notebooks/t2_SKAB.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pipeline of the anomaly detection on the SKAB using Hotelling's $T^2$ statistic"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Details regarding the SKAB one can find in the [SKAB repository](https://github.com/waico/SKAB)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Libraries importing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# libraries importing\n",
    "import sys\n",
    "import warnings\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "sys.path.append(\"..\")\n",
    "from core.t2 import T2\n",
    "from core.metrics import chp_score\n",
    "from core.utils import load_preprocess_skab, plot_results\n",
    "\n",
    "warnings.filterwarnings(\"ignore\", category=UserWarning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xy_traintest_list = load_preprocess_skab()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Method initialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# classifier initializing\n",
    "model = T2(scaling=True, using_pca=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Method fitting and applying"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# inference\n",
    "predicted_outlier, predicted_cp = [], []\n",
    "true_outlier, true_cp = [], []\n",
    "for X_train, X_test, y_train, y_test in Xy_traintest_list:\n",
    "    # classifier fitting\n",
    "    model.fit(X_train)\n",
    "\n",
    "    # results predicting\n",
    "    model.predict(\n",
    "        X_test,\n",
    "        window_size=5,\n",
    "        plot_fig=False,\n",
    "    )\n",
    "    prediction = pd.Series(\n",
    "        (model.t2[\"T2\"].values > (2 * model.t2_ucl)).astype(int),\n",
    "        index=X_test.index,\n",
    "    ).fillna(0)\n",
    "    # predicted outliers saving\n",
    "    predicted_outlier.append(prediction)\n",
    "\n",
    "    # predicted CPs saving\n",
    "    prediction_cp = abs(prediction.diff())\n",
    "    prediction_cp[0] = prediction[0]\n",
    "    predicted_cp.append(prediction_cp)\n",
    "\n",
    "    true_outlier.append(y_test[\"anomaly\"])\n",
    "    true_cp.append(y_test[\"changepoint\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Results visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# [donotremove]\n",
    "plot_results(\n",
    "    (true_outlier[1], predicted_outlier[1]),\n",
    "    (true_cp[1], predicted_cp[1]),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation (metrics calculation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "\n",
    "pickle.dump(\n",
    "    predicted_outlier,\n",
    "    open(f\"../results/results-{model.__class__.__name__}.pkl\", \"wb\"),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Binary classification (outlier detection) metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "False Alarm Rate 19.21 %\n",
      "Missing Alarm Rate 42.6 %\n",
      "F1 metric 0.66\n"
     ]
    }
   ],
   "source": [
    "# [donotremove]\n",
    "# binary classification metrics calculation\n",
    "binary = chp_score(true_outlier, predicted_outlier, metric=\"binary\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Changepoint detection metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Amount of true anomalies 127\n",
      "A number of missed CPs = 84\n",
      "A number of FPs = 203\n",
      "Average time 0 days 00:00:18.232558139\n"
     ]
    }
   ],
   "source": [
    "# [donotremove]\n",
    "# average detection delay metric calculation\n",
    "add = chp_score(\n",
    "    true_cp,\n",
    "    predicted_cp,\n",
    "    metric=\"average_time\",\n",
    "    window_width=\"60s\",\n",
    "    anomaly_window_destination=\"righter\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Standard  -  19.54\n",
      "LowFP  -  10.2\n",
      "LowFN  -  24.31\n"
     ]
    }
   ],
   "source": [
    "# [donotremove]\n",
    "# nab metric calculation\n",
    "nab = chp_score(\n",
    "    true_cp,\n",
    "    predicted_cp,\n",
    "    metric=\"nab\",\n",
    "    window_width=\"60s\",\n",
    "    anomaly_window_destination=\"righter\",\n",
    ")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "",
   "language": "python",
   "name": ""
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.14"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "toc-autonumbering": true
 },
 "nbformat": 4,
 "nbformat_minor": 4
}