17. Engagement Test.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d0c9d0c1",
   "metadata": {},
   "source": [
    "# Goal of Analysis:\n",
    "\n",
    "### 1. A model was built to recommend 5 friends to the user, is the test result successful that we should implement to all the users?\n",
    "\n",
    "### 2. Is the test performing similarly for all user segments or are there differences among different segments?\n",
    "\n",
    "### 3. If you identified segments that responded differently to the test, can you guess the reason? Would this change your point 1 conclusions?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "6ec608dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "84473ea6",
   "metadata": {},
   "source": [
    "### Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "36a731e6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>signup_date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>34</td>\n",
       "      <td>2015-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>59</td>\n",
       "      <td>2015-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>178</td>\n",
       "      <td>2015-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>285</td>\n",
       "      <td>2015-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>383</td>\n",
       "      <td>2015-01-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id signup_date\n",
       "0       34  2015-01-01\n",
       "1       59  2015-01-01\n",
       "2      178  2015-01-01\n",
       "3      285  2015-01-01\n",
       "4      383  2015-01-01"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user = pd.read_csv('user_table.csv')\n",
    "user.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "47ea8eed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>date</th>\n",
       "      <th>browser</th>\n",
       "      <th>test</th>\n",
       "      <th>pages_visited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>600597</td>\n",
       "      <td>2015-08-13</td>\n",
       "      <td>IE</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4410028</td>\n",
       "      <td>2015-08-26</td>\n",
       "      <td>Chrome</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6004777</td>\n",
       "      <td>2015-08-17</td>\n",
       "      <td>Chrome</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5990330</td>\n",
       "      <td>2015-08-27</td>\n",
       "      <td>Safari</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3622310</td>\n",
       "      <td>2015-08-07</td>\n",
       "      <td>Firefox</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id        date  browser  test  pages_visited\n",
       "0   600597  2015-08-13       IE     0              2\n",
       "1  4410028  2015-08-26   Chrome     1              5\n",
       "2  6004777  2015-08-17   Chrome     0              8\n",
       "3  5990330  2015-08-27   Safari     0              8\n",
       "4  3622310  2015-08-07  Firefox     0              1"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test = pd.read_csv('test_table.csv')\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c6d023c0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>signup_date</th>\n",
       "      <th>date</th>\n",
       "      <th>browser</th>\n",
       "      <th>test</th>\n",
       "      <th>pages_visited</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>34</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-15</td>\n",
       "      <td>Chrome</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>59</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-12</td>\n",
       "      <td>Chrome</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>178</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-10</td>\n",
       "      <td>Safari</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>285</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-03</td>\n",
       "      <td>Opera</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>383</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-05</td>\n",
       "      <td>Firefox</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id signup_date        date  browser  test  pages_visited\n",
       "0       34  2015-01-01  2015-08-15   Chrome     0              6\n",
       "1       59  2015-01-01  2015-08-12   Chrome     1              6\n",
       "2      178  2015-01-01  2015-08-10   Safari     1              3\n",
       "3      285  2015-01-01  2015-08-03    Opera     0              5\n",
       "4      383  2015-01-01  2015-08-05  Firefox     1              9"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# merge two datasets to have full set of data\n",
    "\n",
    "data = pd.merge(left = user, right = test, on = 'user_id')\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a31d175b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((100000, 2), (100000, 5), (100000, 6))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user.shape, test.shape, data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7f683ac",
   "metadata": {},
   "source": [
    "### Qusetion 1: Is the test winning?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "722b989c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "test\n",
       "0    4.608394\n",
       "1    4.599693\n",
       "Name: pages_visited, dtype: float64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('test')['pages_visited'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "457b3fc0",
   "metadata": {},
   "source": [
    "We can see that the mean of two groups don't have large difference."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6a83148c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T-test:\t Ttest_indResult(statistic=0.5571118435554797, pvalue=0.5774523171559118)\n"
     ]
    }
   ],
   "source": [
    "# perform student's t-test\n",
    "from scipy.stats import ttest_ind\n",
    "\n",
    "control_page = data[data['test'] == 0]['pages_visited'].values\n",
    "experiment_page = data[data['test'] == 1]['pages_visited'].values\n",
    "\n",
    "print('T-test:\\t', ttest_ind(a=control_page, b=experiment_page, equal_var=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "353161dd",
   "metadata": {},
   "source": [
    "The result shows that the test statistic is 0.577, which is larger than 5%, so we cannot reject the null hypothesis that there is no difference between control and experimet group. \n",
    "\n",
    "Therefore, the test doesn't win at all, and we should not expand it to all the users."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd6928cd",
   "metadata": {},
   "source": [
    "### Question 2: Is the test performing similarly for all user segments?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "92f948bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# first convert the date into datetime in the dataframe\n",
    "\n",
    "data['date'] = pd.to_datetime(data.date)\n",
    "data['signup_date'] = pd.to_datetime(data.signup_date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "b847a12c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>signup_date</th>\n",
       "      <th>date</th>\n",
       "      <th>browser</th>\n",
       "      <th>test</th>\n",
       "      <th>pages_visited</th>\n",
       "      <th>interval</th>\n",
       "      <th>is_first</th>\n",
       "      <th>dayofweek</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>34</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-15</td>\n",
       "      <td>Chrome</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>226</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>59</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-12</td>\n",
       "      <td>Chrome</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>223</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>178</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-10</td>\n",
       "      <td>Safari</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>221</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>285</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-03</td>\n",
       "      <td>Opera</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>214</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>383</td>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>2015-08-05</td>\n",
       "      <td>Firefox</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>216</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id signup_date       date  browser  test  pages_visited  interval  \\\n",
       "0       34  2015-01-01 2015-08-15   Chrome     0              6       226   \n",
       "1       59  2015-01-01 2015-08-12   Chrome     1              6       223   \n",
       "2      178  2015-01-01 2015-08-10   Safari     1              3       221   \n",
       "3      285  2015-01-01 2015-08-03    Opera     0              5       214   \n",
       "4      383  2015-01-01 2015-08-05  Firefox     1              9       216   \n",
       "\n",
       "   is_first  dayofweek  \n",
       "0         0          5  \n",
       "1         0          2  \n",
       "2         0          0  \n",
       "3         0          0  \n",
       "4         0          2  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# process the data\n",
    "\n",
    "data['interval'] = (data['date'] - data['signup_date']).apply(lambda x: x.days)\n",
    "data['is_first'] = (data['interval'] == 0).astype(int)\n",
    "data['dayofweek'] = data['date'].apply(lambda x: x.dayofweek)\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e80d0f5e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization of browser\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(18, 6))\n",
    "sns.countplot(x='browser', hue='test', data=data, ax=ax[0])\n",
    "ax[0].set_title('Box Plot of browser', fontsize=16)\n",
    "\n",
    "sns.barplot(x='browser', y='pages_visited', hue='test', data=data, ax=ax[1])\n",
    "ax[1].set_title('Pages Visited vs. browser', fontsize=16)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ff8b531",
   "metadata": {},
   "source": [
    "For different browser, Chrome has the largest number of usrs, but in terms of pages visited, Chrome is almost the same as Safari."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "01de823d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization of is_first\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(18, 6))\n",
    "sns.countplot(x='is_first', hue='test', data=data, ax=ax[0])\n",
    "ax[0].set_title('Box Plot of is_first', fontsize=16)\n",
    "\n",
    "sns.barplot(x='is_first', y='pages_visited', hue='test', data=data, ax=ax[1])\n",
    "ax[1].set_title('Pages Visited vs. is_first', fontsize=16)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9cd0cd0",
   "metadata": {},
   "source": [
    "In terms of pages visited, it seems that whether it's the first time doesn't seem to make many differences."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5230da97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization of dayofweek\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(18, 6))\n",
    "sns.countplot(x='dayofweek', hue='test', data=data, ax=ax[0])\n",
    "ax[0].set_title('Box Plot of dayofweek', fontsize=16)\n",
    "\n",
    "sns.barplot(x='dayofweek', y='pages_visited', hue='test', data=data, ax=ax[1])\n",
    "ax[1].set_title('Pages Visited vs. dayofweek', fontsize=16)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f974b1fc",
   "metadata": {},
   "source": [
    "For all the users, Sunday (day 0) is the most visited day, but in terms of pages visited, the day of week doesn't seem to have many differences."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d657eefb",
   "metadata": {},
   "source": [
    "### Question 3: If you identified segments that responded differently to the test, can you guess the reason? Would this change your point 1 conclusions?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32b99c2b",
   "metadata": {},
   "source": [
    "From the chart above, we can see that the browser Opera has the lowest visited number and doesn't seem to have a clear distinction in test and control group, so we will remove the Opera and see whether the result will change."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5ef2cf34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "test\n",
       "0    4.609804\n",
       "1    4.694989\n",
       "Name: pages_visited, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subset = data[data['browser'] != 'Opera']\n",
    "subset.groupby('test')['pages_visited'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "2fb1bce0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8, 5))\n",
    "sns.barplot(x='test', y='pages_visited', data=subset, ax=ax)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "f657ce61",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T-test:\t Ttest_indResult(statistic=-5.474295518566027, pvalue=4.403954129457701e-08)\n"
     ]
    }
   ],
   "source": [
    "# perform student's t-test\n",
    "\n",
    "control_page = subset[subset['test'] == 0]['pages_visited'].values\n",
    "experiment_page = subset[subset['test'] == 1]['pages_visited'].values\n",
    "\n",
    "print('T-test:\\t', ttest_ind(a=control_page, b=experiment_page, equal_var=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e9d3eed",
   "metadata": {},
   "source": [
    "We can see that if we remove Opera, the test statistic is -5.47, the new feature seems to be significantly userful now, we can also take a look at the is_first visualization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "97b4a5b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization of is_first\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(18, 6))\n",
    "sns.countplot(x='is_first', hue='test', data=subset, ax=ax[0])\n",
    "ax[0].set_title('Box Plot of is_first', fontsize=16)\n",
    "\n",
    "sns.barplot(x='is_first', y='pages_visited', hue='test', data=subset, ax=ax[1])\n",
    "ax[1].set_title('Pages Visited vs. is_first', fontsize=16)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6d956879",
   "metadata": {},
   "source": [
    "There are clear test and control group now, so removing the Opera will change my conclusion in Question 1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "58bf94ba",
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}