ANOVA Test project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ANOVA TESTS for changes in meal delivery app\n",
    "### Task 1. We changed meal pictures sizw and quality in our app. We showed romb shape pictures to \"A\" group of users, square sized pictures to group \"B\" and not changed variant to \"C\". "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Task 1. В первом тестировали разрешение фотографий блюд в приложении: пользователям показывались либо прямоугольные, либо новые квадратные "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/tljh/user/lib/python3.7/site-packages/outdated/utils.py:18: OutdatedCacheFailedWarning: Failed to use cache while checking for outdated package.\n",
      "Set the environment variable OUTDATED_RAISE_EXCEPTION=1 for a full traceback.\n",
      "Set the environment variable OUTDATED_IGNORE=1 to disable these warnings.\n",
      "  **kwargs\n",
      "/opt/tljh/user/lib/python3.7/site-packages/outdated/utils.py:18: OutdatedPackageWarning: The package outdated is out of date. Your version is 0.2.1, the latest is 0.2.2.\n",
      "Set the environment variable OUTDATED_IGNORE=1 to disable these warnings.\n",
      "  **kwargs\n",
      "/opt/tljh/user/lib/python3.7/site-packages/outdated/utils.py:18: OutdatedPackageWarning: The package pingouin is out of date. Your version is 0.4.0, the latest is 0.5.3.\n",
      "Set the environment variable OUTDATED_IGNORE=1 to disable these warnings.\n",
      "  **kwargs\n"
     ]
    }
   ],
   "source": [
    "# Importing needed libraries \n",
    "# Загружаем необходимые библиотеки\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy.stats as ss\n",
    "from scipy. stats import ttest_ind\n",
    "import pingouin as pg\n",
    "\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reading our file with pandas\n",
    "# Считываем наш файл\n",
    "\n",
    "task_1 = pd.read_csv('/mnt/HC_Volume_18315164/home-jupyter/jupyter-a-lavrukhin/5_task_1.csv')\n",
    "task_2 = pd.read_csv('/mnt/HC_Volume_18315164/home-jupyter/jupyter-a-lavrukhin/5_task_2.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>id</th>\n",
       "      <th>group</th>\n",
       "      <th>events</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>16046</td>\n",
       "      <td>A</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>18712</td>\n",
       "      <td>A</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3268</td>\n",
       "      <td>A</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22633</td>\n",
       "      <td>A</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>28071</td>\n",
       "      <td>A</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id group  events\n",
       "0  16046     A      14\n",
       "1  18712     A      41\n",
       "2   3268     A      30\n",
       "3  22633     A      37\n",
       "4  28071     A      38"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Quick look to our data\n",
    "\n",
    "task_1.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step A"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Проверьте, являются ли дисперсии внутри групп примерно одинаковыми (гомогенными) с помощью теста Левена.\n",
    "#### Checki if our groups have homogen dispersion with Leven test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "gr_A = task_1.query('group == \"A\"').events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "gr_B = task_1.query('group == \"B\"').events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "gr_C = task_1.query('group == \"C\"').events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LeveneResult(statistic=2.2622596339318037, pvalue=0.1041327199523974)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# H0 - all input samples are from populations with equal variances\n",
    "# H1 - all input samples are from populations with diff variances\n",
    "\n",
    "ss.levene(gr_A, gr_B, gr_C)\n",
    "\n",
    "# pvalue=0.104 shows that we can't deny our H0 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### *Вывод* : дисперсии внутри групп являются гомогенными\n",
    "#### *Result* : dispersions are homogen"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step B\n",
    "\n",
    "### Checkig if our distributions are normal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NormaltestResult(statistic=2.1034599402238734, pvalue=0.3493328906085719)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Preparing our data and using our D’Agostino - Pearson test.\n",
    "# Используем тест Агостино-Пирсона для проверки наших распределений на нормальное.\n",
    "\n",
    "gr_A = task_1.query('group == \"A\"').events\n",
    "\n",
    "ss.stats.normaltest(gr_A)\n",
    "\n",
    "# H0 - our distribution is normal\n",
    "# H1 - our distribution is not normal\n",
    "\n",
    "# pvalue=0.349 shows that we can't deny our H0 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NormaltestResult(statistic=0.8410715800223014, pvalue=0.6566948749848232)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Preparing our data and using our D’Agostino - Pearson test.\n",
    "# Используем тест Агостино-Пирсона для проверки наших распределений на нормальное.\n",
    "\n",
    "gr_B = task_1.query('group == \"B\"').events\n",
    "\n",
    "ss.stats.normaltest(gr_B)\n",
    "\n",
    "# H0 - our distribution is normal\n",
    "# H1 - our distribution is not normal\n",
    "\n",
    "# pvalue=0.657 shows that we can't deny our H0  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NormaltestResult(statistic=3.243725809054571, pvalue=0.19753037588314626)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Preparing our data and using our D’Agostino - Pearson test.\n",
    "# Используем тест Агостино-Пирсона для проверки наших распределений на нормальное.\n",
    "\n",
    "gr_C = task_1.query('group == \"C\"').events\n",
    "\n",
    "ss.stats.normaltest(gr_C)\n",
    "\n",
    "# H0 - our distribution is normal\n",
    "# H1 - our distribution is not normal\n",
    "\n",
    "# pvalue=0.197 shows that we can't deny our H0 "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step C. Perform one-way ANOVA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "F_onewayResult(statistic=2886.166657438961, pvalue=0.0)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Для сравнения средних я использую ОДНОФАКТОРНЫЙ ДИСПЕРСИОННЫЙ АНАЛИЗ \n",
    "\n",
    "ss.stats.f_oneway(gr_A, gr_B, gr_C)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step D. Perform Tukey Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>mean(A)</th>\n",
       "      <th>mean(B)</th>\n",
       "      <th>diff</th>\n",
       "      <th>se</th>\n",
       "      <th>T</th>\n",
       "      <th>p-tukey</th>\n",
       "      <th>hedges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>29.5796</td>\n",
       "      <td>37.6136</td>\n",
       "      <td>-8.0340</td>\n",
       "      <td>0.140894</td>\n",
       "      <td>-57.021604</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-0.806377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A</td>\n",
       "      <td>C</td>\n",
       "      <td>29.5796</td>\n",
       "      <td>27.4703</td>\n",
       "      <td>2.1093</td>\n",
       "      <td>0.140894</td>\n",
       "      <td>14.970833</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.211712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>B</td>\n",
       "      <td>C</td>\n",
       "      <td>37.6136</td>\n",
       "      <td>27.4703</td>\n",
       "      <td>10.1433</td>\n",
       "      <td>0.140894</td>\n",
       "      <td>71.992436</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1.018089</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   A  B  mean(A)  mean(B)     diff        se          T  p-tukey    hedges\n",
       "0  A  B  29.5796  37.6136  -8.0340  0.140894 -57.021604    0.001 -0.806377\n",
       "1  A  C  29.5796  27.4703   2.1093  0.140894  14.970833    0.001  0.211712\n",
       "2  B  C  37.6136  27.4703  10.1433  0.140894  71.992436    0.001  1.018089"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Теперь используйте критерий Тьюки и определите, между какими группами есть статистически значимые различия.\n",
    "\n",
    "pg.pairwise_tukey(data=task_1, dv=\"events\", between=\"group\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step E. Perform Games-Showell Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>mean(A)</th>\n",
       "      <th>mean(B)</th>\n",
       "      <th>diff</th>\n",
       "      <th>se</th>\n",
       "      <th>T</th>\n",
       "      <th>df</th>\n",
       "      <th>pval</th>\n",
       "      <th>hedges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>B</td>\n",
       "      <td>29.5796</td>\n",
       "      <td>37.6136</td>\n",
       "      <td>-8.0340</td>\n",
       "      <td>0.140920</td>\n",
       "      <td>-57.011134</td>\n",
       "      <td>19987.258491</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-0.806229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A</td>\n",
       "      <td>C</td>\n",
       "      <td>29.5796</td>\n",
       "      <td>27.4703</td>\n",
       "      <td>2.1093</td>\n",
       "      <td>0.141696</td>\n",
       "      <td>14.886135</td>\n",
       "      <td>19995.116233</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.210514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>B</td>\n",
       "      <td>C</td>\n",
       "      <td>37.6136</td>\n",
       "      <td>27.4703</td>\n",
       "      <td>10.1433</td>\n",
       "      <td>0.140062</td>\n",
       "      <td>72.420220</td>\n",
       "      <td>19995.502469</td>\n",
       "      <td>0.001</td>\n",
       "      <td>1.024138</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   A  B  mean(A)  mean(B)     diff        se          T            df   pval  \\\n",
       "0  A  B  29.5796  37.6136  -8.0340  0.140920 -57.011134  19987.258491  0.001   \n",
       "1  A  C  29.5796  27.4703   2.1093  0.141696  14.886135  19995.116233  0.001   \n",
       "2  B  C  37.6136  27.4703  10.1433  0.140062  72.420220  19995.502469  0.001   \n",
       "\n",
       "     hedges  \n",
       "0 -0.806229  \n",
       "1  0.210514  \n",
       "2  1.024138  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pg.pairwise_gameshowell(data=task_1, dv=\"events\", between=\"group\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Finall Step. \n",
    "\n",
    "## Result: As we can see in our tests, difference between our groups is statisticaly significant and we should pick group \"B\". "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>events</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>group</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>A</th>\n",
       "      <td>295796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>B</th>\n",
       "      <td>376136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>C</th>\n",
       "      <td>274703</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       events\n",
       "group        \n",
       "A      295796\n",
       "B      376136\n",
       "C      274703"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "task_1.groupby('group').agg({'events':'sum'})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Task 2 \n",
    "\n",
    "### Во втором эксперименте требуется проверить, как пользователи отреагируют на изменение формата кнопки оформления заказа, с разбивкой по сегменту клиента. Какой вариант кнопки нам нужно оставить в приложении. \n",
    "\n",
    "##### TEST group - new interface. CONTROL - old interface.\n",
    "\n",
    "### In this task we should say if difference between our groups is statisticaly significant and what format of our \"ORDER\" button should we apply. \n",
    "\n",
    "#####  TEST group - новый вариант кнопки. CONTROL - старый вариант."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>id</th>\n",
       "      <th>group</th>\n",
       "      <th>segment</th>\n",
       "      <th>events</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>83440</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>78.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>96723</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>71.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68719</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>80.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>50399</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>80.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>95613</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>78.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id group segment  events\n",
       "0  83440  test    high    78.0\n",
       "1  96723  test    high    71.0\n",
       "2  68719  test    high    80.0\n",
       "3  50399  test    high    80.0\n",
       "4  95613  test    high    78.0"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Quick look to our data\n",
    "\n",
    "task_2.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step A. Visualising our distributions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# At first we should filter by our experimental group type and then group by region and finaly creating our histogram by delivery time.\n",
    "# Для начала отфильтруем по контрольной группе, сгруппируем по району города и затем построим гисторграмму по веремени доставки. \n",
    "\n",
    "test_group = task_2.query('group ==\"test\"').groupby('segment')['events'].hist(bins=20)\n",
    "\n",
    "# Our hist for \"test\" group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# At first we should filter by our experimental group type and then group by region and finaly creating our histogram by delivery time.\n",
    "# Для начала отфильтруем по контрольной группе, сгруппируем по району города и затем построим гисторграмму по веремени доставки. \n",
    "\n",
    "control_group = task_2.query('group ==\"control\"').groupby('segment')['events'].hist(bins=20)\n",
    "\n",
    "# Our hist for \"control\" group"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step B. Calculating 'std', 'mean', 'median' by our segments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>segment</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>median</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>high</td>\n",
       "      <td>7.805347</td>\n",
       "      <td>59.970800</td>\n",
       "      <td>60.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>low</td>\n",
       "      <td>4.473576</td>\n",
       "      <td>20.031575</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  segment       std       mean  median\n",
       "0    high  7.805347  59.970800    60.0\n",
       "1     low  4.473576  20.031575    20.0"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "task_2.query('group ==\"control\"').groupby('segment')['events'].agg(['std', 'mean', 'median']).reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step C. Perform ANOVA Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>Source</th>\n",
       "      <th>SS</th>\n",
       "      <th>DF</th>\n",
       "      <th>MS</th>\n",
       "      <th>F</th>\n",
       "      <th>p-unc</th>\n",
       "      <th>np2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>group</td>\n",
       "      <td>3.837195e+06</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.837195e+06</td>\n",
       "      <td>112031.864119</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.528383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>segment</td>\n",
       "      <td>2.362480e+07</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.362480e+07</td>\n",
       "      <td>689756.377484</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.873383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>group * segment</td>\n",
       "      <td>3.664244e+04</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.664244e+04</td>\n",
       "      <td>1069.823273</td>\n",
       "      <td>2.060261e-233</td>\n",
       "      <td>0.010585</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Residual</td>\n",
       "      <td>3.424956e+06</td>\n",
       "      <td>99996.0</td>\n",
       "      <td>3.425093e+01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Source            SS       DF            MS              F  \\\n",
       "0            group  3.837195e+06      1.0  3.837195e+06  112031.864119   \n",
       "1          segment  2.362480e+07      1.0  2.362480e+07  689756.377484   \n",
       "2  group * segment  3.664244e+04      1.0  3.664244e+04    1069.823273   \n",
       "3         Residual  3.424956e+06  99996.0  3.425093e+01            NaN   \n",
       "\n",
       "           p-unc       np2  \n",
       "0   0.000000e+00  0.528383  \n",
       "1   0.000000e+00  0.873383  \n",
       "2  2.060261e-233  0.010585  \n",
       "3            NaN       NaN  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Want to check the effect of two veriables - group и segment – for events, with factors interaction.\n",
    "# хотим проверить эффект нескольких переменных – group и segment – на events, с взаимодействием факторов.\n",
    "\n",
    "pg.anova(data=task_2, dv=\"events\", between=[\"group\", \"segment\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step D. Perform Tukey Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>mean(A)</th>\n",
       "      <th>mean(B)</th>\n",
       "      <th>diff</th>\n",
       "      <th>se</th>\n",
       "      <th>T</th>\n",
       "      <th>p-tukey</th>\n",
       "      <th>hedges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>control</td>\n",
       "      <td>test</td>\n",
       "      <td>28.01942</td>\n",
       "      <td>40.40844</td>\n",
       "      <td>-12.38902</td>\n",
       "      <td>0.10409</td>\n",
       "      <td>-119.021936</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-0.752755</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         A     B   mean(A)   mean(B)      diff       se           T  p-tukey  \\\n",
       "0  control  test  28.01942  40.40844 -12.38902  0.10409 -119.021936    0.001   \n",
       "\n",
       "     hedges  \n",
       "0 -0.752755  "
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Performing Tukey test to check if our diff between groups stats significant.\n",
    "\n",
    "pg.pairwise_tukey(data=task_2, dv=\"events\", between=\"group\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>group</th>\n",
       "      <th>sum</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>control</td>\n",
       "      <td>1400971.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test</td>\n",
       "      <td>2020422.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     group        sum\n",
       "0  control  1400971.0\n",
       "1     test  2020422.0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Understanding the diff\n",
    "\n",
    "task_2.groupby('group')['events'].agg(['sum']).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>segment</th>\n",
       "      <th>group</th>\n",
       "      <th>sum</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>high</td>\n",
       "      <td>control</td>\n",
       "      <td>599708.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>high</td>\n",
       "      <td>test</td>\n",
       "      <td>699385.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>low</td>\n",
       "      <td>control</td>\n",
       "      <td>801263.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>low</td>\n",
       "      <td>test</td>\n",
       "      <td>1321037.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  segment    group        sum\n",
       "0    high  control   599708.0\n",
       "1    high     test   699385.0\n",
       "2     low  control   801263.0\n",
       "3     low     test  1321037.0"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Understanding the diff\n",
    "\n",
    "task_2.groupby(['segment', 'group'])['events'].agg(['sum']).reset_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Step E. FINALL TUKEY TEST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Making it more convinient to compare and creatin new column by concatenating column \"group\" and \"segment\"\n",
    "# В данном случае для последующих сравнений между группами можно соединить колонки group и segment в combination, чтобы потом передать в аргумент groups в наших тестах.\n",
    "\n",
    "task_2['combination'] = task_2.group + ' // ' + task_2.segment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>id</th>\n",
       "      <th>group</th>\n",
       "      <th>segment</th>\n",
       "      <th>events</th>\n",
       "      <th>combination</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>83440</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>78.0</td>\n",
       "      <td>test // high</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>96723</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>71.0</td>\n",
       "      <td>test // high</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68719</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>80.0</td>\n",
       "      <td>test // high</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>50399</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>80.0</td>\n",
       "      <td>test // high</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>95613</td>\n",
       "      <td>test</td>\n",
       "      <td>high</td>\n",
       "      <td>78.0</td>\n",
       "      <td>test // high</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id group segment  events   combination\n",
       "0  83440  test    high    78.0  test // high\n",
       "1  96723  test    high    71.0  test // high\n",
       "2  68719  test    high    80.0  test // high\n",
       "3  50399  test    high    80.0  test // high\n",
       "4  95613  test    high    78.0  test // high"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking if new column looks good\n",
    "# Посмотрим все ли у нас хорошо получилось\n",
    "\n",
    "task_2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>A</th>\n",
       "      <th>B</th>\n",
       "      <th>mean(A)</th>\n",
       "      <th>mean(B)</th>\n",
       "      <th>diff</th>\n",
       "      <th>se</th>\n",
       "      <th>T</th>\n",
       "      <th>p-tukey</th>\n",
       "      <th>hedges</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>control // high</td>\n",
       "      <td>control // low</td>\n",
       "      <td>59.970800</td>\n",
       "      <td>20.031575</td>\n",
       "      <td>39.939225</td>\n",
       "      <td>0.065432</td>\n",
       "      <td>610.391461</td>\n",
       "      <td>0.001</td>\n",
       "      <td>6.824282</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>control // high</td>\n",
       "      <td>test // high</td>\n",
       "      <td>59.970800</td>\n",
       "      <td>69.938500</td>\n",
       "      <td>-9.967700</td>\n",
       "      <td>0.082766</td>\n",
       "      <td>-120.432523</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-1.703109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>control // high</td>\n",
       "      <td>test // low</td>\n",
       "      <td>59.970800</td>\n",
       "      <td>33.025925</td>\n",
       "      <td>26.944875</td>\n",
       "      <td>0.065432</td>\n",
       "      <td>411.798717</td>\n",
       "      <td>0.001</td>\n",
       "      <td>4.603981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>control // low</td>\n",
       "      <td>test // high</td>\n",
       "      <td>20.031575</td>\n",
       "      <td>69.938500</td>\n",
       "      <td>-49.906925</td>\n",
       "      <td>0.065432</td>\n",
       "      <td>-762.727892</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-8.527429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>control // low</td>\n",
       "      <td>test // low</td>\n",
       "      <td>20.031575</td>\n",
       "      <td>33.025925</td>\n",
       "      <td>-12.994350</td>\n",
       "      <td>0.041383</td>\n",
       "      <td>-314.002699</td>\n",
       "      <td>0.001</td>\n",
       "      <td>-2.220314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>test // high</td>\n",
       "      <td>test // low</td>\n",
       "      <td>69.938500</td>\n",
       "      <td>33.025925</td>\n",
       "      <td>36.912575</td>\n",
       "      <td>0.065432</td>\n",
       "      <td>564.135148</td>\n",
       "      <td>0.001</td>\n",
       "      <td>6.307128</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 A               B    mean(A)    mean(B)       diff        se  \\\n",
       "0  control // high  control // low  59.970800  20.031575  39.939225  0.065432   \n",
       "1  control // high    test // high  59.970800  69.938500  -9.967700  0.082766   \n",
       "2  control // high     test // low  59.970800  33.025925  26.944875  0.065432   \n",
       "3   control // low    test // high  20.031575  69.938500 -49.906925  0.065432   \n",
       "4   control // low     test // low  20.031575  33.025925 -12.994350  0.041383   \n",
       "5     test // high     test // low  69.938500  33.025925  36.912575  0.065432   \n",
       "\n",
       "            T  p-tukey    hedges  \n",
       "0  610.391461    0.001  6.824282  \n",
       "1 -120.432523    0.001 -1.703109  \n",
       "2  411.798717    0.001  4.603981  \n",
       "3 -762.727892    0.001 -8.527429  \n",
       "4 -314.002699    0.001 -2.220314  \n",
       "5  564.135148    0.001  6.307128  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Performing Tukey test to help us make finall descision.  \n",
    "\n",
    "pg.pairwise_tukey(data=task_2, dv=\"events\", between=\"combination\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7ffb0a939588>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualising our test\n",
    "\n",
    "sns.pointplot(x = 'segment', y = 'events', hue = 'group', data = task_2, capsize = .2)\n",
    "plt.title('NUmber of events in connection with segment and grop')\n",
    "plt.xlabel('Segment')\n",
    "plt.ylabel('Events')\n",
    "plt.legend(title = 'Group type')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## FINAL RESULT\n",
    "\n",
    "## Factor 'group' is significant, factor 'segmet' is significant and their cooperation is significant. According to our pointplot means in \"Test group\" increased and that is way we should go for NEW (test) version.\n",
    "\n",
    "## Фактор group оказался ЗНАЧИМЫМ, фактор segment – ЗНАЧИМЫМ, их взаимодействие – ЗНАЧИМО. Судя по графику, для всех групп среднее значение events УВЕЛИЧИЛОСЬ, поэтому решение: ВЫКАТЫВАЕМ НОВУЮ ВЕРСИЮ."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}