').mean())\n",
    "    .filter(\n",
    "        pl.col('make')\n",
    "        .cast(pl.String)\n",
    "        .is_in(['Alfa Romeo', 'Ferrari', 'Dodge'])\n",
    "    )\n",
    "    .sort(by='make', descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3fe905fc-9033-4297-9fe8-3befcce6b72f",
   "metadata": {},
   "source": [
    "* A quick implication is we can use window functions to do multiple groupings and aggregations in a single operation.\n",
    "* Explore filtering with window functions in Chapter 5."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "266b0451-2f7e-4fb2-b9cc-511e69b6c286",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.10 Grouping to Lists\n",
    "[back to Top](#Top)\n",
    "\n",
    "You can write powerful code with expressions, window functions and lists.\n",
    "\n",
    "* First, aggregate to a column by passing pl.all to the `.agg` context."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "56a05543-0f95-416b-81dc-a9bea3c430bd",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>year</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>cat</td><td>list[i16]</td><td>list[cat]</td><td>list[f32]</td><td>list[u8]</td><td>list[str]</td><td>list[cat]</td><td>list[cat]</td><td>list[cat]</td><td>list[f32]</td><td>list[u8]</td><td>list[u8]</td><td>list[datetime[μs]]</td><td>list[str]</td><td>list[u8]</td></tr></thead><tbody><tr><td>&quot;Excalibur Autos&quot;</td><td>[1987]</td><td>[&quot;Phaeton&quot;]</td><td>[5.0]</td><td>[8]</td><td>[&quot;Automatic 4-spd&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;]</td><td>[&quot;Minicompact Cars&quot;]</td><td>[&quot;Premium&quot;]</td><td>[21.250713]</td><td>[12]</td><td>[17]</td><td>[2013-01-01 00:00:00]</td><td>[&quot;true&quot;]</td><td>[4]</td></tr><tr><td>&quot;Renault&quot;</td><td>[1986, 1986, … 1985]</td><td>[&quot;Alliance Convertible&quot;, &quot;Alliance Convertible&quot;, … &quot;18i 4DR Wagon&quot;]</td><td>[1.7, 1.7, … 2.2]</td><td>[4, 4, … 4]</td><td>[&quot;Automatic 3-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Manual 5-spd&quot;]</td><td>[&quot;Front-Wheel Drive&quot;, &quot;Front-Wheel Drive&quot;, … &quot;Front-Wheel Drive&quot;]</td><td>[&quot;Subcompact Cars&quot;, &quot;Subcompact Cars&quot;, … &quot;Small Station Wagons&quot;]</td><td>[&quot;Regular&quot;, &quot;Regular&quot;, … &quot;Regular&quot;]</td><td>[14.8755, 11.442693, … 12.935218]</td><td>[19, 23, … 20]</td><td>[22, 29, … 28]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;true&quot;, &quot;false&quot;, … &quot;false&quot;]</td><td>[3, 5, … 5]</td></tr><tr><td>&quot;Porsche&quot;</td><td>[1994, 1994, … 1993]</td><td>[&quot;911 Carrera 4/2&quot;, &quot;911 Carrera 4/2&quot;, … &quot;968&quot;]</td><td>[3.6, 3.6, … 3.0]</td><td>[6, 6, … 4]</td><td>[&quot;Automatic 4-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Manual 6-spd&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;, &quot;Rear-Wheel Drive&quot;, … &quot;Rear-Wheel Drive&quot;]</td><td>[&quot;Two Seaters&quot;, &quot;Two Seaters&quot;, … &quot;Minicompact Cars&quot;]</td><td>[&quot;Regular&quot;, &quot;Regular&quot;, … &quot;Premium&quot;]</td><td>[17.500587, 16.528334, … 16.528334]</td><td>[15, 15, … 15]</td><td>[22, 23, … 24]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;true&quot;, &quot;false&quot;, … &quot;false&quot;]</td><td>[4, 5, … 6]</td></tr><tr><td>&quot;Land Rover&quot;</td><td>[1993, 1993, … 1992]</td><td>[&quot;Defender 110&quot;, &quot;Range Rover County&quot;, … &quot;Range Rover&quot;]</td><td>[3.9, 3.9, … 3.9]</td><td>[8, 8, … 8]</td><td>[&quot;Manual 5-spd&quot;, &quot;Automatic 4-spd&quot;, … &quot;Automatic 4-spd&quot;]</td><td>[&quot;4-Wheel or All-Wheel Drive&quot;, &quot;4-Wheel or All-Wheel Drive&quot;, … &quot;4-Wheel or All-Wheel Drive&quot;]</td><td>[&quot;Special Purpose Vehicles&quot;, &quot;Special Purpose Vehicles&quot;, … &quot;Special Purpose Vehicles&quot;]</td><td>[&quot;Premium&quot;, &quot;Premium&quot;, … &quot;Premium&quot;]</td><td>[29.750999, 24.7925, … 22.885386]</td><td>[9, 11, … 11]</td><td>[11, 14, … 15]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;false&quot;, &quot;true&quot;, … &quot;true&quot;]</td><td>[5, 4, … 4]</td></tr><tr><td>&quot;Geo&quot;</td><td>[1993, 1993, … 1993]</td><td>[&quot;Tracker Convertible 2WD&quot;, &quot;Tracker Convertible 2WD&quot;, … &quot;Prizm&quot;]</td><td>[1.6, 1.6, … 1.8]</td><td>[4, 4, … 4]</td><td>[&quot;Automatic 3-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Manual 5-spd&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;, &quot;Rear-Wheel Drive&quot;, … &quot;Front-Wheel Drive&quot;]</td><td>[&quot;Special Purpose Vehicles&quot;, &quot;Special Purpose Vehicles&quot;, … &quot;Compact Cars&quot;]</td><td>[&quot;Regular&quot;, &quot;Regular&quot;, … &quot;Regular&quot;]</td><td>[14.167143, 12.935218, … 11.018888]</td><td>[20, 22, … 24]</td><td>[22, 24, … 31]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;true&quot;, &quot;false&quot;, … &quot;false&quot;]</td><td>[3, 5, … 5]</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Roush Performance&quot;</td><td>[2001, 2001, … 2021]</td><td>[&quot;Stage 3 Mustang&quot;, &quot;Stage 3 Mustang&quot;, … &quot;Mustang&quot;]</td><td>[4.6, 4.6, … 5.0]</td><td>[8, 8, … 8]</td><td>[&quot;Automatic 4-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Automatic (S10)&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;, &quot;Rear-Wheel Drive&quot;, … &quot;Rear-Wheel Drive&quot;]</td><td>[&quot;Subcompact Cars&quot;, &quot;Subcompact Cars&quot;, … &quot;Subcompact Cars&quot;]</td><td>[&quot;Premium&quot;, &quot;Premium&quot;, … &quot;Premium&quot;]</td><td>[16.528334, 16.528334, … 19.834]</td><td>[16, 15, … 12]</td><td>[21, 23, … 19]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2021-07-06 00:00:00]</td><td>[&quot;true&quot;, &quot;false&quot;, … &quot;true&quot;]</td><td>[4, 5, … 10]</td></tr><tr><td>&quot;STI&quot;</td><td>[2019]</td><td>[&quot;S209&quot;]</td><td>[2.5]</td><td>[4]</td><td>[&quot;Manual 6-spd&quot;]</td><td>[&quot;All-Wheel Drive&quot;]</td><td>[&quot;Compact Cars&quot;]</td><td>[&quot;Premium&quot;]</td><td>[17.500587]</td><td>[15]</td><td>[21]</td><td>[2019-10-02 00:00:00]</td><td>[&quot;false&quot;]</td><td>[6]</td></tr><tr><td>&quot;Merkur&quot;</td><td>[1986, 1986, … 1989]</td><td>[&quot;XR4Ti&quot;, &quot;XR4Ti&quot;, … &quot;Scorpio&quot;]</td><td>[2.3, 2.3, … 2.9]</td><td>[4, 4, … 6]</td><td>[&quot;Automatic 3-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Manual 5-spd&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;, &quot;Rear-Wheel Drive&quot;, … &quot;Rear-Wheel Drive&quot;]</td><td>[&quot;Compact Cars&quot;, &quot;Compact Cars&quot;, … &quot;Midsize Cars&quot;]</td><td>[&quot;Regular&quot;, &quot;Regular&quot;, … &quot;Regular&quot;]</td><td>[16.528334, 15.658422, … 16.528334]</td><td>[16, 17, … 16]</td><td>[20, 22, … 21]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;true&quot;, &quot;false&quot;, … &quot;false&quot;]</td><td>[3, 5, … 5]</td></tr><tr><td>&quot;Maserati&quot;</td><td>[1985, 1985, … 1991]</td><td>[&quot;Biturbo&quot;, &quot;Biturbo&quot;, … &quot;430&quot;]</td><td>[2.5, 2.5, … 2.8]</td><td>[6, 6, … 6]</td><td>[&quot;Manual 5-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Automatic 4-spd&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;, &quot;Rear-Wheel Drive&quot;, … &quot;Rear-Wheel Drive&quot;]</td><td>[&quot;Minicompact Cars&quot;, &quot;Subcompact Cars&quot;, … &quot;Subcompact Cars&quot;]</td><td>[&quot;Regular&quot;, &quot;Premium&quot;, … &quot;Premium&quot;]</td><td>[21.250713, 21.250713, … 21.250713]</td><td>[12, 12, … 13]</td><td>[18, 18, … 16]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;false&quot;, &quot;false&quot;, … &quot;true&quot;]</td><td>[5, 5, … 4]</td></tr><tr><td>&quot;AM General&quot;</td><td>[1985, 1985, … 1984]</td><td>[&quot;Post Office DJ5 2WD&quot;, &quot;Post Office DJ8 2WD&quot;, … &quot;DJ Po Vehicle 2WD&quot;]</td><td>[2.5, 4.2, … 2.5]</td><td>[4, 6, … 4]</td><td>[&quot;Automatic 3-spd&quot;, &quot;Automatic 3-spd&quot;, … &quot;Automatic 3-spd&quot;]</td><td>[&quot;Rear-Wheel Drive&quot;, &quot;Rear-Wheel Drive&quot;, … &quot;2-Wheel Drive&quot;]</td><td>[&quot;Special Purpose Vehicle 2WD&quot;, &quot;Special Purpose Vehicle 2WD&quot;, … &quot;Special Purpose Vehicle 2WD&quot;]</td><td>[&quot;Regular&quot;, &quot;Regular&quot;, … &quot;Regular&quot;]</td><td>[18.594376, 22.885386, … 17.500587]</td><td>[16, 13, … 18]</td><td>[17, 13, … 17]</td><td>[2013-01-01 00:00:00, 2013-01-01 00:00:00, … 2013-01-01 00:00:00]</td><td>[&quot;true&quot;, &quot;true&quot;, … &quot;true&quot;]</td><td>[3, 3, … 3]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| mak | yea | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| e   | r   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| cat | lis | lis | lis | --- | lis | lis | lis | --- | --- | lis | list[u | list[ | ---   | list[ |\n",
       "|     | t[i | t[c | t[f | lis | t[s | t[c | t[c | lis | lis | t[u | 8]     | datet | list[ | u8]   |\n",
       "|     | 16] | at] | 32] | t[u | tr] | at] | at] | t[c | t[f | 8]  |        | ime[μ | str]  |       |\n",
       "|     |     |     |     | 8]  |     |     |     | at] | 32] |     |        | s]]   |       |       |\n",
       "+==================================================================================================+\n",
       "| Exc | [19 | [\"P | [5. | [8] | [\"A | [\"R | [\"M | [\"P | [21 | [12 | [17]   | [2013 | [\"tru | [4]   |\n",
       "| ali | 87] | hae | 0]  |     | uto | ear | ini | rem | .25 | ]   |        | -01-0 | e\"]   |       |\n",
       "| bur |     | ton |     |     | mat | -Wh | com | ium | 071 |     |        | 1 00: |       |       |\n",
       "| Aut |     | \"]  |     |     | ic  | eel | pac | \"]  | 3]  |     |        | 00:00 |       |       |\n",
       "| os  |     |     |     |     | 4-s | Dri | t   |     |     |     |        | ]     |       |       |\n",
       "|     |     |     |     |     | pd\" | ve\" | Car |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     | ]   | ]   | s\"] |     |     |     |        |       |       |       |\n",
       "| Ren | [19 | [\"A | [1. | [4, | [\"A | [\"F | [\"S | [\"R | [14 | [19 | [22,   | [2013 | [\"tru | [3,   |\n",
       "| aul | 86, | lli | 7,  | 4,  | uto | ron | ubc | egu | .87 | ,   | 29, …  | -01-0 | e\",   | 5, …  |\n",
       "| t   | 198 | anc | 1.7 | …   | mat | t-W | omp | lar | 55, | 23, | 28]    | 1 00: | \"fals | 5]    |\n",
       "|     | 6,  | e   | , … | 4]  | ic  | hee | act | \",  | 11. | …   |        | 00:00 | e\", … |       |\n",
       "|     | …   | Con | 2.2 |     | 3-s | l   | Car | \"Re | 442 | 20] |        | , 201 | \"fals |       |\n",
       "|     | 198 | ver | ]   |     | pd\" | Dri | s\", | gul | 693 |     |        | 3-01- | e\"]   |       |\n",
       "|     | 5]  | tib |     |     | ,   | ve\" | \"Su | ar\" | , … |     |        | …     |       |       |\n",
       "|     |     | le\" |     |     | \"Ma | ,   | bco | , … | 12. |     |        |       |       |       |\n",
       "|     |     | ,   |     |     | nua | \"Fr | mpa | \"Re | 935 |     |        |       |       |       |\n",
       "|     |     | \"Al |     |     | l   | ont | c…  | gu… | 2…  |     |        |       |       |       |\n",
       "|     |     | li… |     |     | 5-… | -W… |     |     |     |     |        |       |       |       |\n",
       "| Por | [19 | [\"9 | [3. | [6, | [\"A | [\"R | [\"T | [\"R | [17 | [15 | [22,   | [2013 | [\"tru | [4,   |\n",
       "| sch | 94, | 11  | 6,  | 6,  | uto | ear | wo  | egu | .50 | ,   | 23, …  | -01-0 | e\",   | 5, …  |\n",
       "| e   | 199 | Car | 3.6 | …   | mat | -Wh | Sea | lar | 058 | 15, | 24]    | 1 00: | \"fals | 6]    |\n",
       "|     | 4,  | rer | , … | 4]  | ic  | eel | ter | \",  | 7,  | …   |        | 00:00 | e\", … |       |\n",
       "|     | …   | a   | 3.0 |     | 4-s | Dri | s\", | \"Re | 16. | 15] |        | , 201 | \"fals |       |\n",
       "|     | 199 | 4/2 | ]   |     | pd\" | ve\" | \"Tw | gul | 528 |     |        | 3-01- | e\"]   |       |\n",
       "|     | 3]  | \",  |     |     | ,   | ,   | o   | ar\" | 334 |     |        | …     |       |       |\n",
       "|     |     | \"91 |     |     | \"Ma | \"Re | Sea | , … | , … |     |        |       |       |       |\n",
       "|     |     | 1   |     |     | nua | ar- | ter | \"Pr | 16. |     |        |       |       |       |\n",
       "|     |     | Car |     |     | l   | Whe | s\", | em… | 52… |     |        |       |       |       |\n",
       "|     |     | re… |     |     | 5-… | …   | …   |     |     |     |        |       |       |       |\n",
       "| Lan | [19 | [\"D | [3. | [8, | [\"M | [\"4 | [\"S | [\"P | [29 | [9, | [11,   | [2013 | [\"fal | [5,   |\n",
       "| d   | 93, | efe | 9,  | 8,  | anu | -Wh | pec | rem | .75 | 11, | 14, …  | -01-0 | se\",  | 4, …  |\n",
       "| Rov | 199 | nde | 3.9 | …   | al  | eel | ial | ium | 099 | …   | 15]    | 1 00: | \"true | 4]    |\n",
       "| er  | 3,  | r   | , … | 8]  | 5-s | or  | Pur | \",  | 9,  | 11] |        | 00:00 | \", …  |       |\n",
       "|     | …   | 110 | 3.9 |     | pd\" | All | pos | \"Pr | 24. |     |        | , 201 | \"true |       |\n",
       "|     | 199 | \",  | ]   |     | ,   | -Wh | e   | emi | 792 |     |        | 3-01- | \"]    |       |\n",
       "|     | 2]  | \"Ra |     |     | \"Au | eel | Veh | um\" | 5,  |     |        | …     |       |       |\n",
       "|     |     | nge |     |     | tom | Dri | icl | , … | …   |     |        |       |       |       |\n",
       "|     |     | Rov |     |     | ati | ve\" | es\" | \"Pr | 22. |     |        |       |       |       |\n",
       "|     |     | er  |     |     | c   | ,…  | ,   | em… | 885 |     |        |       |       |       |\n",
       "|     |     | …   |     |     | 4-… |     | \"…  |     | 3…  |     |        |       |       |       |\n",
       "| Geo | [19 | [\"T | [1. | [4, | [\"A | [\"R | [\"S | [\"R | [14 | [20 | [22,   | [2013 | [\"tru | [3,   |\n",
       "|     | 93, | rac | 6,  | 4,  | uto | ear | pec | egu | .16 | ,   | 24, …  | -01-0 | e\",   | 5, …  |\n",
       "|     | 199 | ker | 1.6 | …   | mat | -Wh | ial | lar | 714 | 22, | 31]    | 1 00: | \"fals | 5]    |\n",
       "|     | 3,  | Con | , … | 4]  | ic  | eel | Pur | \",  | 3,  | …   |        | 00:00 | e\", … |       |\n",
       "|     | …   | ver | 1.8 |     | 3-s | Dri | pos | \"Re | 12. | 24] |        | , 201 | \"fals |       |\n",
       "|     | 199 | tib | ]   |     | pd\" | ve\" | e   | gul | 935 |     |        | 3-01- | e\"]   |       |\n",
       "|     | 3]  | le  |     |     | ,   | ,   | Veh | ar\" | 218 |     |        | …     |       |       |\n",
       "|     |     | 2WD |     |     | \"Ma | \"Re | icl | , … | , … |     |        |       |       |       |\n",
       "|     |     | \",  |     |     | nua | ar- | es\" | \"Re | 11. |     |        |       |       |       |\n",
       "|     |     | \"T… |     |     | l   | Whe | ,   | gu… | 01… |     |        |       |       |       |\n",
       "|     |     |     |     |     | 5-… | …   | \"…  |     |     |     |        |       |       |       |\n",
       "| …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …      | …     | …     | …     |\n",
       "| Rou | [20 | [\"S | [4. | [8, | [\"A | [\"R | [\"S | [\"P | [16 | [16 | [21,   | [2013 | [\"tru | [4,   |\n",
       "| sh  | 01, | tag | 6,  | 8,  | uto | ear | ubc | rem | .52 | ,   | 23, …  | -01-0 | e\",   | 5, …  |\n",
       "| Per | 200 | e 3 | 4.6 | …   | mat | -Wh | omp | ium | 833 | 15, | 19]    | 1 00: | \"fals | 10]   |\n",
       "| for | 1,  | Mus | , … | 8]  | ic  | eel | act | \",  | 4,  | …   |        | 00:00 | e\", … |       |\n",
       "| man | …   | tan | 5.0 |     | 4-s | Dri | Car | \"Pr | 16. | 12] |        | , 201 | \"true |       |\n",
       "| ce  | 202 | g\", | ]   |     | pd\" | ve\" | s\", | emi | 528 |     |        | 3-01- | \"]    |       |\n",
       "|     | 1]  | \"St |     |     | ,   | ,   | \"Su | um\" | 334 |     |        | …     |       |       |\n",
       "|     |     | age |     |     | \"Ma | \"Re | bco | , … | , … |     |        |       |       |       |\n",
       "|     |     | 3   |     |     | nua | ar- | mpa | \"Pr | 19. |     |        |       |       |       |\n",
       "|     |     | M…  |     |     | l   | Whe | c…  | em… | 83… |     |        |       |       |       |\n",
       "|     |     |     |     |     | 5-… | …   |     |     |     |     |        |       |       |       |\n",
       "| STI | [20 | [\"S | [2. | [4] | [\"M | [\"A | [\"C | [\"P | [17 | [15 | [21]   | [2019 | [\"fal | [6]   |\n",
       "|     | 19] | 209 | 5]  |     | anu | ll- | omp | rem | .50 | ]   |        | -10-0 | se\"]  |       |\n",
       "|     |     | \"]  |     |     | al  | Whe | act | ium | 058 |     |        | 2 00: |       |       |\n",
       "|     |     |     |     |     | 6-s | el  | Car | \"]  | 7]  |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd\" | Dri | s\"] |     |     |     |        | ]     |       |       |\n",
       "|     |     |     |     |     | ]   | ve\" |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ]   |     |     |     |     |        |       |       |       |\n",
       "| Mer | [19 | [\"X | [2. | [4, | [\"A | [\"R | [\"C | [\"R | [16 | [16 | [20,   | [2013 | [\"tru | [3,   |\n",
       "| kur | 86, | R4T | 3,  | 4,  | uto | ear | omp | egu | .52 | ,   | 22, …  | -01-0 | e\",   | 5, …  |\n",
       "|     | 198 | i\", | 2.3 | …   | mat | -Wh | act | lar | 833 | 17, | 21]    | 1 00: | \"fals | 5]    |\n",
       "|     | 6,  | \"XR | , … | 6]  | ic  | eel | Car | \",  | 4,  | …   |        | 00:00 | e\", … |       |\n",
       "|     | …   | 4Ti | 2.9 |     | 3-s | Dri | s\", | \"Re | 15. | 16] |        | , 201 | \"fals |       |\n",
       "|     | 198 | \",  | ]   |     | pd\" | ve\" | \"Co | gul | 658 |     |        | 3-01- | e\"]   |       |\n",
       "|     | 9]  | …   |     |     | ,   | ,   | mpa | ar\" | 422 |     |        | …     |       |       |\n",
       "|     |     | \"Sc |     |     | \"Ma | \"Re | ct  | , … | , … |     |        |       |       |       |\n",
       "|     |     | orp |     |     | nua | ar- | Car | \"Re | 16. |     |        |       |       |       |\n",
       "|     |     | io\" |     |     | l   | Whe | s…  | gu… | 52… |     |        |       |       |       |\n",
       "|     |     | …   |     |     | 5-… | …   |     |     |     |     |        |       |       |       |\n",
       "| Mas | [19 | [\"B | [2. | [6, | [\"M | [\"R | [\"M | [\"R | [21 | [12 | [18,   | [2013 | [\"fal | [5,   |\n",
       "| era | 85, | itu | 5,  | 6,  | anu | ear | ini | egu | .25 | ,   | 18, …  | -01-0 | se\",  | 5, …  |\n",
       "| ti  | 198 | rbo | 2.5 | …   | al  | -Wh | com | lar | 071 | 12, | 16]    | 1 00: | \"fals | 4]    |\n",
       "|     | 5,  | \",  | , … | 6]  | 5-s | eel | pac | \",  | 3,  | …   |        | 00:00 | e\", … |       |\n",
       "|     | …   | \"Bi | 2.8 |     | pd\" | Dri | t   | \"Pr | 21. | 13] |        | , 201 | \"true |       |\n",
       "|     | 199 | tur | ]   |     | ,   | ve\" | Car | emi | 250 |     |        | 3-01- | \"]    |       |\n",
       "|     | 1]  | bo\" |     |     | \"Ma | ,   | s\", | um\" | 713 |     |        | …     |       |       |\n",
       "|     |     | , … |     |     | nua | \"Re | \"Su | , … | , … |     |        |       |       |       |\n",
       "|     |     | \"43 |     |     | l   | ar- | bco | \"Pr | 21. |     |        |       |       |       |\n",
       "|     |     | 0\"… |     |     | 5-s | Whe | mpa | em… | 25… |     |        |       |       |       |\n",
       "|     |     |     |     |     | pd… | …   | …   |     |     |     |        |       |       |       |\n",
       "| AM  | [19 | [\"P | [2. | [4, | [\"A | [\"R | [\"S | [\"R | [18 | [16 | [17,   | [2013 | [\"tru | [3,   |\n",
       "| Gen | 85, | ost | 5,  | 6,  | uto | ear | pec | egu | .59 | ,   | 13, …  | -01-0 | e\",   | 3, …  |\n",
       "| era | 198 | Off | 4.2 | …   | mat | -Wh | ial | lar | 437 | 13, | 17]    | 1 00: | \"true | 3]    |\n",
       "| l   | 5,  | ice | , … | 4]  | ic  | eel | Pur | \",  | 6,  | …   |        | 00:00 | \", …  |       |\n",
       "|     | …   | DJ5 | 2.5 |     | 3-s | Dri | pos | \"Re | 22. | 18] |        | , 201 | \"true |       |\n",
       "|     | 198 | 2WD | ]   |     | pd\" | ve\" | e   | gul | 885 |     |        | 3-01- | \"]    |       |\n",
       "|     | 4]  | \",  |     |     | ,   | ,   | Veh | ar\" | 386 |     |        | …     |       |       |\n",
       "|     |     | \"Po |     |     | \"Au | \"Re | icl | , … | , … |     |        |       |       |       |\n",
       "|     |     | st  |     |     | tom | ar- | e   | \"Re | 17. |     |        |       |       |       |\n",
       "|     |     | …   |     |     | ati | Whe | 2WD | gu… | 50… |     |        |       |       |       |\n",
       "|     |     |     |     |     | c…  | …   | \"…  |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .agg(pl.all())\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76362ed5-880a-495e-99e9-1055412af3b8",
   "metadata": {},
   "source": [
    "* Limit to make, model and cylinders so the data is easier to see.\n",
    "* Get unique values for each list by using chained expression to:\n",
    "    - remove duplicates\n",
    "    - drop nulls\n",
    "    - get the length\n",
    "    - add a suffix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "04531a6c-e103-4ec8-8596-b68a54709df9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>model_len</th><th>cylinders_len</th></tr><tr><td>cat</td><td>u32</td><td>u32</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>13</td><td>3</td></tr><tr><td>&quot;Ferrari&quot;</td><td>83</td><td>3</td></tr><tr><td>&quot;Dodge&quot;</td><td>124</td><td>4</td></tr><tr><td>&quot;Subaru&quot;</td><td>78</td><td>3</td></tr><tr><td>&quot;Toyota&quot;</td><td>179</td><td>4</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Quantum Technologies&quot;</td><td>1</td><td>1</td></tr><tr><td>&quot;London Taxi&quot;</td><td>1</td><td>1</td></tr><tr><td>&quot;Excalibur Autos&quot;</td><td>1</td><td>1</td></tr><tr><td>&quot;Mahindra&quot;</td><td>1</td><td>1</td></tr><tr><td>&quot;Azure Dynamics&quot;</td><td>1</td><td>0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 3)\n",
       "+----------------------+-----------+---------------+\n",
       "| make                 | model_len | cylinders_len |\n",
       "| ---                  | ---       | ---           |\n",
       "| cat                  | u32       | u32           |\n",
       "+==================================================+\n",
       "| Alfa Romeo           | 13        | 3             |\n",
       "| Ferrari              | 83        | 3             |\n",
       "| Dodge                | 124       | 4             |\n",
       "| Subaru               | 78        | 3             |\n",
       "| Toyota               | 179       | 4             |\n",
       "| …                    | …         | …             |\n",
       "| Quantum Technologies | 1         | 1             |\n",
       "| London Taxi          | 1         | 1             |\n",
       "| Excalibur Autos      | 1         | 1             |\n",
       "| Mahindra             | 1         | 1             |\n",
       "| Azure Dynamics       | 1         | 0             |\n",
       "+----------------------+-----------+---------------+"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select('make', 'model', 'cylinders')\n",
    "    .group_by('make')\n",
    "    .agg(\n",
    "        pl.all()\n",
    "        .unique()\n",
    "        .drop_nulls()\n",
    "        .len() # akin to value_counts\n",
    "        .name.suffix('_len')\n",
    "    )\n",
    "    .sort(by='make')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eade53fe-d92e-4543-aaaf-55dc023d1c24",
   "metadata": {},
   "source": [
    "* These are the aggregated values.\n",
    "* There are only 136 makes, so this dataframe has only 136 rows.\n",
    "  \n",
    "If we wanted these *'_len'* columns added to the **original dataframe with no loss of rows**, use window functions.\n",
    "* Move the `.group_by('make')` to an `.over('make')`.\n",
    "* Use `.with_columns` instead of `.select`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "a45db136-f737-4c39-b6e4-e2b51f0d7ec0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 17)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th><th>model_len</th><th>cylinders_len</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td><td>u32</td><td>u32</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>13</td><td>3</td></tr><tr><td>1985</td><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>4.9</td><td>12</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>9</td><td>14</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>83</td><td>3</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;Charger&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Subcompact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.018888</td><td>23</td><td>33</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>124</td><td>4</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;B150/B250 Wagon 2WD&quot;</td><td>5.2</td><td>8</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Vans&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>10</td><td>12</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td><td>124</td><td>4</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>15.658422</td><td>17</td><td>23</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>78</td><td>3</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>13.523182</td><td>19</td><td>26</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td><td>78</td><td>3</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>12.935218</td><td>20</td><td>28</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>78</td><td>3</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td><td>78</td><td>3</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td><td>78</td><td>3</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>16.528334</td><td>16</td><td>21</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td><td>78</td><td>3</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 17)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | hig | cre | is_ | num | mod | cyl |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | hwa | ate | aut | _ge | el_ | ind |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | y08 | dOn | oma | ars | len | ers |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | --- | --- | tic | --- | --- | _le |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     | u8  | dat | --- | u8  | u32 | n   |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | eti | str |     |     | --- |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | me[ |     |     |     | u32 |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | μs] |     |     |     |     |\n",
       "+=====================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25  | 201 | fal | 5   | 13  | 3   |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |     | 3-0 | se  |     |     |     |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |     | 1-0 |     |     |     |     |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |     |     |\n",
       "| 198 | Fer | Tes | 4.9 | 12  | Man | Rea | Two | Reg | 27. | 9   | 14  | 201 | fal | 5   | 83  | 3   |\n",
       "| 5   | rar | tar |     |     | ual | r-W | Sea | ula | 046 |     |     | 3-0 | se  |     |     |     |\n",
       "|     | i   | oss |     |     | 5-s | hee | ter | r   | 364 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     | a   |     |     | pd  | l   | s   |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |     |     |\n",
       "| 198 | Dod | Cha | 2.2 | 4   | Man | Fro | Sub | Reg | 11. | 23  | 33  | 201 | fal | 5   | 124 | 4   |\n",
       "| 5   | ge  | rge |     |     | ual | nt- | com | ula | 018 |     |     | 3-0 | se  |     |     |     |\n",
       "|     |     | r   |     |     | 5-s | Whe | pac | r   | 888 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     |     |     |     | pd  | el  | t   |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri | Car |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |     |     |\n",
       "| 198 | Dod | B15 | 5.2 | 8   | Aut | Rea | Van | Reg | 27. | 10  | 12  | 201 | tru | 3   | 124 | 4   |\n",
       "| 5   | ge  | 0/B |     |     | oma | r-W | s   | ula | 046 |     |     | 3-0 | e   |     |     |     |\n",
       "|     |     | 250 |     |     | tic | hee |     | r   | 364 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     | Wag |     |     | 3-s | l   |     |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     | on  |     |     | pd  | Dri |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     | 2WD |     |     |     | ve  |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |     |     |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | 4-W | Com | Pre | 15. | 17  | 23  | 201 | fal | 5   | 78  | 3   |\n",
       "| 3   | aru | acy |     |     | ual | hee | pac | miu | 658 |     |     | 3-0 | se  |     |     |     |\n",
       "|     |     | AWD |     |     | 5-s | l   | t   | m   | 422 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     | Tur |     |     | pd  | or  | Car |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     | bo  |     |     |     | All | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |     |     |\n",
       "| …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | Fro | Com | Reg | 13. | 19  | 26  | 201 | tru | 4   | 78  | 3   |\n",
       "| 3   | aru | acy |     |     | oma | nt- | pac | ula | 523 |     |     | 3-0 | e   |     |     |     |\n",
       "|     |     |     |     |     | tic | Whe | t   | r   | 182 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     |     |     |     | 4-s | el  | Car |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |     |     |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | Fro | Com | Reg | 12. | 20  | 28  | 201 | fal | 5   | 78  | 3   |\n",
       "| 3   | aru | acy |     |     | ual | nt- | pac | ula | 935 |     |     | 3-0 | se  |     |     |     |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 218 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |     | 00  |     |     |     |     |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | 4-W | Com | Reg | 14. | 18  | 24  | 201 | tru | 4   | 78  | 3   |\n",
       "| 3   | aru | acy |     |     | oma | hee | pac | ula | 167 |     |     | 3-0 | e   |     |     |     |\n",
       "|     |     | AWD |     |     | tic | l   | t   | r   | 143 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     |     |     |     | 4-s | or  | Car |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     |     |     |     | pd  | All | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |     |     |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | 4-W | Com | Reg | 14. | 18  | 24  | 201 | fal | 5   | 78  | 3   |\n",
       "| 3   | aru | acy |     |     | ual | hee | pac | ula | 167 |     |     | 3-0 | se  |     |     |     |\n",
       "|     |     | AWD |     |     | 5-s | l   | t   | r   | 143 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     |     |     |     | pd  | or  | Car |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     |     |     |     |     | All | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |     |     |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | 4-W | Com | Pre | 16. | 16  | 21  | 201 | tru | 4   | 78  | 3   |\n",
       "| 3   | aru | acy |     |     | oma | hee | pac | miu | 528 |     |     | 3-0 | e   |     |     |     |\n",
       "|     |     | AWD |     |     | tic | l   | t   | m   | 334 |     |     | 1-0 |     |     |     |     |\n",
       "|     |     | Tur |     |     | 4-s | or  | Car |     |     |     |     | 1   |     |     |     |     |\n",
       "|     |     | bo  |     |     | pd  | All | s   |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |     | 00: |     |     |     |     |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |     | 00  |     |     |     |     |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |     |     |     |     |     |     |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |     |     |     |     |     |     |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .with_columns(\n",
    "        pl.col('model', 'cylinders')\n",
    "            .unique().drop_nulls().len()\n",
    "            .over('make')  \n",
    "            .name.suffix('_len')\n",
    "        )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3dd6da3a-07a6-4485-8fad-715d15bceae1",
   "metadata": {},
   "source": [
    "---\n",
    "To get the count of every column:\n",
    "\n",
    "* Change the selector to use all of the columns.\n",
    "* Remove the *make* column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "f82ae398-4bd1-4a88-92c6-1a9d8f62ac20",
   "metadata": {},
   "outputs": [],
   "source": [
    "# (\n",
    "#     autos\n",
    "#     .select(\n",
    "#         (cs.all() - pl.col('make'))\n",
    "#             .unique().drop_nulls().len()\n",
    "#             .over('make')\n",
    "#             #.name.suffix('_count')\n",
    "#     )\n",
    "# )\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8db85e39-51dd-4919-9858-2d5b4c6eeb1c",
   "metadata": {},
   "source": [
    "---\n",
    "# 4.11 Chapter Methods and Functions\n",
    "\n",
    "[back to Top](#Top)\n",
    "\n",
    "See full list of aggregation functions here:\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/aggregation.html\n",
    "\n",
    "Selected list below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "d74e06ad-24b5-4d24-86cd-d5e96d26123c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (145, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>len</th></tr><tr><td>cat</td><td>u32</td></tr></thead><tbody><tr><td>&quot;Chevrolet&quot;</td><td>4472</td></tr><tr><td>&quot;Ford&quot;</td><td>3834</td></tr><tr><td>&quot;GMC&quot;</td><td>2823</td></tr><tr><td>&quot;Dodge&quot;</td><td>2695</td></tr><tr><td>&quot;BMW&quot;</td><td>2536</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Goldacre&quot;</td><td>1</td></tr><tr><td>&quot;Panos&quot;</td><td>1</td></tr><tr><td>&quot;General Motors&quot;</td><td>1</td></tr><tr><td>&quot;Lordstown&quot;</td><td>1</td></tr><tr><td>&quot;London Coach Co Inc&quot;</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (145, 2)\n",
       "+---------------------+------+\n",
       "| make                | len  |\n",
       "| ---                 | ---  |\n",
       "| cat                 | u32  |\n",
       "+============================+\n",
       "| Chevrolet           | 4472 |\n",
       "| Ford                | 3834 |\n",
       "| GMC                 | 2823 |\n",
       "| Dodge               | 2695 |\n",
       "| BMW                 | 2536 |\n",
       "| …                   | …    |\n",
       "| Goldacre            | 1    |\n",
       "| Panos               | 1    |\n",
       "| General Motors      | 1    |\n",
       "| Lordstown           | 1    |\n",
       "| London Coach Co Inc | 1    |\n",
       "+---------------------+------+"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by('make')\n",
    "    .len()\n",
    "    .sort(by=['len'], descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "af0e94b1-43e0-466f-91f2-c5d5e6c443af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1985</td><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>4.9</td><td>12</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>9</td><td>14</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;Charger&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Subcompact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.018888</td><td>23</td><td>33</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25     | 2013- | false | 5     |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 198 | Fer | Tes | 4.9 | 12  | Man | Rea | Two | Reg | 27. | 9   | 14     | 2013- | false | 5     |\n",
       "| 5   | rar | tar |     |     | ual | r-W | Sea | ula | 046 |     |        | 01-01 |       |       |\n",
       "|     | i   | oss |     |     | 5-s | hee | ter | r   | 364 |     |        | 00:00 |       |       |\n",
       "|     |     | a   |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 198 | Dod | Cha | 2.2 | 4   | Man | Fro | Sub | Reg | 11. | 23  | 33     | 2013- | false | 5     |\n",
       "| 5   | ge  | rge |     |     | ual | nt- | com | ula | 018 |     |        | 01-01 |       |       |\n",
       "|     |     | r   |     |     | 5-s | Whe | pac | r   | 888 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | t   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | Car |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  | s   |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.all.html\n",
    "autos.select(pl.all(ignore_nulls=False)).head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "e56acd3e-8724-4559-970a-6c115d8705c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.any.html\n",
    "# Syntactic sugar for col(names).any().\n",
    "# autos.select(pl.any(\"make\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "bb056da6-c282-4588-bf06-0d3626eefb71",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th></tr><tr><td>u32</td></tr></thead><tbody><tr><td>24039</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 1)\n",
       "+-------+\n",
       "| make  |\n",
       "| ---   |\n",
       "| u32   |\n",
       "+=======+\n",
       "| 24039 |\n",
       "+-------+"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.arg_max.html\n",
    "# Get the index of the maximal value.\n",
    "autos.select(pl.col(\"make\").arg_max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "70b31713-07dd-4668-94ee-bbd0e713c5aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th></tr><tr><td>u32</td></tr></thead><tbody><tr><td>0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 1)\n",
       "+------+\n",
       "| make |\n",
       "| ---  |\n",
       "| u32  |\n",
       "+======+\n",
       "| 0    |\n",
       "+------+"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.min.html#polars.Expr.min\n",
    "# Get minimum value.\n",
    "autos.select(pl.col(\"make\").arg_min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "6b24f659-d284-462c-b358-1075aa5a8f2c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th></tr><tr><td>cat</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td></tr><tr><td>&quot;Ferrari&quot;</td></tr><tr><td>&quot;Dodge&quot;</td></tr><tr><td>&quot;Dodge&quot;</td></tr><tr><td>&quot;Subaru&quot;</td></tr><tr><td>&hellip;</td></tr><tr><td>&quot;Subaru&quot;</td></tr><tr><td>&quot;Subaru&quot;</td></tr><tr><td>&quot;Subaru&quot;</td></tr><tr><td>&quot;Subaru&quot;</td></tr><tr><td>&quot;Subaru&quot;</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 1)\n",
       "+------------+\n",
       "| make       |\n",
       "| ---        |\n",
       "| cat        |\n",
       "+============+\n",
       "| Alfa Romeo |\n",
       "| Ferrari    |\n",
       "| Dodge      |\n",
       "| Dodge      |\n",
       "| Subaru     |\n",
       "| …          |\n",
       "| Subaru     |\n",
       "| Subaru     |\n",
       "| Subaru     |\n",
       "| Subaru     |\n",
       "| Subaru     |\n",
       "+------------+"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autos.select(pl.col('make'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "42019480-044e-4ab8-974d-88aebbfa9d01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th></tr><tr><td>list[cat]</td></tr></thead><tbody><tr><td>[&quot;Alfa Romeo&quot;, &quot;Ferrari&quot;, … &quot;Subaru&quot;]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 1)\n",
       "+---------------------------------+\n",
       "| make                            |\n",
       "| ---                             |\n",
       "| list[cat]                       |\n",
       "+=================================+\n",
       "| [\"Alfa Romeo\", \"Ferrari\", … \"S… |\n",
       "+---------------------------------+"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.implode.html\n",
    "# Aggregate all column values into a list.\n",
    "# This function is syntactic sugar for pl.col(name).implode().\n",
    "autos.select(pl.col('make').implode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "b8209f59-864b-4ca9-8b29-a7edc498ac38",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td></tr></thead><tbody><tr><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td><td>48202</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| u32 | u32 | u32 | u32 | --- | u32 | u32 | u32 | --- | --- | u32 | u32    | u32   | ---   | u32   |\n",
       "|     |     |     |     | u32 |     |     |     | u32 | u32 |     |        |       | u32   |       |\n",
       "+==================================================================================================+\n",
       "| 482 | 482 | 482 | 482 | 482 | 482 | 482 | 482 | 482 | 482 | 482 | 48202  | 48202 | 48202 | 48202 |\n",
       "| 02  | 02  | 02  | 02  | 02  | 02  | 02  | 02  | 02  | 02  | 02  |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.len.html\n",
    "# Return the number of rows in the context.\n",
    "autos.select(pl.all().len())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "c0d37468-67c3-4bc2-a38c-35e94eed8163",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>str</td><td>str</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>2025</td><td>&quot;Azure Dynamics&quot;</td><td>&quot;ActiveHybrid 7&quot;</td><td>8.4</td><td>16</td><td>&quot;Manual 7-spd&quot;</td><td>&quot;Part-time 4-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicles/4wd&quot;</td><td>&quot;Hydrogen&quot;</td><td>42.501427</td><td>153</td><td>142</td><td>2024-10-17 00:00:00</td><td>&quot;true&quot;</td><td>10</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | str | str | f32 | --- | str | str | str | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | str | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 202 | Azu | Act | 8.4 | 16  | Man | Par | Spe | Hyd | 42. | 153 | 142    | 2024- | true  | 10    |\n",
       "| 5   | re  | ive |     |     | ual | t-t | cia | rog | 501 |     |        | 10-17 |       |       |\n",
       "|     | Dyn | Hyb |     |     | 7-s | ime | l   | en  | 427 |     |        | 00:00 |       |       |\n",
       "|     | ami | rid |     |     | pd  | 4-W | Pur |     |     |     |        | :00   |       |       |\n",
       "|     | cs  | 7   |     |     |     | hee | pos |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | l   | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | Dri | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | es/ |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 4wd |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.max.html\n",
    "# Get maximum value.\n",
    "autos.select(pl.all().max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "b61bb051-bead-41af-bec1-15e5e5ede7e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>f64</td><td>cat</td><td>cat</td><td>f32</td><td>f64</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>f64</td><td>f64</td><td>datetime[μs]</td><td>str</td><td>f64</td></tr></thead><tbody><tr><td>2004.59344</td><td>null</td><td>null</td><td>3.27066</td><td>5.698838</td><td>null</td><td>null</td><td>null</td><td>null</td><td>15.106042</td><td>19.929235</td><td>25.821294</td><td>2014-11-14 05:34:17.590971</td><td>null</td><td>5.346666</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| f64 | cat | cat | f32 | --- | str | cat | cat | --- | --- | f64 | f64    | datet | ---   | f64   |\n",
       "|     |     |     |     | f64 |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 200 | nul | nul | 3.2 | 5.6 | nul | nul | nul | nul | 15. | 19. | 25.821 | 2014- | null  | 5.346 |\n",
       "| 4.5 | l   | l   | 706 | 988 | l   | l   | l   | l   | 106 | 929 | 294    | 11-14 |       | 666   |\n",
       "| 934 |     |     | 6   | 38  |     |     |     |     | 042 | 235 |        | 05:34 |       |       |\n",
       "| 4   |     |     |     |     |     |     |     |     |     |     |        | :17.5 |       |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | 90971 |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.mean.html\n",
    "# Get mean value.\n",
    "autos.select(pl.all().mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "d5b21266-c1f5-45ff-a41d-dd7e7a0ecece",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "dtype_columns([UInt16, Float32, Int8, Decimal(None, None), Int64, UInt32, Float64, Int16, UInt8, UInt64, Int32])"
      ],
      "text/plain": [
       "cs.numeric()"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# .median() can be applied only on numeric types\n",
    "# pl.col(pl.NUMERIC_DTYPES) DEPRECATED, use\n",
    "cs.numeric()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "d677937e-22e1-42ed-a424-0dcf91d680c2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>f64</td><td>f32</td><td>f64</td><td>f32</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>2006.0</td><td>3.0</td><td>6.0</td><td>14.8755</td><td>18.0</td><td>24.0</td><td>5.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+--------+-------+-----------+-----------+--------+-----------+-----------+\n",
       "| year   | displ | cylinders | barrels08 | city08 | highway08 | num_gears |\n",
       "| ---    | ---   | ---       | ---       | ---    | ---       | ---       |\n",
       "| f64    | f32   | f64       | f32       | f64    | f64       | f64       |\n",
       "+=========================================================================+\n",
       "| 2006.0 | 3.0   | 6.0       | 14.8755   | 18.0   | 24.0      | 5.0       |\n",
       "+--------+-------+-----------+-----------+--------+-----------+-----------+"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.median.html\n",
    "# Get median value using linear interpolation.\n",
    "autos.select(cs.numeric()).median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "f7517a81-cf10-434c-8293-9795001e28cf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>f64</td><td>f32</td><td>f64</td><td>f32</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>2006.0</td><td>3.0</td><td>6.0</td><td>14.8755</td><td>18.0</td><td>24.0</td><td>5.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+--------+-------+-----------+-----------+--------+-----------+-----------+\n",
       "| year   | displ | cylinders | barrels08 | city08 | highway08 | num_gears |\n",
       "| ---    | ---   | ---       | ---       | ---    | ---       | ---       |\n",
       "| f64    | f32   | f64       | f32       | f64    | f64       | f64       |\n",
       "+=========================================================================+\n",
       "| 2006.0 | 3.0   | 6.0       | 14.8755   | 18.0   | 24.0      | 5.0       |\n",
       "+--------+-------+-----------+-----------+--------+-----------+-----------+"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Equivalent\n",
    "autos.select(cs.numeric()).median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "8084d6fa-2c11-4b25-b0c2-fffd334f2699",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>str</td><td>str</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1984</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Eagle&quot;</td><td>0.0</td><td>2</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>0.047081</td><td>6</td><td>9</td><td>2013-01-01 00:00:00</td><td>&quot;Automatic&quot;</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | str | str | f32 | --- | str | str | str | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | str | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Eag | 0.0 | 2   | Aut | Rea | Two | Reg | 0.0 | 6   | 9      | 2013- | Autom | 1     |\n",
       "| 4   | a   | le  |     |     | oma | r-W | Sea | ula | 470 |     |        | 01-01 | atic  |       |\n",
       "|     | Rom |     |     |     | tic | hee | ter | r   | 81  |     |        | 00:00 |       |       |\n",
       "|     | eo  |     |     |     | (A1 | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.min.html\n",
    "# Get minimum value.\n",
    "autos.select(pl.all().min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "25d3faf0-f4bd-4cff-9005-c72c07f7e87a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td><td>u32</td></tr></thead><tbody><tr><td>42</td><td>145</td><td>5166</td><td>67</td><td>10</td><td>41</td><td>8</td><td>34</td><td>15</td><td>323</td><td>139</td><td>125</td><td>459</td><td>3</td><td>10</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| u32 | u32 | u32 | u32 | --- | u32 | u32 | u32 | --- | --- | u32 | u32    | u32   | ---   | u32   |\n",
       "|     |     |     |     | u32 |     |     |     | u32 | u32 |     |        |       | u32   |       |\n",
       "+==================================================================================================+\n",
       "| 42  | 145 | 516 | 67  | 10  | 41  | 8   | 34  | 15  | 323 | 139 | 125    | 459   | 3     | 10    |\n",
       "|     |     | 6   |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.n_unique.html#polars.Expr.n_unique\n",
    "# Count unique values.\n",
    "# null is considered to be a unique value for the purposes of this operation.\n",
    "autos.select(pl.all().n_unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "e30d3b31-286c-48b8-a992-e7bf9c407230",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>str</td><td>str</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>2025</td><td>&quot;Azure Dynamics&quot;</td><td>&quot;ActiveHybrid 7&quot;</td><td>8.4</td><td>16</td><td>&quot;Manual 7-spd&quot;</td><td>&quot;Part-time 4-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicles/4wd&quot;</td><td>&quot;Hydrogen&quot;</td><td>42.501427</td><td>153</td><td>142</td><td>2024-10-17 00:00:00</td><td>&quot;true&quot;</td><td>10</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | str | str | f32 | --- | str | str | str | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | str | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 202 | Azu | Act | 8.4 | 16  | Man | Par | Spe | Hyd | 42. | 153 | 142    | 2024- | true  | 10    |\n",
       "| 5   | re  | ive |     |     | ual | t-t | cia | rog | 501 |     |        | 10-17 |       |       |\n",
       "|     | Dyn | Hyb |     |     | 7-s | ime | l   | en  | 427 |     |        | 00:00 |       |       |\n",
       "|     | ami | rid |     |     | pd  | 4-W | Pur |     |     |     |        | :00   |       |       |\n",
       "|     | cs  | 7   |     |     |     | hee | pos |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | l   | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | Dri | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | es/ |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 4wd |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.nan_max.html\n",
    "# Get maximum value, but propagate/poison encountered NaN values.\n",
    "# This differs from numpy’s nanmax as numpy defaults to propagating NaN values, whereas polars defaults to ignoring them.\n",
    "autos.select(pl.all().nan_max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "98b39b4e-a5f8-4f2d-9b6b-8d1774d5f8c7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>str</td><td>str</td><td>str</td><td>str</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1984</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Eagle&quot;</td><td>0.0</td><td>2</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>0.047081</td><td>6</td><td>9</td><td>2013-01-01 00:00:00</td><td>&quot;Automatic&quot;</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | str | str | f32 | --- | str | str | str | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | str | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Eag | 0.0 | 2   | Aut | Rea | Two | Reg | 0.0 | 6   | 9      | 2013- | Autom | 1     |\n",
       "| 4   | a   | le  |     |     | oma | r-W | Sea | ula | 470 |     |        | 01-01 | atic  |       |\n",
       "|     | Rom |     |     |     | tic | hee | ter | r   | 81  |     |        | 00:00 |       |       |\n",
       "|     | eo  |     |     |     | (A1 | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | )   | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.nan_min.html\n",
    "# Get minimum value, but propagate/poison encountered NaN values.\n",
    "# This differs from numpy’s nanmax as numpy defaults to propagating NaN values, whereas polars defaults to ignoring them.\n",
    "autos.select(pl.all().nan_min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "64a8351a-1d17-4857-ace8-de268fa1f434",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>i64</td><td>f32</td><td>i64</td><td>f32</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>0</td><td>NaN</td><td>0</td><td>inf</td><td>0</td><td>0</td><td>0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+------+-------+-----------+-----------+--------+-----------+-----------+\n",
       "| year | displ | cylinders | barrels08 | city08 | highway08 | num_gears |\n",
       "| ---  | ---   | ---       | ---       | ---    | ---       | ---       |\n",
       "| i64  | f32   | i64       | f32       | i64    | i64       | i64       |\n",
       "+=======================================================================+\n",
       "| 0    | NaN   | 0         | inf       | 0      | 0         | 0         |\n",
       "+------+-------+-----------+-----------+--------+-----------+-----------+"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.product.html\n",
    "# Compute the product of an expression.\n",
    "autos.select(cs.numeric().product())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "4d24f897-39d9-423a-be45-66cb9731a4f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>i64</td><td>f32</td><td>i64</td><td>f32</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>0</td><td>inf</td><td>0</td><td>inf</td><td>0</td><td>0</td><td>0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+------+-------+-----------+-----------+--------+-----------+-----------+\n",
       "| year | displ | cylinders | barrels08 | city08 | highway08 | num_gears |\n",
       "| ---  | ---   | ---       | ---       | ---    | ---       | ---       |\n",
       "| i64  | f32   | i64       | f32       | i64    | i64       | i64       |\n",
       "+=======================================================================+\n",
       "| 0    | inf   | 0         | inf       | 0      | 0         | 0         |\n",
       "+------+-------+-----------+-----------+--------+-----------+-----------+"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# variation\n",
    "autos.select(cs.numeric().filter(cs.numeric() != 0).product())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "ee817472-0c83-4f24-8f38-c4dc36297c41",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>f64</td><td>f32</td><td>f64</td><td>f32</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>1995.0</td><td>2.3</td><td>4.0</td><td>12.935218</td><td>15.0</td><td>21.0</td><td>4.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+--------+-------+-----------+-----------+--------+-----------+-----------+\n",
       "| year   | displ | cylinders | barrels08 | city08 | highway08 | num_gears |\n",
       "| ---    | ---   | ---       | ---       | ---    | ---       | ---       |\n",
       "| f64    | f32   | f64       | f32       | f64    | f64       | f64       |\n",
       "+=========================================================================+\n",
       "| 1995.0 | 2.3   | 4.0       | 12.935218 | 15.0   | 21.0      | 4.0       |\n",
       "+--------+-------+-----------+-----------+--------+-----------+-----------+"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.quantile.html\n",
    "# Get quantile value.\n",
    "autos.select(cs.numeric().quantile(0.3, interpolation='nearest'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "2e134e14-663c-4dd7-b599-9d60ddacefa3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>f64</td><td>f32</td><td>f64</td><td>f32</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>12.681444</td><td>1.353053</td><td>1.772964</td><td>4.560156</td><td>12.658018</td><td>11.087154</td><td>1.73496</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+-----------+----------+-----------+-----------+-----------+-----------+-----------+\n",
       "| year      | displ    | cylinders | barrels08 | city08    | highway08 | num_gears |\n",
       "| ---       | ---      | ---       | ---       | ---       | ---       | ---       |\n",
       "| f64       | f32      | f64       | f32       | f64       | f64       | f64       |\n",
       "+==================================================================================+\n",
       "| 12.681444 | 1.353053 | 1.772964  | 4.560156  | 12.658018 | 11.087154 | 1.73496   |\n",
       "+-----------+----------+-----------+-----------+-----------+-----------+-----------+"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.std.html\n",
    "# Get standard deviation.\n",
    "autos.select(cs.numeric().std(ddof=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "174a0bde-60a2-462f-ad9a-fb1b82599b29",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>i64</td><td>f32</td><td>i64</td><td>f32</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>96625413</td><td>154502.703125</td><td>269196</td><td>728141.4375</td><td>960629</td><td>1244638</td><td>257720</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+----------+---------------+-----------+-------------+--------+-----------+-----------+\n",
       "| year     | displ         | cylinders | barrels08   | city08 | highway08 | num_gears |\n",
       "| ---      | ---           | ---       | ---         | ---    | ---       | ---       |\n",
       "| i64      | f32           | i64       | f32         | i64    | i64       | i64       |\n",
       "+=====================================================================================+\n",
       "| 96625413 | 154502.703125 | 269196    | 728141.4375 | 960629 | 1244638   | 257720    |\n",
       "+----------+---------------+-----------+-------------+--------+-----------+-----------+"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.sum.html\n",
    "# Get sum value.\n",
    "# Dtypes in {Int8, UInt8, Int16, UInt16} are cast to Int64 before summing to prevent overflow issues.\n",
    "autos.select(cs.numeric().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "e797b056-bb45-4672-bda1-a4375ba03b2b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 7)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>displ</th><th>cylinders</th><th>barrels08</th><th>city08</th><th>highway08</th><th>num_gears</th></tr><tr><td>f64</td><td>f32</td><td>f64</td><td>f32</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>160.819021</td><td>1.830753</td><td>3.143401</td><td>20.795021</td><td>160.225423</td><td>122.924973</td><td>3.010087</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 7)\n",
       "+------------+----------+-----------+-----------+------------+------------+-----------+\n",
       "| year       | displ    | cylinders | barrels08 | city08     | highway08  | num_gears |\n",
       "| ---        | ---      | ---       | ---       | ---        | ---        | ---       |\n",
       "| f64        | f32      | f64       | f32       | f64        | f64        | f64       |\n",
       "+=====================================================================================+\n",
       "| 160.819021 | 1.830753 | 3.143401  | 20.795021 | 160.225423 | 122.924973 | 3.010087  |\n",
       "+------------+----------+-----------+-----------+------------+------------+-----------+"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.var.html\n",
    "# Get variance.\n",
    "autos.select(cs.numeric().var(ddof=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "523c7f54-1e49-4c2b-8f32-581c69a6fc8e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>model</th><th>fuelType</th></tr><tr><td>cat</td><td>cat</td><td>cat</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Dodge&quot;</td><td>&quot;Charger&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Dodge&quot;</td><td>&quot;B150/B250 Wagon 2WD&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>&quot;Premium&quot;</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>&quot;Regular&quot;</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>&quot;Premium&quot;</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 3)\n",
       "+------------+---------------------+----------+\n",
       "| make       | model               | fuelType |\n",
       "| ---        | ---                 | ---      |\n",
       "| cat        | cat                 | cat      |\n",
       "+=============================================+\n",
       "| Alfa Romeo | Spider Veloce 2000  | Regular  |\n",
       "| Ferrari    | Testarossa          | Regular  |\n",
       "| Dodge      | Charger             | Regular  |\n",
       "| Dodge      | B150/B250 Wagon 2WD | Regular  |\n",
       "| Subaru     | Legacy AWD Turbo    | Premium  |\n",
       "| …          | …                   | …        |\n",
       "| Subaru     | Legacy              | Regular  |\n",
       "| Subaru     | Legacy              | Regular  |\n",
       "| Subaru     | Legacy AWD          | Regular  |\n",
       "| Subaru     | Legacy AWD          | Regular  |\n",
       "| Subaru     | Legacy AWD Turbo    | Premium  |\n",
       "+------------+---------------------+----------+"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/selectors.html#polars.selectors.matches\n",
    "# Select all columns that match the given regex pattern.\n",
    "# pattern: A valid regular expression pattern, compatible with the regex crate.(\n",
    "autos.select(cs.matches(\"(^f|^m)\")) # select columns starting with f or m"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "100ee1d2-d81b-4f4a-9fd2-9bc2e4b4d2df",
   "metadata": {},
   "source": [
    "## pl.concat_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "32d7f5f2-e0a8-4fbf-b4b7-654f26958f3c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>concatenated</th><th>concatenated2</th></tr><tr><td>list[cat]</td><td>list[f32]</td></tr></thead><tbody><tr><td>[&quot;Alfa Romeo&quot;, &quot;Spider Veloce 2000&quot;]</td><td>[2.0, 4.0]</td></tr><tr><td>[&quot;Ferrari&quot;, &quot;Testarossa&quot;]</td><td>[4.9, 12.0]</td></tr><tr><td>[&quot;Dodge&quot;, &quot;Charger&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Dodge&quot;, &quot;B150/B250 Wagon 2WD&quot;]</td><td>[5.2, 8.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD Turbo&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD Turbo&quot;]</td><td>[2.2, 4.0]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 2)\n",
       "+---------------------------------+---------------+\n",
       "| concatenated                    | concatenated2 |\n",
       "| ---                             | ---           |\n",
       "| list[cat]                       | list[f32]     |\n",
       "+=================================================+\n",
       "| [\"Alfa Romeo\", \"Spider Veloce … | [2.0, 4.0]    |\n",
       "| [\"Ferrari\", \"Testarossa\"]       | [4.9, 12.0]   |\n",
       "| [\"Dodge\", \"Charger\"]            | [2.2, 4.0]    |\n",
       "| [\"Dodge\", \"B150/B250 Wagon 2WD… | [5.2, 8.0]    |\n",
       "| [\"Subaru\", \"Legacy AWD Turbo\"]  | [2.2, 4.0]    |\n",
       "| …                               | …             |\n",
       "| [\"Subaru\", \"Legacy\"]            | [2.2, 4.0]    |\n",
       "| [\"Subaru\", \"Legacy\"]            | [2.2, 4.0]    |\n",
       "| [\"Subaru\", \"Legacy AWD\"]        | [2.2, 4.0]    |\n",
       "| [\"Subaru\", \"Legacy AWD\"]        | [2.2, 4.0]    |\n",
       "| [\"Subaru\", \"Legacy AWD Turbo\"]  | [2.2, 4.0]    |\n",
       "+---------------------------------+---------------+"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.concat_list.html\n",
    "# Horizontally concatenate columns into a single list column.\n",
    "# Concatenate multiple columns into a single list column\n",
    "# Note that pl.concat_list is marked for deprecation in favor of the \n",
    "# pl.col(\"column_name\").list.concat(...) syntax, which provides a more consistent API.\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        # very slow with strings\n",
    "        concatenated = pl.concat_list(['make', 'model']), \n",
    "\n",
    "        # very fast with numeric\n",
    "        concatenated2 = pl.concat_list(['displ', 'cylinders']) \n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a2f2590a-55b3-46d8-b172-40ff359fa59b",
   "metadata": {},
   "source": [
    "## pl.col().list.concat()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "fff06a28-e647-4fa3-a26d-10cd614edd8e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>concatenated3</th><th>concatenated</th></tr><tr><td>list[cat]</td><td>list[f32]</td></tr></thead><tbody><tr><td>[&quot;Alfa Romeo&quot;, &quot;Spider Veloce 2000&quot;]</td><td>[2.0, 4.0]</td></tr><tr><td>[&quot;Ferrari&quot;, &quot;Testarossa&quot;]</td><td>[4.9, 12.0]</td></tr><tr><td>[&quot;Dodge&quot;, &quot;Charger&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Dodge&quot;, &quot;B150/B250 Wagon 2WD&quot;]</td><td>[5.2, 8.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD Turbo&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD&quot;]</td><td>[2.2, 4.0]</td></tr><tr><td>[&quot;Subaru&quot;, &quot;Legacy AWD Turbo&quot;]</td><td>[2.2, 4.0]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 2)\n",
       "+---------------------------------+--------------+\n",
       "| concatenated3                   | concatenated |\n",
       "| ---                             | ---          |\n",
       "| list[cat]                       | list[f32]    |\n",
       "+================================================+\n",
       "| [\"Alfa Romeo\", \"Spider Veloce … | [2.0, 4.0]   |\n",
       "| [\"Ferrari\", \"Testarossa\"]       | [4.9, 12.0]  |\n",
       "| [\"Dodge\", \"Charger\"]            | [2.2, 4.0]   |\n",
       "| [\"Dodge\", \"B150/B250 Wagon 2WD… | [5.2, 8.0]   |\n",
       "| [\"Subaru\", \"Legacy AWD Turbo\"]  | [2.2, 4.0]   |\n",
       "| …                               | …            |\n",
       "| [\"Subaru\", \"Legacy\"]            | [2.2, 4.0]   |\n",
       "| [\"Subaru\", \"Legacy\"]            | [2.2, 4.0]   |\n",
       "| [\"Subaru\", \"Legacy AWD\"]        | [2.2, 4.0]   |\n",
       "| [\"Subaru\", \"Legacy AWD\"]        | [2.2, 4.0]   |\n",
       "| [\"Subaru\", \"Legacy AWD Turbo\"]  | [2.2, 4.0]   |\n",
       "+---------------------------------+--------------+"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Note that pl.concat_list is marked for deprecation in favor of the pl.col(\"column_name\").list.concat(...) syntax, which provides a more consistent API.\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        concatenated3 = pl.col(['make']).list.concat('model'),\n",
    "        concatenated = pl.col(['displ']).list.concat('cylinders')\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "4b76c6a1-71bb-4537-9d8a-030b10a49789",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: (5, 1)\n",
      "+------+\n",
      "| A    |\n",
      "| ---  |\n",
      "| f64  |\n",
      "+======+\n",
      "| 1.0  |\n",
      "| 2.0  |\n",
      "| 9.0  |\n",
      "| 2.0  |\n",
      "| 13.0 |\n",
      "+------+\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>A_rolling</th></tr><tr><td>list[f64]</td></tr></thead><tbody><tr><td>[null, null, 1.0]</td></tr><tr><td>[null, 1.0, 2.0]</td></tr><tr><td>[1.0, 2.0, 9.0]</td></tr><tr><td>[2.0, 9.0, 2.0]</td></tr><tr><td>[9.0, 2.0, 13.0]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 1)\n",
       "+-------------------+\n",
       "| A_rolling         |\n",
       "| ---               |\n",
       "| list[f64]         |\n",
       "+===================+\n",
       "| [null, null, 1.0] |\n",
       "| [null, 1.0, 2.0]  |\n",
       "| [1.0, 2.0, 9.0]   |\n",
       "| [2.0, 9.0, 2.0]   |\n",
       "| [9.0, 2.0, 13.0]  |\n",
       "+-------------------+"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Another example of using pl.concat_list.\n",
    "# This mimics a rolling window.\n",
    "start = pl.DataFrame({\"A\": [1.0, 2.0, 9.0, 2.0, 13.0]})\n",
    "print(start)\n",
    "\n",
    "rolling_window = (\n",
    "    start\n",
    "    .select([pl.col(\"A\").shift(i).alias(f\"A_lag_{i}\") for i in range(3)])\n",
    "    .select(pl.concat_list([f\"A_lag_{i}\" for i in range(3)][::-1]).alias(\"A_rolling\"))\n",
    ")\n",
    "rolling_window"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65c06468-5c09-47e2-b614-8fa1e6a8a9a2",
   "metadata": {},
   "source": [
    "## pl.col.list.sort()\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.list.sort.html\n",
    "* Sort the lists in this column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "742c00a7-f2c3-4ff0-a74f-309a329ae355",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>A_rolling</th></tr><tr><td>list[f64]</td></tr></thead><tbody><tr><td>[null, null, 1.0]</td></tr><tr><td>[null, 1.0, 2.0]</td></tr><tr><td>[1.0, 2.0, 9.0]</td></tr><tr><td>[2.0, 2.0, 9.0]</td></tr><tr><td>[2.0, 9.0, 13.0]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 1)\n",
       "+-------------------+\n",
       "| A_rolling         |\n",
       "| ---               |\n",
       "| list[f64]         |\n",
       "+===================+\n",
       "| [null, null, 1.0] |\n",
       "| [null, 1.0, 2.0]  |\n",
       "| [1.0, 2.0, 9.0]   |\n",
       "| [2.0, 2.0, 9.0]   |\n",
       "| [2.0, 9.0, 13.0]  |\n",
       "+-------------------+"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    rolling_window\n",
    "    .select(pl.col('A_rolling').list.sort(descending=False))\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bb2c81f-edca-421b-9100-b4e5c1f20aec",
   "metadata": {},
   "source": [
    "## pl.col().list.slice() \n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.list.slice.html\n",
    "* Slice every sublist."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "3169491c-08fa-43dd-bc5c-4b9f1aec650c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>A_rolling</th></tr><tr><td>list[f64]</td></tr></thead><tbody><tr><td>[null, 1.0]</td></tr><tr><td>[1.0, 2.0]</td></tr><tr><td>[2.0, 9.0]</td></tr><tr><td>[9.0, 2.0]</td></tr><tr><td>[2.0, 13.0]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 1)\n",
       "+-------------+\n",
       "| A_rolling   |\n",
       "| ---         |\n",
       "| list[f64]   |\n",
       "+=============+\n",
       "| [null, 1.0] |\n",
       "| [1.0, 2.0]  |\n",
       "| [2.0, 9.0]  |\n",
       "| [9.0, 2.0]  |\n",
       "| [2.0, 13.0] |\n",
       "+-------------+"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    rolling_window\n",
    "    .select(\n",
    "        pl.col('A_rolling').list.slice(1, 2) # slice 2nd and 3rd elements\n",
    "    ) \n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0649cfae-ecb2-4890-bb79-e877bba4e88c",
   "metadata": {},
   "source": [
    "## pl.col().list.sum()\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.list.sum.html#polars.Expr.list.sum\n",
    "* Sum all the lists in the array."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "bf025f75-0e0f-4c84-9a18-339324ef5b94",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>A_rolling</th></tr><tr><td>f64</td></tr></thead><tbody><tr><td>1.0</td></tr><tr><td>3.0</td></tr><tr><td>12.0</td></tr><tr><td>13.0</td></tr><tr><td>24.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 1)\n",
       "+-----------+\n",
       "| A_rolling |\n",
       "| ---       |\n",
       "| f64       |\n",
       "+===========+\n",
       "| 1.0       |\n",
       "| 3.0       |\n",
       "| 12.0      |\n",
       "| 13.0      |\n",
       "| 24.0      |\n",
       "+-----------+"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    rolling_window\n",
    "    .select(\n",
    "        pl.col('A_rolling').list.sum()\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1982d47a-770f-422d-b763-a60ff3373fc3",
   "metadata": {},
   "source": [
    "## df.sort()\n",
    "* https://docs.pola.rs/py-polars/html/reference/dataframe/api/polars.DataFrame.sort.html#polars.DataFrame.sort\n",
    "* Sort the dataframe by the given columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "231cb453-1806-41e0-8dc2-05d8e2d50f55",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1994</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Premium&quot;</td><td>13.523182</td><td>19</td><td>27</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1994</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;164&quot;</td><td>3.0</td><td>6</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>18.594376</td><td>14</td><td>20</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>1994</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;164&quot;</td><td>3.0</td><td>6</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>16.528334</td><td>15</td><td>22</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1995</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;164&quot;</td><td>3.0</td><td>6</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>18.594376</td><td>14</td><td>20</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>1999</td><td>&quot;Quantum Technologies&quot;</td><td>&quot;Chevrolet Cavalier&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Subcompact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>2003</td><td>&quot;London Taxi&quot;</td><td>&quot;London Taxi&quot;</td><td>2.4</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle&quot;</td><td>&quot;Diesel&quot;</td><td>16.236135</td><td>21</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>1987</td><td>&quot;Excalibur Autos&quot;</td><td>&quot;Phaeton&quot;</td><td>5.0</td><td>8</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Minicompact Cars&quot;</td><td>&quot;Premium&quot;</td><td>21.250713</td><td>12</td><td>17</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>2011</td><td>&quot;Mahindra&quot;</td><td>&quot;TR40&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 6-spd&quot;</td><td>&quot;4-Wheel Drive&quot;</td><td>&quot;Standard Pickup Trucks 4WD&quot;</td><td>&quot;Diesel&quot;</td><td>17.859751</td><td>19</td><td>21</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>6</td></tr><tr><td>2012</td><td>&quot;Azure Dynamics&quot;</td><td>&quot;Transit Connect Electric Van/W…</td><td>null</td><td>null</td><td>&quot;Automatic (A1)&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Special Purpose Vehicle 2WD&quot;</td><td>&quot;Electricity&quot;</td><td>0.1296</td><td>62</td><td>62</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25     | 2013- | false | 5     |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Pre | 13. | 19  | 27     | 2013- | false | 5     |\n",
       "| 4   | a   | der |     |     | ual | r-W | Sea | miu | 523 |     |        | 01-01 |       |       |\n",
       "|     | Rom |     |     |     | 5-s | hee | ter | m   | 182 |     |        | 00:00 |       |       |\n",
       "|     | eo  |     |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Alf | 164 | 3.0 | 6   | Aut | Fro | Com | Pre | 18. | 14  | 20     | 2013- | true  | 4     |\n",
       "| 4   | a   |     |     |     | oma | nt- | pac | miu | 594 |     |        | 01-01 |       |       |\n",
       "|     | Rom |     |     |     | tic | Whe | t   | m   | 376 |     |        | 00:00 |       |       |\n",
       "|     | eo  |     |     |     | 4-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Alf | 164 | 3.0 | 6   | Man | Fro | Com | Pre | 16. | 15  | 22     | 2013- | false | 5     |\n",
       "| 4   | a   |     |     |     | ual | nt- | pac | miu | 528 |     |        | 01-01 |       |       |\n",
       "|     | Rom |     |     |     | 5-s | Whe | t   | m   | 334 |     |        | 00:00 |       |       |\n",
       "|     | eo  |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Alf | 164 | 3.0 | 6   | Aut | Fro | Com | Pre | 18. | 14  | 20     | 2013- | true  | 4     |\n",
       "| 5   | a   |     |     |     | oma | nt- | pac | miu | 594 |     |        | 01-01 |       |       |\n",
       "|     | Rom |     |     |     | tic | Whe | t   | m   | 376 |     |        | 00:00 |       |       |\n",
       "|     | eo  |     |     |     | 4-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …      | …     | …     | …     |\n",
       "| 199 | Qua | Che | 2.2 | 4   | Aut | Fro | Sub | Reg | 14. | 19  | 25     | 2013- | true  | 3     |\n",
       "| 9   | ntu | vro |     |     | oma | nt- | com | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     | m   | let |     |     | tic | Whe | pac | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     | Tec | Cav |     |     | 3-s | el  | t   |     |     |     |        | :00   |       |       |\n",
       "|     | hno | ali |     |     | pd  | Dri | Car |     |     |     |        |       |       |       |\n",
       "|     | log | er  |     |     |     | ve  | s   |     |     |     |        |       |       |       |\n",
       "|     | ies |     |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "| 200 | Lon | Lon | 2.4 | 4   | Aut | Rea | Spe | Die | 16. | 21  | 24     | 2013- | true  | 4     |\n",
       "| 3   | don | don |     |     | oma | r-W | cia | sel | 236 |     |        | 01-01 |       |       |\n",
       "|     | Tax | Tax |     |     | tic | hee | l   |     | 135 |     |        | 00:00 |       |       |\n",
       "|     | i   | i   |     |     | 4-s | l   | Pur |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | pos |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  | e   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "| 198 | Exc | Pha | 5.0 | 8   | Aut | Rea | Min | Pre | 21. | 12  | 17     | 2013- | true  | 4     |\n",
       "| 7   | ali | eto |     |     | oma | r-W | ico | miu | 250 |     |        | 01-01 |       |       |\n",
       "|     | bur | n   |     |     | tic | hee | mpa | m   | 713 |     |        | 00:00 |       |       |\n",
       "|     | Aut |     |     |     | 4-s | l   | ct  |     |     |     |        | :00   |       |       |\n",
       "|     | os  |     |     |     | pd  | Dri | Car |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  | s   |     |     |     |        |       |       |       |\n",
       "| 201 | Mah | TR4 | 2.2 | 4   | Aut | 4-W | Sta | Die | 17. | 19  | 21     | 2013- | true  | 6     |\n",
       "| 1   | ind | 0   |     |     | oma | hee | nda | sel | 859 |     |        | 01-01 |       |       |\n",
       "|     | ra  |     |     |     | tic | l   | rd  |     | 751 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 6-s | Dri | Pic |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | ve  | kup |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | Tru |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | cks |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     |     | 4WD |     |     |     |        |       |       |       |\n",
       "| 201 | Azu | Tra | nul | nul | Aut | Fro | Spe | Ele | 0.1 | 62  | 62     | 2013- | true  | 1     |\n",
       "| 2   | re  | nsi | l   | l   | oma | nt- | cia | ctr | 296 |     |        | 01-01 |       |       |\n",
       "|     | Dyn | t   |     |     | tic | Whe | l   | ici |     |     |        | 00:00 |       |       |\n",
       "|     | ami | Con |     |     | (A1 | el  | Pur | ty  |     |     |        | :00   |       |       |\n",
       "|     | cs  | nec |     |     | )   | Dri | pos |     |     |     |        |       |       |       |\n",
       "|     |     | t   |     |     |     | ve  | e   |     |     |     |        |       |       |       |\n",
       "|     |     | Ele |     |     |     |     | Veh |     |     |     |        |       |       |       |\n",
       "|     |     | ctr |     |     |     |     | icl |     |     |     |        |       |       |       |\n",
       "|     |     | ic  |     |     |     |     | e   |     |     |     |        |       |       |       |\n",
       "|     |     | Van |     |     |     |     | 2WD |     |     |     |        |       |       |       |\n",
       "|     |     | /W… |     |     |     |     |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select(pl.all())\n",
    "    .sort(by='make')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b76d2aee-29b2-46ae-adde-ad4f0db0963d",
   "metadata": {},
   "source": [
    "## df.group_by()\n",
    "* https://docs.pola.rs/py-polars/html/reference/dataframe/api/polars.DataFrame.group_by.html#polars.DataFrame.group_by\n",
    "* Start a group by operation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "57424d58-37de-46d1-8624-5b85398f5d9f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "estimated unique values: 22060\n",
      "estimated unique count: 22060 exceeded the boundary: 1000, running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5_238, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>model</th><th>len</th></tr><tr><td>cat</td><td>cat</td><td>u32</td></tr></thead><tbody><tr><td>&quot;Ford&quot;</td><td>&quot;F150 Pickup 2WD&quot;</td><td>229</td></tr><tr><td>&quot;Chevrolet&quot;</td><td>&quot;Camaro&quot;</td><td>220</td></tr><tr><td>&quot;Ford&quot;</td><td>&quot;Mustang&quot;</td><td>215</td></tr><tr><td>&quot;Ford&quot;</td><td>&quot;F150 Pickup 4WD&quot;</td><td>211</td></tr><tr><td>&quot;Volkswagen&quot;</td><td>&quot;Jetta&quot;</td><td>203</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Polestar&quot;</td><td>&quot;3 Long Range Dual Motor (20 In…</td><td>1</td></tr><tr><td>&quot;Chrysler&quot;</td><td>&quot;TC by Maserati&quot;</td><td>1</td></tr><tr><td>&quot;Jaguar&quot;</td><td>&quot;F-Type R Coupe&quot;</td><td>1</td></tr><tr><td>&quot;GMC&quot;</td><td>&quot;Sierra 2500 HD 4WD CNG&quot;</td><td>1</td></tr><tr><td>&quot;Rivian&quot;</td><td>&quot;R1S Performance Dual Large (22…</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5_238, 3)\n",
       "+------------+---------------------------------+-----+\n",
       "| make       | model                           | len |\n",
       "| ---        | ---                             | --- |\n",
       "| cat        | cat                             | u32 |\n",
       "+====================================================+\n",
       "| Ford       | F150 Pickup 2WD                 | 229 |\n",
       "| Chevrolet  | Camaro                          | 220 |\n",
       "| Ford       | Mustang                         | 215 |\n",
       "| Ford       | F150 Pickup 4WD                 | 211 |\n",
       "| Volkswagen | Jetta                           | 203 |\n",
       "| …          | …                               | …   |\n",
       "| Polestar   | 3 Long Range Dual Motor (20 In… | 1   |\n",
       "| Chrysler   | TC by Maserati                  | 1   |\n",
       "| Jaguar     | F-Type R Coupe                  | 1   |\n",
       "| GMC        | Sierra 2500 HD 4WD CNG          | 1   |\n",
       "| Rivian     | R1S Performance Dual Large (22… | 1   |\n",
       "+------------+---------------------------------+-----+"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by(\"make\", \"model\")\n",
    "    .len()\n",
    "    .sort(by=\"len\", descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b97e7bf1-c1ec-45e6-96e6-cf350fd9d73e",
   "metadata": {},
   "source": [
    "## df.grouped.agg()\n",
    "* Essentially just using the aggregation function on the groupby object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "40537311-5f1c-4eed-a403-e1992b7cb2eb",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5_238, 6)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>model</th><th>len</th><th>trany</th><th>year</th><th>highway08</th></tr><tr><td>cat</td><td>cat</td><td>u32</td><td>list[str]</td><td>list[i16]</td><td>list[u8]</td></tr></thead><tbody><tr><td>&quot;Ford&quot;</td><td>&quot;F150 Pickup 2WD&quot;</td><td>229</td><td>[&quot;Automatic 4-spd&quot;, &quot;Manual 4-spd&quot;, … &quot;Automatic 4-spd&quot;]</td><td>[1993, 1993, … 1992]</td><td>[18, 17, … 15]</td></tr><tr><td>&quot;Chevrolet&quot;</td><td>&quot;Camaro&quot;</td><td>220</td><td>[&quot;Automatic 4-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Manual 6-spd&quot;]</td><td>[1994, 1994, … 1993]</td><td>[26, 26, … 23]</td></tr><tr><td>&quot;Ford&quot;</td><td>&quot;Mustang&quot;</td><td>215</td><td>[&quot;Automatic 4-spd&quot;, &quot;Manual 5-spd&quot;, … &quot;Automatic 4-spd&quot;]</td><td>[1994, 1994, … 1993]</td><td>[26, 27, … 22]</td></tr><tr><td>&quot;Ford&quot;</td><td>&quot;F150 Pickup 4WD&quot;</td><td>211</td><td>[&quot;Automatic 4-spd&quot;, &quot;Manual 4-spd&quot;, … &quot;Automatic 4-spd&quot;]</td><td>[1993, 1993, … 1992]</td><td>[17, 16, … 15]</td></tr><tr><td>&quot;Volkswagen&quot;</td><td>&quot;Jetta&quot;</td><td>203</td><td>[&quot;Manual 5-spd&quot;, &quot;Automatic 4-spd&quot;, … &quot;Manual 5-spd&quot;]</td><td>[1996, 1996, … 1992]</td><td>[44, 26, … 29]</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Mercedes-Benz&quot;</td><td>&quot;400SEL&quot;</td><td>1</td><td>[&quot;Automatic 4-spd&quot;]</td><td>[1993]</td><td>[17]</td></tr><tr><td>&quot;Chevrolet&quot;</td><td>&quot;Equinox EV AWD&quot;</td><td>1</td><td>[&quot;Automatic (A1)&quot;]</td><td>[2024]</td><td>[90]</td></tr><tr><td>&quot;Chevrolet&quot;</td><td>&quot;Express Cargo (Bi-fuel)&quot;</td><td>1</td><td>[&quot;Automatic 4-spd&quot;]</td><td>[2004]</td><td>[15]</td></tr><tr><td>&quot;Porsche&quot;</td><td>&quot;Taycan 4S Performance Battery …</td><td>1</td><td>[&quot;Automatic (A2)&quot;]</td><td>[2024]</td><td>[81]</td></tr><tr><td>&quot;Tesla&quot;</td><td>&quot;Model X AWD - 100D&quot;</td><td>1</td><td>[&quot;Automatic (A1)&quot;]</td><td>[2017]</td><td>[89]</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5_238, 6)\n",
       "+---------------+---------------------+-----+--------------------+----------------+----------------+\n",
       "| make          | model               | len | trany              | year           | highway08      |\n",
       "| ---           | ---                 | --- | ---                | ---            | ---            |\n",
       "| cat           | cat                 | u32 | list[str]          | list[i16]      | list[u8]       |\n",
       "+==================================================================================================+\n",
       "| Ford          | F150 Pickup 2WD     | 229 | [\"Automatic        | [1993, 1993, … | [18, 17, … 15] |\n",
       "|               |                     |     | 4-spd\", \"Manual    | 1992]          |                |\n",
       "|               |                     |     | 4-…                |                |                |\n",
       "| Chevrolet     | Camaro              | 220 | [\"Automatic        | [1994, 1994, … | [26, 26, … 23] |\n",
       "|               |                     |     | 4-spd\", \"Manual    | 1993]          |                |\n",
       "|               |                     |     | 5-…                |                |                |\n",
       "| Ford          | Mustang             | 215 | [\"Automatic        | [1994, 1994, … | [26, 27, … 22] |\n",
       "|               |                     |     | 4-spd\", \"Manual    | 1993]          |                |\n",
       "|               |                     |     | 5-…                |                |                |\n",
       "| Ford          | F150 Pickup 4WD     | 211 | [\"Automatic        | [1993, 1993, … | [17, 16, … 15] |\n",
       "|               |                     |     | 4-spd\", \"Manual    | 1992]          |                |\n",
       "|               |                     |     | 4-…                |                |                |\n",
       "| Volkswagen    | Jetta               | 203 | [\"Manual 5-spd\",   | [1996, 1996, … | [44, 26, … 29] |\n",
       "|               |                     |     | \"Automatic 4-…     | 1992]          |                |\n",
       "| …             | …                   | …   | …                  | …              | …              |\n",
       "| Mercedes-Benz | 400SEL              | 1   | [\"Automatic        | [1993]         | [17]           |\n",
       "|               |                     |     | 4-spd\"]            |                |                |\n",
       "| Chevrolet     | Equinox EV AWD      | 1   | [\"Automatic (A1)\"] | [2024]         | [90]           |\n",
       "| Chevrolet     | Express Cargo       | 1   | [\"Automatic        | [2004]         | [15]           |\n",
       "|               | (Bi-fuel)           |     | 4-spd\"]            |                |                |\n",
       "| Porsche       | Taycan 4S           | 1   | [\"Automatic (A2)\"] | [2024]         | [81]           |\n",
       "|               | Performance Battery |     |                    |                |                |\n",
       "|               | …                   |     |                    |                |                |\n",
       "| Tesla         | Model X AWD - 100D  | 1   | [\"Automatic (A1)\"] | [2017]         | [89]           |\n",
       "+---------------+---------------------+-----+--------------------+----------------+----------------+"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .group_by(\"make\", \"model\")\n",
    "    .agg(\n",
    "        pl.len(),\n",
    "        pl.col(\"trany\"),\n",
    "        pl.col(\"year\"),\n",
    "        pl.col(\"highway08\")\n",
    "    )\n",
    "    .sort(by=\"len\", descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "accff641-44e3-4aef-bef8-c85ae2b7adf8",
   "metadata": {},
   "source": [
    "## df.pivot()\n",
    "* https://docs.pola.rs/py-polars/html/reference/dataframe/api/polars.DataFrame.pivot.html\n",
    "* Create a spreadsheet-style pivot table as a DataFrame.\n",
    "* Only available in eager mode. If you know the unique column values in advance, you can do a \"lazy pivot\". "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "37e49c97-ff31-4075-adee-f9f9b516697f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5_238, 5)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>model</th><th>Automatic</th><th>false</th><th>true</th></tr><tr><td>cat</td><td>cat</td><td>f64</td><td>f64</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>null</td><td>18.333333</td><td>null</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>null</td><td>9.111111</td><td>null</td></tr><tr><td>&quot;Dodge&quot;</td><td>&quot;Charger&quot;</td><td>null</td><td>21.555556</td><td>17.146667</td></tr><tr><td>&quot;Dodge&quot;</td><td>&quot;B150/B250 Wagon 2WD&quot;</td><td>null</td><td>11.894737</td><td>11.580645</td></tr><tr><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>null</td><td>17.0</td><td>16.0</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Dodge&quot;</td><td>&quot;Caravan C/V/Grand Caravan 4WD&quot;</td><td>null</td><td>null</td><td>15.0</td></tr><tr><td>&quot;PAS Inc - GMC&quot;</td><td>&quot;Pas-typhoon&quot;</td><td>null</td><td>null</td><td>14.0</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Ferrari 348 TB/TS&quot;</td><td>null</td><td>12.0</td><td>null</td></tr><tr><td>&quot;Jaguar&quot;</td><td>&quot;XJRS Convertble&quot;</td><td>null</td><td>null</td><td>11.0</td></tr><tr><td>&quot;Jaguar&quot;</td><td>&quot;XJRS Coupe&quot;</td><td>null</td><td>null</td><td>11.0</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5_238, 5)\n",
       "+---------------+-------------------------------+-----------+-----------+-----------+\n",
       "| make          | model                         | Automatic | false     | true      |\n",
       "| ---           | ---                           | ---       | ---       | ---       |\n",
       "| cat           | cat                           | f64       | f64       | f64       |\n",
       "+===================================================================================+\n",
       "| Alfa Romeo    | Spider Veloce 2000            | null      | 18.333333 | null      |\n",
       "| Ferrari       | Testarossa                    | null      | 9.111111  | null      |\n",
       "| Dodge         | Charger                       | null      | 21.555556 | 17.146667 |\n",
       "| Dodge         | B150/B250 Wagon 2WD           | null      | 11.894737 | 11.580645 |\n",
       "| Subaru        | Legacy AWD Turbo              | null      | 17.0      | 16.0      |\n",
       "| …             | …                             | …         | …         | …         |\n",
       "| Dodge         | Caravan C/V/Grand Caravan 4WD | null      | null      | 15.0      |\n",
       "| PAS Inc - GMC | Pas-typhoon                   | null      | null      | 14.0      |\n",
       "| Ferrari       | Ferrari 348 TB/TS             | null      | 12.0      | null      |\n",
       "| Jaguar        | XJRS Convertble               | null      | null      | 11.0      |\n",
       "| Jaguar        | XJRS Coupe                    | null      | null      | 11.0      |\n",
       "+---------------+-------------------------------+-----------+-----------+-----------+"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .pivot(\n",
    "        index=['make', 'model'], # One or multiple keys to group by.\n",
    "        on='is_automatic', # Name of the column(s) whose values will be used as the header of the output DataFrame.\n",
    "        values='city08', # Column values to aggregate. If None, all remaining columns will be used.\n",
    "        aggregate_function=\"mean\",\n",
    "        sort_columns=True,\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fc562bb0-e225-4e34-979c-34f8d129f5d0",
   "metadata": {},
   "source": [
    "## value counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "dbd1ff11-a610-4566-9775-bd813d869361",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---- year ---\n",
      "shape: (42, 2)\n",
      "+------+-------+\n",
      "| year | count |\n",
      "| ---  | ---   |\n",
      "| i16  | u32   |\n",
      "+==============+\n",
      "| 1984 | 1964  |\n",
      "| 1985 | 1701  |\n",
      "| 2018 | 1350  |\n",
      "| 2019 | 1347  |\n",
      "| 2023 | 1330  |\n",
      "| …    | …     |\n",
      "| 2000 | 840   |\n",
      "| 1998 | 812   |\n",
      "| 2025 | 800   |\n",
      "| 1996 | 773   |\n",
      "| 1997 | 762   |\n",
      "+------+-------+\n",
      "---- make ---\n",
      "shape: (145, 2)\n",
      "+---------------------------------+-------+\n",
      "| make                            | count |\n",
      "| ---                             | ---   |\n",
      "| cat                             | u32   |\n",
      "+=========================================+\n",
      "| Chevrolet                       | 4472  |\n",
      "| Ford                            | 3834  |\n",
      "| GMC                             | 2823  |\n",
      "| Dodge                           | 2695  |\n",
      "| BMW                             | 2536  |\n",
      "| …                               | …     |\n",
      "| Grumman Allied Industries       | 1     |\n",
      "| Environmental Rsch and Devp Co… | 1     |\n",
      "| General Motors                  | 1     |\n",
      "| Goldacre                        | 1     |\n",
      "| Isis Imports Ltd                | 1     |\n",
      "+---------------------------------+-------+\n",
      "---- model ---\n",
      "shape: (5_166, 2)\n",
      "+-------------------------------+-------+\n",
      "| model                         | count |\n",
      "| ---                           | ---   |\n",
      "| cat                           | u32   |\n",
      "+=======================================+\n",
      "| F150 Pickup 2WD               | 236   |\n",
      "| Camaro                        | 220   |\n",
      "| F150 Pickup 4WD               | 218   |\n",
      "| Mustang                       | 218   |\n",
      "| Jetta                         | 203   |\n",
      "| …                             | …     |\n",
      "| Caravan C/V/Grand Caravan 4WD | 1     |\n",
      "| Pas-typhoon                   | 1     |\n",
      "| Ferrari 348 TB/TS             | 1     |\n",
      "| XJRS Convertble               | 1     |\n",
      "| XJRS Coupe                    | 1     |\n",
      "+-------------------------------+-------+\n",
      "---- displ ---\n",
      "shape: (67, 2)\n",
      "+-------+-------+\n",
      "| displ | count |\n",
      "| ---   | ---   |\n",
      "| f32   | u32   |\n",
      "+===============+\n",
      "| 2.0   | 5809  |\n",
      "| 3.0   | 4313  |\n",
      "| 2.5   | 3050  |\n",
      "| 2.4   | 2162  |\n",
      "| 3.5   | 2077  |\n",
      "| …     | …     |\n",
      "| 7.0   | 10    |\n",
      "| 8.3   | 9     |\n",
      "| 0.9   | 6     |\n",
      "| 7.4   | 4     |\n",
      "| 0.0   | 1     |\n",
      "+-------+-------+\n",
      "---- cylinders ---\n",
      "shape: (10, 2)\n",
      "+-----------+-------+\n",
      "| cylinders | count |\n",
      "| ---       | ---   |\n",
      "| u8        | u32   |\n",
      "+===================+\n",
      "| 4         | 18727 |\n",
      "| 6         | 16239 |\n",
      "| 8         | 10040 |\n",
      "| null      | 965   |\n",
      "| 5         | 778   |\n",
      "| 12        | 741   |\n",
      "| 3         | 428   |\n",
      "| 10        | 199   |\n",
      "| 2         | 63    |\n",
      "| 16        | 22    |\n",
      "+-----------+-------+\n",
      "---- trany ---\n",
      "shape: (41, 2)\n",
      "+--------------------+-------+\n",
      "| trany              | count |\n",
      "| ---                | ---   |\n",
      "| str                | u32   |\n",
      "+============================+\n",
      "| Automatic 4-spd    | 11048 |\n",
      "| Manual 5-spd       | 8392  |\n",
      "| Automatic (S6)     | 3352  |\n",
      "| Automatic (S8)     | 3292  |\n",
      "| Automatic 3-spd    | 3151  |\n",
      "| …                  | …     |\n",
      "| null               | 11    |\n",
      "| Automatic (AM-S9)  | 6     |\n",
      "| Automatic (L4)     | 2     |\n",
      "| Automatic (L3)     | 2     |\n",
      "| Automatic (AM-S10) | 2     |\n",
      "+--------------------+-------+\n",
      "---- drive ---\n",
      "shape: (8, 2)\n",
      "+----------------------------+-------+\n",
      "| drive                      | count |\n",
      "| ---                        | ---   |\n",
      "| cat                        | u32   |\n",
      "+====================================+\n",
      "| Front-Wheel Drive          | 15599 |\n",
      "| Rear-Wheel Drive           | 15452 |\n",
      "| 4-Wheel or All-Wheel Drive | 6645  |\n",
      "| All-Wheel Drive            | 5772  |\n",
      "| 4-Wheel Drive              | 2428  |\n",
      "| null                       | 1186  |\n",
      "| Part-time 4-Wheel Drive    | 613   |\n",
      "| 2-Wheel Drive              | 507   |\n",
      "+----------------------------+-------+\n",
      "---- VClass ---\n",
      "shape: (34, 2)\n",
      "+------------------------------+-------+\n",
      "| VClass                       | count |\n",
      "| ---                          | ---   |\n",
      "| cat                          | u32   |\n",
      "+======================================+\n",
      "| Compact Cars                 | 6474  |\n",
      "| Midsize Cars                 | 5750  |\n",
      "| Subcompact Cars              | 5717  |\n",
      "| Large Cars                   | 2680  |\n",
      "| Two Seaters                  | 2450  |\n",
      "| …                            | …     |\n",
      "| Standard Pickup Trucks/2wd   | 4     |\n",
      "| Vans Passenger               | 2     |\n",
      "| Special Purpose Vehicles/2wd | 2     |\n",
      "| Special Purpose Vehicles/4wd | 2     |\n",
      "| Special Purpose Vehicle      | 1     |\n",
      "+------------------------------+-------+\n",
      "---- fuelType ---\n",
      "shape: (15, 2)\n",
      "+----------------------------+-------+\n",
      "| fuelType                   | count |\n",
      "| ---                        | ---   |\n",
      "| cat                        | u32   |\n",
      "+====================================+\n",
      "| Regular                    | 29131 |\n",
      "| Premium                    | 14699 |\n",
      "| Gasoline or E85            | 1402  |\n",
      "| Diesel                     | 1284  |\n",
      "| Electricity                | 930   |\n",
      "| …                          | …     |\n",
      "| Premium Gas or Electricity | 55    |\n",
      "| Hydrogen                   | 32    |\n",
      "| Gasoline or natural gas    | 20    |\n",
      "| Gasoline or propane        | 8     |\n",
      "| Regular Gas or Electricity | 4     |\n",
      "+----------------------------+-------+\n",
      "---- barrels08 ---\n",
      "shape: (323, 2)\n",
      "+-----------+-------+\n",
      "| barrels08 | count |\n",
      "| ---       | ---   |\n",
      "| f32       | u32   |\n",
      "+===================+\n",
      "| 16.528334 | 4202  |\n",
      "| 15.658422 | 4049  |\n",
      "| 14.167143 | 3998  |\n",
      "| 14.8755   | 3842  |\n",
      "| 17.500587 | 2983  |\n",
      "| …         | …     |\n",
      "| 1.324673  | 1     |\n",
      "| 1.780144  | 1     |\n",
      "| 1.7376    | 1     |\n",
      "| 5.833529  | 1     |\n",
      "| 0.0552    | 1     |\n",
      "+-----------+-------+\n",
      "---- city08 ---\n",
      "shape: (139, 2)\n",
      "+--------+-------+\n",
      "| city08 | count |\n",
      "| ---    | ---   |\n",
      "| u8     | u32   |\n",
      "+================+\n",
      "| 15     | 5003  |\n",
      "| 18     | 4518  |\n",
      "| 17     | 4470  |\n",
      "| 16     | 4418  |\n",
      "| 19     | 3527  |\n",
      "| …      | …     |\n",
      "| 148    | 1     |\n",
      "| 135    | 1     |\n",
      "| 144    | 1     |\n",
      "| 149    | 1     |\n",
      "| 70     | 1     |\n",
      "+--------+-------+\n",
      "---- highway08 ---\n",
      "shape: (125, 2)\n",
      "+-----------+-------+\n",
      "| highway08 | count |\n",
      "| ---       | ---   |\n",
      "| u8        | u32   |\n",
      "+===================+\n",
      "| 24        | 3854  |\n",
      "| 26        | 3767  |\n",
      "| 22        | 3508  |\n",
      "| 23        | 3196  |\n",
      "| 20        | 2922  |\n",
      "| …         | …     |\n",
      "| 131       | 1     |\n",
      "| 142       | 1     |\n",
      "| 130       | 1     |\n",
      "| 113       | 1     |\n",
      "| 70        | 1     |\n",
      "+-----------+-------+\n",
      "---- createdOn ---\n",
      "shape: (459, 2)\n",
      "+---------------------+-------+\n",
      "| createdOn           | count |\n",
      "| ---                 | ---   |\n",
      "| datetime[μs]        | u32   |\n",
      "+=============================+\n",
      "| 2013-01-01 00:00:00 | 34193 |\n",
      "| 2024-08-25 00:00:00 | 329   |\n",
      "| 2023-07-31 00:00:00 | 247   |\n",
      "| 2023-11-06 00:00:00 | 235   |\n",
      "| 2023-09-19 00:00:00 | 206   |\n",
      "| …                   | …     |\n",
      "| 2019-04-10 00:00:00 | 1     |\n",
      "| 2020-02-19 00:00:00 | 1     |\n",
      "| 2020-07-16 00:00:00 | 1     |\n",
      "| 2020-12-17 00:00:00 | 1     |\n",
      "| 2022-04-18 00:00:00 | 1     |\n",
      "+---------------------+-------+\n",
      "---- is_automatic ---\n",
      "shape: (3, 2)\n",
      "+--------------+-------+\n",
      "| is_automatic | count |\n",
      "| ---          | ---   |\n",
      "| str          | u32   |\n",
      "+======================+\n",
      "| true         | 34950 |\n",
      "| false        | 13241 |\n",
      "| Automatic    | 11    |\n",
      "+--------------+-------+\n",
      "---- num_gears ---\n",
      "shape: (10, 2)\n",
      "+-----------+-------+\n",
      "| num_gears | count |\n",
      "| ---       | ---   |\n",
      "| u8        | u32   |\n",
      "+===================+\n",
      "| 4         | 12783 |\n",
      "| 5         | 11457 |\n",
      "| 6         | 10037 |\n",
      "| 8         | 4896  |\n",
      "| 3         | 3230  |\n",
      "| 7         | 2609  |\n",
      "| 10        | 1119  |\n",
      "| 9         | 1079  |\n",
      "| 1         | 925   |\n",
      "| 2         | 67    |\n",
      "+-----------+-------+\n"
     ]
    }
   ],
   "source": [
    "def all_value_counts(df_):\n",
    "    for x in df_.columns:\n",
    "        print(f\"---- {x} ---\")\n",
    "        print(autos[x].value_counts(sort=True))\n",
    "\n",
    "all_value_counts(autos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "82aed2ad-1867-4513-b0e1-f78439728db9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year              42\n",
       "make             145\n",
       "model           5166\n",
       "displ             66\n",
       "cylinders          9\n",
       "trany             40\n",
       "drive              7\n",
       "VClass            34\n",
       "fuelType          15\n",
       "barrels08        323\n",
       "city08           139\n",
       "highway08        125\n",
       "createdOn        459\n",
       "is_automatic       3\n",
       "num_gears         10\n",
       "dtype: int64"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autos.to_pandas().nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "5371b214-4beb-4ed7-997e-8dad424cd14e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "year          : 42\n",
      "make          : 145\n",
      "model         : 5166\n",
      "displ         : 67\n",
      "cylinders     : 10\n",
      "trany         : 41\n",
      "drive         : 8\n",
      "VClass        : 34\n",
      "fuelType      : 15\n",
      "barrels08     : 323\n",
      "city08        : 139\n",
      "highway08     : 125\n",
      "createdOn     : 459\n",
      "is_automatic  : 3\n",
      "num_gears     : 10\n"
     ]
    }
   ],
   "source": [
    "def all_unique_counts(df_):\n",
    "    for x in df_.columns:\n",
    "        print(f\"{x: <14}: {autos[x].n_unique()}\")\n",
    "        \n",
    "all_unique_counts(autos)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1de2309d-afdb-43bd-b40e-025b49ca10e3",
   "metadata": {},
   "source": [
    "## df.filter\n",
    "\n",
    "    DataFrame.filter(\n",
    "        *predicates: IntoExprColumn | Iterable[IntoExprColumn] | bool | list[bool] | np.ndarray[Any, Any],\n",
    "        **constraints: Any,\n",
    "    ) → DataFrame\n",
    "\n",
    "* https://docs.pola.rs/py-polars/html/reference/dataframe/api/polars.DataFrame.filter.html\n",
    "* Filter the rows in the DataFrame based on one or more predicate expressions.\n",
    "* The original order of the remaining rows is preserved."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "aafc0e7f-1dca-4698-9c01-764b36dd2c07",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (2, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1993</td><td>&quot;Toyota&quot;</td><td>&quot;Corolla&quot;</td><td>1.6</td><td>4</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>12.39625</td><td>23</td><td>26</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1993</td><td>&quot;Toyota&quot;</td><td>&quot;Corolla&quot;</td><td>1.6</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.442693</td><td>23</td><td>31</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (2, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 199 | Toy | Cor | 1.6 | 4   | Aut | Fro | Com | Reg | 12. | 23  | 26     | 2013- | true  | 3     |\n",
       "| 3   | ota | oll |     |     | oma | nt- | pac | ula | 396 |     |        | 01-01 |       |       |\n",
       "|     |     | a   |     |     | tic | Whe | t   | r   | 25  |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 3-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Toy | Cor | 1.6 | 4   | Man | Fro | Com | Reg | 11. | 23  | 31     | 2013- | false | 5     |\n",
       "| 3   | ota | oll |     |     | ual | nt- | pac | ula | 442 |     |        | 01-01 |       |       |\n",
       "|     |     | a   |     |     | 5-s | Whe | t   | r   | 693 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .filter(pl.col('make') == \"Toyota\" )\n",
    ").head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "917015f7-fc31-4d3a-a535-7f1891486106",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n",
      "known unique values: 145\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (11, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>len</th></tr><tr><td>cat</td><td>u32</td></tr></thead><tbody><tr><td>&quot;Toyota&quot;</td><td>2470</td></tr><tr><td>&quot;Nissan&quot;</td><td>1686</td></tr><tr><td>&quot;Honda&quot;</td><td>1197</td></tr><tr><td>&quot;Mitsubishi&quot;</td><td>1131</td></tr><tr><td>&quot;Mazda&quot;</td><td>1129</td></tr><tr><td>&quot;Subaru&quot;</td><td>1013</td></tr><tr><td>&quot;Lexus&quot;</td><td>698</td></tr><tr><td>&quot;Suzuki&quot;</td><td>515</td></tr><tr><td>&quot;Infiniti&quot;</td><td>484</td></tr><tr><td>&quot;Acura&quot;</td><td>424</td></tr><tr><td>&quot;Daihatsu&quot;</td><td>17</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (11, 2)\n",
       "+------------+------+\n",
       "| make       | len  |\n",
       "| ---        | ---  |\n",
       "| cat        | u32  |\n",
       "+===================+\n",
       "| Toyota     | 2470 |\n",
       "| Nissan     | 1686 |\n",
       "| Honda      | 1197 |\n",
       "| Mitsubishi | 1131 |\n",
       "| Mazda      | 1129 |\n",
       "| Subaru     | 1013 |\n",
       "| Lexus      | 698  |\n",
       "| Suzuki     | 515  |\n",
       "| Infiniti   | 484  |\n",
       "| Acura      | 424  |\n",
       "| Daihatsu   | 17   |\n",
       "+------------+------+"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl.Config.set_tbl_rows(20)\n",
    "\n",
    "ja_cars = sorted([\n",
    "    'Acura', 'Aspark', 'Daihatsu', 'Hino', 'Honda', \n",
    "    'Infiniti', 'Izusu', 'Kawasaki', 'Lexus', 'Mazda', \n",
    "    'Mitsubishi', 'Nissan', 'Subaru', 'Suzuki', 'Toyota'\n",
    "])\n",
    "\n",
    "ja_cars_count = (\n",
    "    autos\n",
    "    .filter(pl.col('make').is_in(ja_cars))\n",
    "    .group_by('make')\n",
    "    .len()\n",
    "    .sort(by='len', descending=True)\n",
    ")\n",
    "ja_cars_count"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47b28f0f-31db-4918-820f-93a17ac5e8cc",
   "metadata": {},
   "source": [
    "## col.cat.set_ordering()\n",
    "\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.cat.set_ordering.html#polars-expr-cat-set-ordering\n",
    "* Determine how this categorical series should be sorted.\n",
    "\n",
    "**DeprecationWarning: `set_ordering` is deprecated. Set the ordering directly on the datatype `pl.Categorical('lexical')` or `pl.Categorical('physical')` or `cast()` to the intended data type. This method will be removed in the next breaking change**\n",
    "\n",
    "## pl.Categorical('lexical'|'physical')\n",
    "\n",
    "* https://docs.pola.rs/py-polars/html/reference/api/polars.datatypes.Categorical.html\n",
    "* A categorical encoding of a set of strings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "9039948d-d259-4c11-8370-39b479f299b0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "estimated unique values: 7306\n",
      "estimated unique count: 7306 exceeded the boundary: 1000, running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (698, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>trany</th><th>make</th><th>len</th></tr><tr><td>cat</td><td>cat</td><td>u32</td></tr></thead><tbody><tr><td>null</td><td>&quot;Toyota&quot;</td><td>4</td></tr><tr><td>null</td><td>&quot;Ford&quot;</td><td>4</td></tr><tr><td>null</td><td>&quot;Nissan&quot;</td><td>2</td></tr><tr><td>null</td><td>&quot;GMC&quot;</td><td>1</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>&quot;Bugatti Rimac&quot;</td><td>2</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>&quot;Audi&quot;</td><td>36</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>&quot;Kia&quot;</td><td>34</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>&quot;Hyundai&quot;</td><td>50</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>&quot;Volvo&quot;</td><td>15</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>&quot;Subaru&quot;</td><td>4</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Manual 6-spd&quot;</td><td>&quot;Hyundai&quot;</td><td>99</td></tr><tr><td>&quot;Manual 6-spd&quot;</td><td>&quot;Toyota&quot;</td><td>109</td></tr><tr><td>&quot;Manual 6-spd&quot;</td><td>&quot;SRT&quot;</td><td>2</td></tr><tr><td>&quot;Manual 6-spd&quot;</td><td>&quot;Genesis&quot;</td><td>3</td></tr><tr><td>&quot;Manual 7-spd&quot;</td><td>&quot;Ford&quot;</td><td>15</td></tr><tr><td>&quot;Manual 7-spd&quot;</td><td>&quot;Porsche&quot;</td><td>140</td></tr><tr><td>&quot;Manual 7-spd&quot;</td><td>&quot;RUF Automobile&quot;</td><td>4</td></tr><tr><td>&quot;Manual 7-spd&quot;</td><td>&quot;Lamborghini&quot;</td><td>2</td></tr><tr><td>&quot;Manual 7-spd&quot;</td><td>&quot;Chevrolet&quot;</td><td>12</td></tr><tr><td>&quot;Manual 7-spd&quot;</td><td>&quot;Aston Martin&quot;</td><td>3</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (698, 3)\n",
       "+----------------+----------------+-----+\n",
       "| trany          | make           | len |\n",
       "| ---            | ---            | --- |\n",
       "| cat            | cat            | u32 |\n",
       "+=======================================+\n",
       "| null           | Toyota         | 4   |\n",
       "| null           | Ford           | 4   |\n",
       "| null           | Nissan         | 2   |\n",
       "| null           | GMC            | 1   |\n",
       "| Automatic (A1) | Bugatti Rimac  | 2   |\n",
       "| Automatic (A1) | Audi           | 36  |\n",
       "| Automatic (A1) | Kia            | 34  |\n",
       "| Automatic (A1) | Hyundai        | 50  |\n",
       "| Automatic (A1) | Volvo          | 15  |\n",
       "| Automatic (A1) | Subaru         | 4   |\n",
       "| …              | …              | …   |\n",
       "| Manual 6-spd   | Hyundai        | 99  |\n",
       "| Manual 6-spd   | Toyota         | 109 |\n",
       "| Manual 6-spd   | SRT            | 2   |\n",
       "| Manual 6-spd   | Genesis        | 3   |\n",
       "| Manual 7-spd   | Ford           | 15  |\n",
       "| Manual 7-spd   | Porsche        | 140 |\n",
       "| Manual 7-spd   | RUF Automobile | 4   |\n",
       "| Manual 7-spd   | Lamborghini    | 2   |\n",
       "| Manual 7-spd   | Chevrolet      | 12  |\n",
       "| Manual 7-spd   | Aston Martin   | 3   |\n",
       "+----------------+----------------+-----+"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        pl.col('trany').cast(pl.Categorical('lexical')),\n",
    "        pl.col('make')\n",
    "    )\n",
    "    .group_by('trany', 'make')\n",
    "    .len()\n",
    "    .sort(by='trany')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c7ad52c-3f23-4e16-8ef9-142c120045e6",
   "metadata": {},
   "source": [
    "## col.over() (can be tricky at first)\n",
    "\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.over.html\n",
    "* Create a window expression.\n",
    "* Group over the specified expressions.\n",
    "* The mapping_strategy parameter can be 'group_to_rows', 'join', or 'explode'.\n",
    "* Generally, an aggregation function is used before the .over method.\n",
    "* Compute expressions over the given groups.\n",
    "* This expression is similar to performing a group by aggregation and joining the result back into the original DataFrame.\n",
    "* The outcome is similar to how window functions work in PostgreSQL.\n",
    "\n",
    "---\n",
    "\n",
    "## Explain the use of .over in Polars\n",
    "\n",
    "The `.over()` method in Polars is used to perform window functions or aggregations over groups within a DataFrame. It allows you to compute expressions or aggregations on subsets of the data, grouped by one or more columns, without modifying the original size of the DataFrame.\n",
    "\n",
    "Window functions are powerful because they enable you to perform multiple group-by operations and aggregations in a single query, projecting the results back to the original rows. This eliminates the need for separate `group_by` and `explode` operations, making the API cleaner and more efficient.\n",
    "\n",
    "Here are some key points about using `.over()` in Polars:\n",
    "\n",
    "1. **Grouping columns**: The columns to group by are specified as arguments to `.over()`. For example, `pl.col(\"value\").mean().over(\"group\")` computes the mean of \"value\" for each group defined by the \"group\" column.\n",
    "\n",
    "2. **Multiple grouping columns**: You can group by multiple columns by passing them as a list or tuple to `.over()`, e.g., `.over([\"group1\", \"group2\"])`. This allows you to compute aggregations over combinations of groups.\n",
    "\n",
    "3. **Mapping strategies**: When an expression results in multiple values per group, Polars provides three mapping strategies to link the values back to the DataFrame rows: 'group_to_rows', 'join', and 'explode'. The appropriate strategy depends on your use case.\n",
    "\n",
    "4. **Parallel computation**: Window expressions are computed in parallel, making them efficient for large datasets.\n",
    "\n",
    "5. **Caching groups**: Polars caches and shares computed groups between different window expressions, optimizing performance when multiple aggregations are performed over the same groups.\n",
    "\n",
    "6. **Nested over calls**: While nested `.over()` calls are generally not allowed in Polars, there are workarounds to achieve similar functionality by breaking down the expressions into separate steps.\n",
    "\n",
    "By using `.over()` and window functions, you can perform complex data manipulations and aggregations in a concise and efficient manner within Polars, without modifying the original DataFrame size unless explicitly desired."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb9cff75-104e-48b3-9635-69f1ca44912a",
   "metadata": {},
   "source": [
    "The best examples to work through for .over() are combinations found in model/features/characteristics examples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "b3df77d5-33a5-4cd5-8e8e-b747511065f5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['depname', 'empno', 'salary']\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (10, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>depname</th><th>empno</th><th>salary</th><th>avg</th></tr><tr><td>str</td><td>i64</td><td>i64</td><td>f64</td></tr></thead><tbody><tr><td>&quot;develop&quot;</td><td>11</td><td>5200</td><td>5020.0</td></tr><tr><td>&quot;develop&quot;</td><td>7</td><td>4200</td><td>5020.0</td></tr><tr><td>&quot;develop&quot;</td><td>9</td><td>4500</td><td>5020.0</td></tr><tr><td>&quot;develop&quot;</td><td>8</td><td>6000</td><td>5020.0</td></tr><tr><td>&quot;develop&quot;</td><td>10</td><td>5200</td><td>5020.0</td></tr><tr><td>&quot;personnel&quot;</td><td>5</td><td>3500</td><td>3700.0</td></tr><tr><td>&quot;personnel&quot;</td><td>2</td><td>3900</td><td>3700.0</td></tr><tr><td>&quot;sales&quot;</td><td>3</td><td>4800</td><td>4866.67</td></tr><tr><td>&quot;sales&quot;</td><td>1</td><td>5000</td><td>4866.67</td></tr><tr><td>&quot;sales&quot;</td><td>4</td><td>4800</td><td>4866.67</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (10, 4)\n",
       "+-----------+-------+--------+---------+\n",
       "| depname   | empno | salary | avg     |\n",
       "| ---       | ---   | ---    | ---     |\n",
       "| str       | i64   | i64    | f64     |\n",
       "+======================================+\n",
       "| develop   | 11    | 5200   | 5020.0  |\n",
       "| develop   | 7     | 4200   | 5020.0  |\n",
       "| develop   | 9     | 4500   | 5020.0  |\n",
       "| develop   | 8     | 6000   | 5020.0  |\n",
       "| develop   | 10    | 5200   | 5020.0  |\n",
       "| personnel | 5     | 3500   | 3700.0  |\n",
       "| personnel | 2     | 3900   | 3700.0  |\n",
       "| sales     | 3     | 4800   | 4866.67 |\n",
       "| sales     | 1     | 5000   | 4866.67 |\n",
       "| sales     | 4     | 4800   | 4866.67 |\n",
       "+-----------+-------+--------+---------+"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This is like applying .mean on a groupby object \n",
    "# eg .mean on develop-groupby object (5020.0)\n",
    "# eg .mean on personnel-groupby object (3700.0)\n",
    "# eg .mean on sales-groupby object (4866.67)\n",
    "# df_windows = pl.read_csv('./data/data_psql_windows.csv')\n",
    "\n",
    "df_windows = pl.DataFrame({\n",
    "    \"depname\": [\"develop\",\"develop\",\"develop\",\"develop\",\"develop\",\"personnel\",\"personnel\",\"sales\",\"sales\",\"sales\"],\n",
    "    \"empno\": [11, 7, 9, 8, 10, 5, 2, 3, 1, 4],\n",
    "    \"salary\": [5200,4200,4500,6000,5200,3500,3900,4800,5000,4800]\n",
    "})\n",
    "\n",
    "print(df_windows.columns)\n",
    "\n",
    "(\n",
    "    df_windows\n",
    "    .with_columns(avg=pl.col('salary').mean().over('depname').round(2))\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "0a89c0e5-9a98-4e6c-aa04-c26e838cf35c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: (5, 3)\n",
      "+-----+-----+-----+\n",
      "| a   | b   | c   |\n",
      "| --- | --- | --- |\n",
      "| i64 | i64 | i64 |\n",
      "+=================+\n",
      "| 9   | 1   | 5   |\n",
      "| 8   | 2   | 4   |\n",
      "| 7   | 3   | 3   |\n",
      "| 6   | 5   | 2   |\n",
      "| 5   | 3   | 1   |\n",
      "+-----+-----+-----+\n",
      "----------------------------------------\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>a</th><th>b</th><th>c</th><th>c_max</th></tr><tr><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>9</td><td>1</td><td>5</td><td>5</td></tr><tr><td>8</td><td>2</td><td>4</td><td>4</td></tr><tr><td>7</td><td>3</td><td>3</td><td>3</td></tr><tr><td>6</td><td>5</td><td>2</td><td>2</td></tr><tr><td>5</td><td>3</td><td>1</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 4)\n",
       "+-----+-----+-----+-------+\n",
       "| a   | b   | c   | c_max |\n",
       "| --- | --- | --- | ---   |\n",
       "| i64 | i64 | i64 | i64   |\n",
       "+=========================+\n",
       "| 9   | 1   | 5   | 5     |\n",
       "| 8   | 2   | 4   | 4     |\n",
       "| 7   | 3   | 3   | 3     |\n",
       "| 6   | 5   | 2   | 2     |\n",
       "| 5   | 3   | 1   | 1     |\n",
       "+-----+-----+-----+-------+"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Pass the name of a column to compute the expression over that column.\n",
    "df = pl.DataFrame({\n",
    "    \"a\": [9, 8, 7, 6, 5],\n",
    "    \"b\": [1, 2, 3, 5, 3],\n",
    "    \"c\": [5, 4, 3, 2, 1],\n",
    "})\n",
    "print(df)\n",
    "HR()\n",
    "\n",
    "df.with_columns(\n",
    "    pl.col(\"c\")\n",
    "    .max()\n",
    "    .over(\"a\", mapping_strategy='group_to_rows')\n",
    "    .name.suffix(\"_max\"),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "d312e736-2ae9-4339-a434-3dee9df5b656",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>a</th><th>b</th><th>c</th><th>c_max</th></tr><tr><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>9</td><td>1</td><td>5</td><td>5</td></tr><tr><td>8</td><td>2</td><td>4</td><td>4</td></tr><tr><td>7</td><td>3</td><td>3</td><td>3</td></tr><tr><td>6</td><td>5</td><td>2</td><td>2</td></tr><tr><td>5</td><td>3</td><td>1</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 4)\n",
       "+-----+-----+-----+-------+\n",
       "| a   | b   | c   | c_max |\n",
       "| --- | --- | --- | ---   |\n",
       "| i64 | i64 | i64 | i64   |\n",
       "+=========================+\n",
       "| 9   | 1   | 5   | 5     |\n",
       "| 8   | 2   | 4   | 4     |\n",
       "| 7   | 3   | 3   | 3     |\n",
       "| 6   | 5   | 2   | 2     |\n",
       "| 5   | 3   | 1   | 1     |\n",
       "+-----+-----+-----+-------+"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.with_columns(\n",
    "    pl.col(\"c\")\n",
    "    .max()\n",
    "    .over(\"a\", mapping_strategy='join')\n",
    "    .name.suffix(\"_max\"),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "dfdc36f2-9b76-4123-bb9e-2024514f29c7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (5, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>a</th><th>b</th><th>c</th><th>c_min</th></tr><tr><td>i64</td><td>i64</td><td>i64</td><td>i64</td></tr></thead><tbody><tr><td>9</td><td>1</td><td>5</td><td>5</td></tr><tr><td>8</td><td>2</td><td>4</td><td>4</td></tr><tr><td>7</td><td>3</td><td>3</td><td>3</td></tr><tr><td>6</td><td>5</td><td>2</td><td>2</td></tr><tr><td>5</td><td>3</td><td>1</td><td>1</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (5, 4)\n",
       "+-----+-----+-----+-------+\n",
       "| a   | b   | c   | c_min |\n",
       "| --- | --- | --- | ---   |\n",
       "| i64 | i64 | i64 | i64   |\n",
       "+=========================+\n",
       "| 9   | 1   | 5   | 5     |\n",
       "| 8   | 2   | 4   | 4     |\n",
       "| 7   | 3   | 3   | 3     |\n",
       "| 6   | 5   | 2   | 2     |\n",
       "| 5   | 3   | 1   | 1     |\n",
       "+-----+-----+-----+-------+"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Group by multiple columns by passing a list of column names or expressions.\n",
    "df.with_columns(\n",
    "    pl.col(\"c\")\n",
    "    .min()\n",
    "    .over([\"a\", \"b\"])\n",
    "    .name.suffix(\"_min\"),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "6f26e477-6200-45bd-bb40-44b1dbf9fa8c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 4)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>city08</th><th>mean_city08</th></tr><tr><td>i16</td><td>cat</td><td>u8</td><td>f64</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>19</td><td>19.402062</td></tr><tr><td>1985</td><td>&quot;Ferrari&quot;</td><td>9</td><td>11.842294</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>23</td><td>15.461967</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 4)\n",
       "+------+------------+--------+-------------+\n",
       "| year | make       | city08 | mean_city08 |\n",
       "| ---  | ---        | ---    | ---         |\n",
       "| i16  | cat        | u8     | f64         |\n",
       "+==========================================+\n",
       "| 1985 | Alfa Romeo | 19     | 19.402062   |\n",
       "| 1985 | Ferrari    | 9      | 11.842294   |\n",
       "| 1985 | Dodge      | 23     | 15.461967   |\n",
       "+------+------------+--------+-------------+"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        pl.col('year', 'make', 'city08')\n",
    "    )\n",
    "    .with_columns(\n",
    "        mean_city08=pl.col('city08')\n",
    "            .mean()\n",
    "            .over('make')\n",
    "    )\n",
    ").head(3)\n",
    "\n",
    "# Grouping columns: The columns to group by are specified as \n",
    "# arguments to .over(). \n",
    "# For example, pl.col(\"value\").mean().over(\"group\") \n",
    "# computes the mean of \"value\" for each group defined by the \"group\" column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "8b171f9c-8eb0-4554-a76e-5bf79ff2400b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>trany</th><th>mean_city08</th></tr><tr><td>str</td><td>f64</td></tr></thead><tbody><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&quot;Automatic (A1)&quot;</td><td>97.17284</td></tr><tr><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Manual 3-spd&quot;</td><td>14.25974</td></tr><tr><td>&quot;Manual 3-spd&quot;</td><td>14.25974</td></tr><tr><td>&quot;Manual 3-spd&quot;</td><td>14.25974</td></tr><tr><td>&quot;Manual 3-spd&quot;</td><td>14.25974</td></tr><tr><td>&quot;Manual 3-spd&quot;</td><td>14.25974</td></tr><tr><td>&quot;Manual 3-spd&quot;</td><td>14.25974</td></tr><tr><td>&quot;Automatic (L3)&quot;</td><td>14.0</td></tr><tr><td>&quot;Automatic (L3)&quot;</td><td>14.0</td></tr><tr><td>&quot;Automatic (L4)&quot;</td><td>13.5</td></tr><tr><td>&quot;Automatic (L4)&quot;</td><td>13.5</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 2)\n",
       "+----------------+-------------+\n",
       "| trany          | mean_city08 |\n",
       "| ---            | ---         |\n",
       "| str            | f64         |\n",
       "+==============================+\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| Automatic (A1) | 97.17284    |\n",
       "| …              | …           |\n",
       "| Manual 3-spd   | 14.25974    |\n",
       "| Manual 3-spd   | 14.25974    |\n",
       "| Manual 3-spd   | 14.25974    |\n",
       "| Manual 3-spd   | 14.25974    |\n",
       "| Manual 3-spd   | 14.25974    |\n",
       "| Manual 3-spd   | 14.25974    |\n",
       "| Automatic (L3) | 14.0        |\n",
       "| Automatic (L3) | 14.0        |\n",
       "| Automatic (L4) | 13.5        |\n",
       "| Automatic (L4) | 13.5        |\n",
       "+----------------+-------------+"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Multiple grouping columns\n",
    "# Compute the mean of city08 for every group defined by .over\n",
    "# (you can think of .over as akin to .group_by)\n",
    "\n",
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        pl.col('trany'),\n",
    "        mean_city08=pl.col('city08')\n",
    "            .mean()\n",
    "            .over(['trany'])\n",
    "    )\n",
    "    .sort(by='mean_city08', descending=True)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5b987819-b780-4e30-9aff-b0a9b29d0236",
   "metadata": {},
   "source": [
    "## col.is_in\n",
    "\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.is_in.html\n",
    "* Check if elements of this expression are present in the other Series."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "cbe8c3a1-8247-4def-852b-882093a95890",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: (3, 2)\n",
      "+-----------+------------------+\n",
      "| sets      | optional_members |\n",
      "| ---       | ---              |\n",
      "| list[i64] | i64              |\n",
      "+==============================+\n",
      "| [1, 2, 3] | 1                |\n",
      "| [1, 2]    | 2                |\n",
      "| [9, 10]   | 3                |\n",
      "+-----------+------------------+\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (3, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>sets</th><th>optional_members</th><th>om_is_in_sets</th></tr><tr><td>list[i64]</td><td>i64</td><td>bool</td></tr></thead><tbody><tr><td>[1, 2, 3]</td><td>1</td><td>true</td></tr><tr><td>[1, 2]</td><td>2</td><td>true</td></tr><tr><td>[9, 10]</td><td>3</td><td>false</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (3, 3)\n",
       "+-----------+------------------+---------------+\n",
       "| sets      | optional_members | om_is_in_sets |\n",
       "| ---       | ---              | ---           |\n",
       "| list[i64] | i64              | bool          |\n",
       "+==============================================+\n",
       "| [1, 2, 3] | 1                | true          |\n",
       "| [1, 2]    | 2                | true          |\n",
       "| [9, 10]   | 3                | false         |\n",
       "+-----------+------------------+---------------+"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pl.DataFrame(\n",
    "    {\n",
    "        \"sets\": [[1, 2, 3], [1, 2], [9, 10]], \n",
    "        \"optional_members\": [1, 2, 3]\n",
    "    }\n",
    ")\n",
    "print(df)\n",
    "df.with_columns(om_is_in_sets=pl.col(\"optional_members\").is_in(\"sets\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "ff5a9286-405c-48c8-9ea2-f2153a3bf7e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1985</td><td>&quot;Ferrari&quot;</td><td>&quot;Testarossa&quot;</td><td>4.9</td><td>12</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>9</td><td>14</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;Charger&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Subcompact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.018888</td><td>23</td><td>33</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1985</td><td>&quot;Dodge&quot;</td><td>&quot;B150/B250 Wagon 2WD&quot;</td><td>5.2</td><td>8</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Vans&quot;</td><td>&quot;Regular&quot;</td><td>27.046364</td><td>10</td><td>12</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>15.658422</td><td>17</td><td>23</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Loyale&quot;</td><td>1.8</td><td>4</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>13.523182</td><td>21</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Loyale&quot;</td><td>1.8</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.9004</td><td>22</td><td>29</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Toyota&quot;</td><td>&quot;Corolla&quot;</td><td>1.6</td><td>4</td><td>&quot;Automatic 3-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>12.39625</td><td>23</td><td>26</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>3</td></tr><tr><td>1993</td><td>&quot;Toyota&quot;</td><td>&quot;Corolla&quot;</td><td>1.6</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.442693</td><td>23</td><td>31</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Toyota&quot;</td><td>&quot;Corolla&quot;</td><td>1.8</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.9004</td><td>23</td><td>30</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>1993</td><td>&quot;Saab&quot;</td><td>&quot;900&quot;</td><td>2.1</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.8755</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Saturn&quot;</td><td>&quot;SL&quot;</td><td>1.9</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.442693</td><td>23</td><td>33</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>1993</td><td>&quot;Saturn&quot;</td><td>&quot;SL&quot;</td><td>1.9</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>12.39625</td><td>21</td><td>30</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>1993</td><td>&quot;Saturn&quot;</td><td>&quot;SL&quot;</td><td>1.9</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>10.625357</td><td>24</td><td>33</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Saturn&quot;</td><td>&quot;SL&quot;</td><td>1.9</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>11.9004</td><td>21</td><td>32</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>13.523182</td><td>19</td><td>26</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Front-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>12.935218</td><td>20</td><td>28</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD&quot;</td><td>2.2</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>18</td><td>24</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr><tr><td>1993</td><td>&quot;Subaru&quot;</td><td>&quot;Legacy AWD Turbo&quot;</td><td>2.2</td><td>4</td><td>&quot;Automatic 4-spd&quot;</td><td>&quot;4-Wheel or All-Wheel Drive&quot;</td><td>&quot;Compact Cars&quot;</td><td>&quot;Premium&quot;</td><td>16.528334</td><td>16</td><td>21</td><td>2013-01-01 00:00:00</td><td>&quot;true&quot;</td><td>4</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25     | 2013- | false | 5     |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 198 | Fer | Tes | 4.9 | 12  | Man | Rea | Two | Reg | 27. | 9   | 14     | 2013- | false | 5     |\n",
       "| 5   | rar | tar |     |     | ual | r-W | Sea | ula | 046 |     |        | 01-01 |       |       |\n",
       "|     | i   | oss |     |     | 5-s | hee | ter | r   | 364 |     |        | 00:00 |       |       |\n",
       "|     |     | a   |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 198 | Dod | Cha | 2.2 | 4   | Man | Fro | Sub | Reg | 11. | 23  | 33     | 2013- | false | 5     |\n",
       "| 5   | ge  | rge |     |     | ual | nt- | com | ula | 018 |     |        | 01-01 |       |       |\n",
       "|     |     | r   |     |     | 5-s | Whe | pac | r   | 888 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | t   |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | Car |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  | s   |     |     |     |        |       |       |       |\n",
       "| 198 | Dod | B15 | 5.2 | 8   | Aut | Rea | Van | Reg | 27. | 10  | 12     | 2013- | true  | 3     |\n",
       "| 5   | ge  | 0/B |     |     | oma | r-W | s   | ula | 046 |     |        | 01-01 |       |       |\n",
       "|     |     | 250 |     |     | tic | hee |     | r   | 364 |     |        | 00:00 |       |       |\n",
       "|     |     | Wag |     |     | 3-s | l   |     |     |     |     |        | :00   |       |       |\n",
       "|     |     | on  |     |     | pd  | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | 2WD |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | 4-W | Com | Pre | 15. | 17  | 23     | 2013- | false | 5     |\n",
       "| 3   | aru | acy |     |     | ual | hee | pac | miu | 658 |     |        | 01-01 |       |       |\n",
       "|     |     | AWD |     |     | 5-s | l   | t   | m   | 422 |     |        | 00:00 |       |       |\n",
       "|     |     | Tur |     |     | pd  | or  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     | bo  |     |     |     | All | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Loy | 1.8 | 4   | Aut | Fro | Com | Reg | 13. | 21  | 24     | 2013- | true  | 3     |\n",
       "| 3   | aru | ale |     |     | oma | nt- | pac | ula | 523 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | tic | Whe | t   | r   | 182 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 3-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Loy | 1.8 | 4   | Man | Fro | Com | Reg | 11. | 22  | 29     | 2013- | false | 5     |\n",
       "| 3   | aru | ale |     |     | ual | nt- | pac | ula | 900 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 4   |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Toy | Cor | 1.6 | 4   | Aut | Fro | Com | Reg | 12. | 23  | 26     | 2013- | true  | 3     |\n",
       "| 3   | ota | oll |     |     | oma | nt- | pac | ula | 396 |     |        | 01-01 |       |       |\n",
       "|     |     | a   |     |     | tic | Whe | t   | r   | 25  |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 3-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Toy | Cor | 1.6 | 4   | Man | Fro | Com | Reg | 11. | 23  | 31     | 2013- | false | 5     |\n",
       "| 3   | ota | oll |     |     | ual | nt- | pac | ula | 442 |     |        | 01-01 |       |       |\n",
       "|     |     | a   |     |     | 5-s | Whe | t   | r   | 693 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Toy | Cor | 1.8 | 4   | Aut | Fro | Com | Reg | 11. | 23  | 30     | 2013- | true  | 4     |\n",
       "| 3   | ota | oll |     |     | oma | nt- | pac | ula | 900 |     |        | 01-01 |       |       |\n",
       "|     |     | a   |     |     | tic | Whe | t   | r   | 4   |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 4-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …   | …      | …     | …     | …     |\n",
       "| 199 | Saa | 900 | 2.1 | 4   | Man | Fro | Com | Reg | 14. | 18  | 24     | 2013- | false | 5     |\n",
       "| 3   | b   |     |     |     | ual | nt- | pac | ula | 875 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 5   |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sat | SL  | 1.9 | 4   | Aut | Fro | Com | Reg | 11. | 23  | 33     | 2013- | true  | 4     |\n",
       "| 3   | urn |     |     |     | oma | nt- | pac | ula | 442 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | tic | Whe | t   | r   | 693 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 4-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sat | SL  | 1.9 | 4   | Aut | Fro | Com | Reg | 12. | 21  | 30     | 2013- | true  | 4     |\n",
       "| 3   | urn |     |     |     | oma | nt- | pac | ula | 396 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | tic | Whe | t   | r   | 25  |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 4-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sat | SL  | 1.9 | 4   | Man | Fro | Com | Reg | 10. | 24  | 33     | 2013- | false | 5     |\n",
       "| 3   | urn |     |     |     | ual | nt- | pac | ula | 625 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 357 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sat | SL  | 1.9 | 4   | Man | Fro | Com | Reg | 11. | 21  | 32     | 2013- | false | 5     |\n",
       "| 3   | urn |     |     |     | ual | nt- | pac | ula | 900 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 4   |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | Fro | Com | Reg | 13. | 19  | 26     | 2013- | true  | 4     |\n",
       "| 3   | aru | acy |     |     | oma | nt- | pac | ula | 523 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | tic | Whe | t   | r   | 182 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 4-s | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | Fro | Com | Reg | 12. | 20  | 28     | 2013- | false | 5     |\n",
       "| 3   | aru | acy |     |     | ual | nt- | pac | ula | 935 |     |        | 01-01 |       |       |\n",
       "|     |     |     |     |     | 5-s | Whe | t   | r   | 218 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | el  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | Dri | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | 4-W | Com | Reg | 14. | 18  | 24     | 2013- | true  | 4     |\n",
       "| 3   | aru | acy |     |     | oma | hee | pac | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     |     | AWD |     |     | tic | l   | t   | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | 4-s | or  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     | pd  | All | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Man | 4-W | Com | Reg | 14. | 18  | 24     | 2013- | false | 5     |\n",
       "| 3   | aru | acy |     |     | ual | hee | pac | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     |     | AWD |     |     | 5-s | l   | t   | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     |     |     |     |     | pd  | or  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     |     |     |     |     | All | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "| 199 | Sub | Leg | 2.2 | 4   | Aut | 4-W | Com | Pre | 16. | 16  | 21     | 2013- | true  | 4     |\n",
       "| 3   | aru | acy |     |     | oma | hee | pac | miu | 528 |     |        | 01-01 |       |       |\n",
       "|     |     | AWD |     |     | tic | l   | t   | m   | 334 |     |        | 00:00 |       |       |\n",
       "|     |     | Tur |     |     | 4-s | or  | Car |     |     |     |        | :00   |       |       |\n",
       "|     |     | bo  |     |     | pd  | All | s   |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | -Wh |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | eel |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     |     |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "baf5794b-a2d0-4e0d-8104-ee8882754995",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "keys/aggregates are not partitionable: running default HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (2, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>group</th><th>filtered_sum</th></tr><tr><td>str</td><td>i64</td></tr></thead><tbody><tr><td>&quot;A&quot;</td><td>0</td></tr><tr><td>&quot;B&quot;</td><td>12</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (2, 2)\n",
       "+-------+--------------+\n",
       "| group | filtered_sum |\n",
       "| ---   | ---          |\n",
       "| str   | i64          |\n",
       "+======================+\n",
       "| A     | 0            |\n",
       "| B     | 12           |\n",
       "+-------+--------------+"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a sample DataFrame\n",
    "df = pl.DataFrame({\n",
    "    \"group\": [\"A\", \"A\", \"B\", \"B\", \"B\"],\n",
    "    \"value\": [1, 2, 3, 4, 5]\n",
    "})\n",
    "\n",
    "# Group by 'group' column and filter 'value' column\n",
    "filtered = df.group_by(\"group\").agg(\n",
    "    filtered_sum=pl.col(\"value\").filter(pl.col(\"value\") > 2).sum()\n",
    ")\n",
    "filtered"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "5d7a0a7d-a3fa-4123-b995-ac03ec024f2a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (279, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>trany</th><th>IS_5_SPEED</th></tr><tr><td>cat</td><td>str</td><td>bool</td></tr></thead><tbody><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Automatic (S5)&quot;</td><td>false</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 6-spd&quot;</td><td>false</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 6-spd&quot;</td><td>false</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Automatic (S5)&quot;</td><td>false</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Automatic (S5)&quot;</td><td>false</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (279, 3)\n",
       "+---------+----------------+------------+\n",
       "| make    | trany          | IS_5_SPEED |\n",
       "| ---     | ---            | ---        |\n",
       "| cat     | str            | bool       |\n",
       "+=======================================+\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Automatic (S5) | false      |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 6-spd   | false      |\n",
       "| Ferrari | Manual 6-spd   | false      |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| …       | …              | …          |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Automatic (S5) | false      |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Manual 5-spd   | true       |\n",
       "| Ferrari | Automatic (S5) | false      |\n",
       "+---------+----------------+------------+"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Variation using .is_in, essentially a full-match\n",
    "(\n",
    "    autos\n",
    "    .select(pl.col(['make', 'trany']))\n",
    "    .filter(pl.col('make')=='Ferrari')\n",
    "    .with_columns(\n",
    "        IS_5_SPEED=pl.col(\"trany\").is_in([\"Manual 5-spd\"])\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "8971631c-ac49-423a-aa2d-898881ce5d41",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "dataframe filtered\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (279, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>make</th><th>trany</th><th>IS_MANUAL</th></tr><tr><td>cat</td><td>str</td><td>bool</td></tr></thead><tbody><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Automatic (S5)&quot;</td><td>false</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 6-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 6-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&hellip;</td><td>&hellip;</td><td>&hellip;</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Automatic (S5)&quot;</td><td>false</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Manual 5-spd&quot;</td><td>true</td></tr><tr><td>&quot;Ferrari&quot;</td><td>&quot;Automatic (S5)&quot;</td><td>false</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (279, 3)\n",
       "+---------+----------------+-----------+\n",
       "| make    | trany          | IS_MANUAL |\n",
       "| ---     | ---            | ---       |\n",
       "| cat     | str            | bool      |\n",
       "+======================================+\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Automatic (S5) | false     |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 6-spd   | true      |\n",
       "| Ferrari | Manual 6-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| …       | …              | …         |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Automatic (S5) | false     |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Manual 5-spd   | true      |\n",
       "| Ferrari | Automatic (S5) | false     |\n",
       "+---------+----------------+-----------+"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Variation using .str.contains, allowing partial-matches\n",
    "(\n",
    "    autos\n",
    "    .select(pl.col(['make', 'trany']))\n",
    "    .filter(pl.col('make')=='Ferrari')\n",
    "    .with_columns(\n",
    "        IS_MANUAL=pl.col('trany').str.contains('Manual') \n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14483458-818d-4e82-b39e-b500411e6343",
   "metadata": {},
   "source": [
    "## col.str.extract\n",
    "\n",
    "* https://docs.pola.rs/py-polars/html/reference/expressions/api/polars.Expr.str.extract.html\n",
    "* Extract the target capture group from provided patterns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "08c49ddd-97ca-4ac7-a162-672a1d12d79e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 15)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>year</th><th>make</th><th>model</th><th>displ</th><th>cylinders</th><th>trany</th><th>drive</th><th>VClass</th><th>fuelType</th><th>barrels08</th><th>city08</th><th>highway08</th><th>createdOn</th><th>is_automatic</th><th>num_gears</th></tr><tr><td>i16</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>str</td><td>cat</td><td>cat</td><td>cat</td><td>f32</td><td>u8</td><td>u8</td><td>datetime[μs]</td><td>str</td><td>u8</td></tr></thead><tbody><tr><td>1985</td><td>&quot;Alfa Romeo&quot;</td><td>&quot;Spider Veloce 2000&quot;</td><td>2.0</td><td>4</td><td>&quot;Manual 5-spd&quot;</td><td>&quot;Rear-Wheel Drive&quot;</td><td>&quot;Two Seaters&quot;</td><td>&quot;Regular&quot;</td><td>14.167143</td><td>19</td><td>25</td><td>2013-01-01 00:00:00</td><td>&quot;false&quot;</td><td>5</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 15)\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+\n",
       "| yea | mak | mod | dis | cyl | tra | dri | VCl | fue | bar | cit | highwa | creat | is_au | num_g |\n",
       "| r   | e   | el  | pl  | ind | ny  | ve  | ass | lTy | rel | y08 | y08    | edOn  | tomat | ears  |\n",
       "| --- | --- | --- | --- | ers | --- | --- | --- | pe  | s08 | --- | ---    | ---   | ic    | ---   |\n",
       "| i16 | cat | cat | f32 | --- | str | cat | cat | --- | --- | u8  | u8     | datet | ---   | u8    |\n",
       "|     |     |     |     | u8  |     |     |     | cat | f32 |     |        | ime[μ | str   |       |\n",
       "|     |     |     |     |     |     |     |     |     |     |     |        | s]    |       |       |\n",
       "+==================================================================================================+\n",
       "| 198 | Alf | Spi | 2.0 | 4   | Man | Rea | Two | Reg | 14. | 19  | 25     | 2013- | false | 5     |\n",
       "| 5   | a   | der |     |     | ual | r-W | Sea | ula | 167 |     |        | 01-01 |       |       |\n",
       "|     | Rom | Vel |     |     | 5-s | hee | ter | r   | 143 |     |        | 00:00 |       |       |\n",
       "|     | eo  | oce |     |     | pd  | l   | s   |     |     |     |        | :00   |       |       |\n",
       "|     |     | 200 |     |     |     | Dri |     |     |     |     |        |       |       |       |\n",
       "|     |     | 0   |     |     |     | ve  |     |     |     |     |        |       |       |       |\n",
       "+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+--------+-------+-------+-------+"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autos.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "c9f3fa93-f40f-4dcc-8222-75588bd07eb3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (48_202, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>num_gears</th></tr><tr><td>u16</td></tr></thead><tbody><tr><td>5</td></tr><tr><td>5</td></tr><tr><td>5</td></tr><tr><td>3</td></tr><tr><td>5</td></tr><tr><td>3</td></tr><tr><td>5</td></tr><tr><td>3</td></tr><tr><td>5</td></tr><tr><td>4</td></tr><tr><td>&hellip;</td></tr><tr><td>5</td></tr><tr><td>4</td></tr><tr><td>4</td></tr><tr><td>5</td></tr><tr><td>5</td></tr><tr><td>4</td></tr><tr><td>5</td></tr><tr><td>4</td></tr><tr><td>5</td></tr><tr><td>4</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (48_202, 1)\n",
       "+-----------+\n",
       "| num_gears |\n",
       "| ---       |\n",
       "| u16       |\n",
       "+===========+\n",
       "| 5         |\n",
       "| 5         |\n",
       "| 5         |\n",
       "| 3         |\n",
       "| 5         |\n",
       "| 3         |\n",
       "| 5         |\n",
       "| 3         |\n",
       "| 5         |\n",
       "| 4         |\n",
       "| …         |\n",
       "| 5         |\n",
       "| 4         |\n",
       "| 4         |\n",
       "| 5         |\n",
       "| 5         |\n",
       "| 4         |\n",
       "| 5         |\n",
       "| 4         |\n",
       "| 5         |\n",
       "| 4         |\n",
       "+-----------+"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        num_gears = pl.col('trany')\n",
    "            .str.extract(r'(\\d+)')\n",
    "            .cast(pl.UInt8)\n",
    "            .fill_null(999)\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "c01d1312-3a9c-42a7-bd61-91ecbd9ba273",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (2, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>num_gears</th></tr><tr><td>str</td></tr></thead><tbody><tr><td>&quot;999&quot;</td></tr><tr><td>&quot;999&quot;</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (2, 1)\n",
       "+-----------+\n",
       "| num_gears |\n",
       "| ---       |\n",
       "| str       |\n",
       "+===========+\n",
       "| 999       |\n",
       "| 999       |\n",
       "+-----------+"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        num_gears = pl.col('make')\n",
    "            .cast(pl.String)\n",
    "            .str.extract(r'(\\d+)')\n",
    "\n",
    "            .fill_null(999)\n",
    "    )\n",
    ").head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "ad0da578-8637-4dd8-b35e-aa1a88080eb3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (20, 1)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>VClass</th></tr><tr><td>cat</td></tr></thead><tbody><tr><td>&quot;Two Seaters&quot;</td></tr><tr><td>&quot;Two Seaters&quot;</td></tr><tr><td>&quot;Subcompact Cars&quot;</td></tr><tr><td>&quot;Vans&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Vans&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Compact Cars&quot;</td></tr><tr><td>&quot;Midsize Cars&quot;</td></tr><tr><td>&quot;Midsize Cars&quot;</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (20, 1)\n",
       "+-----------------+\n",
       "| VClass          |\n",
       "| ---             |\n",
       "| cat             |\n",
       "+=================+\n",
       "| Two Seaters     |\n",
       "| Two Seaters     |\n",
       "| Subcompact Cars |\n",
       "| Vans            |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Vans            |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Compact Cars    |\n",
       "| Midsize Cars    |\n",
       "| Midsize Cars    |\n",
       "+-----------------+"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "autos.select(pl.col('VClass').head(20))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "12e39e48-5fb0-4d2e-9bea-8edecf9a4048",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "estimated unique values: 2\n",
      "run PARTITIONED HASH AGGREGATION\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (2, 2)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>partial</th><th>len</th></tr><tr><td>str</td><td>u32</td></tr></thead><tbody><tr><td>null</td><td>46132</td></tr><tr><td>&quot;Mini&quot;</td><td>2070</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (2, 2)\n",
       "+---------+-------+\n",
       "| partial | len   |\n",
       "| ---     | ---   |\n",
       "| str     | u32   |\n",
       "+=================+\n",
       "| null    | 46132 |\n",
       "| Mini    | 2070  |\n",
       "+---------+-------+"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .select(\n",
    "        total=pl.col('VClass').len(),\n",
    "        #full=pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)compact\\b\", group_index=0), #\n",
    "        partial=pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)mini\", group_index=0),\n",
    "    )\n",
    "    .group_by('partial')\n",
    "    .len()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "0ac8dbad-17b2-42b5-9271-87d853f61c57",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>partial</th><th>total</th><th>normalized</th></tr><tr><td>u32</td><td>u32</td><td>f64</td></tr></thead><tbody><tr><td>2070</td><td>48202</td><td>0.042944</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 3)\n",
       "+---------+-------+------------+\n",
       "| partial | total | normalized |\n",
       "| ---     | ---   | ---        |\n",
       "| u32     | u32   | f64        |\n",
       "+==============================+\n",
       "| 2070    | 48202 | 0.042944   |\n",
       "+---------+-------+------------+"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    autos\n",
    "    .with_columns(\n",
    "        total=pl.col('VClass').len()\n",
    "    )\n",
    "    .with_columns(\n",
    "        #full=pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)compact\\b\", group_index=0), #\n",
    "        # Use .count to count non-null values\n",
    "        partial=pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)mini\", group_index=0).count()\n",
    "    )\n",
    "    .with_columns(\n",
    "        normalized = pl.col('partial') / pl.col('total')\n",
    "    )\n",
    "    .select(['partial', 'total', 'normalized']).head(1)\n",
    "    #.group_by('partial')\n",
    "    #.len()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "4ca53061-8eb4-4bf1-8e02-73acc49a9b0c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>mini</th><th>total</th><th>normalized</th></tr><tr><td>u32</td><td>u32</td><td>f64</td></tr></thead><tbody><tr><td>2070</td><td>48202</td><td>0.042944</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 3)\n",
       "+------+-------+------------+\n",
       "| mini | total | normalized |\n",
       "| ---  | ---   | ---        |\n",
       "| u32  | u32   | f64        |\n",
       "+===========================+\n",
       "| 2070 | 48202 | 0.042944   |\n",
       "+------+-------+------------+"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hack to ensure sequential processing, \n",
    "# and enable access to immediately created variables\n",
    "(\n",
    "    autos\n",
    "    .select(pl.col('VClass'))\n",
    "    .with_columns(total=pl.col('VClass').len())\n",
    "    .with_columns(mini=pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)mini\", group_index=0).count())\n",
    "    .with_columns(normalized = (pl.col('mini') / pl.col('total')))\n",
    "    .select(['mini', 'total', 'normalized'])\n",
    "    .head(1)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "089a6932-4b81-43b1-a47c-f2563fac00c1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><style>\n",
       ".dataframe > thead > tr,\n",
       ".dataframe > tbody > tr {\n",
       "  text-align: right;\n",
       "  white-space: pre-wrap;\n",
       "}\n",
       "</style>\n",
       "<small>shape: (1, 3)</small><table border=\"1\" class=\"dataframe\"><thead><tr><th>mini</th><th>total</th><th>ratio</th></tr><tr><td>u32</td><td>u32</td><td>f64</td></tr></thead><tbody><tr><td>2070</td><td>48202</td><td>0.042944</td></tr></tbody></table></div>"
      ],
      "text/plain": [
       "shape: (1, 3)\n",
       "+------+-------+----------+\n",
       "| mini | total | ratio    |\n",
       "| ---  | ---   | ---      |\n",
       "| u32  | u32   | f64      |\n",
       "+=========================+\n",
       "| 2070 | 48202 | 0.042944 |\n",
       "+------+-------+----------+"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Hack to ensure sequential processing, \n",
    "# and enable access to immediately created variables\n",
    "(\n",
    "    autos\n",
    "    .select_seq(\n",
    "        mini=pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)mini\", group_index=0).count(),\n",
    "        total=pl.col('VClass').len(),        \n",
    "        ratio=(\n",
    "            pl.col('VClass').cast(pl.String).str.extract(r\"\\b(?i)mini\", group_index=0).count() / pl.col('VClass').len()\n",
    "        )\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "25211587-c35a-4a0e-bd6f-44c7f2ffa930",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}