project_9_ab_test_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "5542b5fb",
   "metadata": {},
   "source": [
    "# Подготовка и анализ A/B-теста для интернет-магазина"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b6f8d22",
   "metadata": {},
   "source": [
    "**Описание проекта**\n",
    "\n",
    "Цель проекта — помочь отделу маркетинга крупного интернет-магазина увеличить выручку, проверив гипотезы с помощью A/B-теста.\n",
    "\n",
    "Необходимо:\n",
    "- приоритизировать гипотезы;\n",
    "- запустить A/B-тест;\n",
    "- проанализировать результаты.\n",
    "\n",
    "**Описание данных**\n",
    "\n",
    "Понадобятся 3 датасета:\n",
    "- файл с гипотезами по увеличению выручки интернет-магазина `hypothesis.csv`,\n",
    "- результаты A/B-теста по заказам `orders.csv`,\n",
    "- результаты A/B-теста по пользователям `visitors.csv`.\n",
    "\n",
    "Структура `hypothesis.csv`:\n",
    "> - `Hypothesis` — краткое описание гипотезы;\n",
    "> - `Reach` — охват пользователей по 10-балльной шкале;\n",
    "> - `Impact` — влияние на пользователей по 10-балльной шкале;\n",
    "> - `Confidence` — уверенность в гипотезе по 10-балльной шкале;\n",
    "> - `Efforts` —  затраты ресурсов на проверку гипотезы по 10-балльной шкале.\n",
    "\n",
    "Структура `orders.csv`:\n",
    "> - `transactionId` — идентификатор заказа;\n",
    "> - `visitorId` — идентификатор пользователя, совершившего заказ;\n",
    "> - `date` — дата, когда был совершён заказ;\n",
    "> - `revenue` — выручка заказа;\n",
    "> - `group` — группа A/B-теста, в которую попал заказ.\n",
    "\n",
    "Структура `visitors.csv`:\n",
    "> - `date` — дата;\n",
    "> - `group` — группа A/B-теста;\n",
    "> - `visitors` — количество пользователей в указанную дату в указанной группе A/B-теста."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec48a2b0",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Приоритизация-гипотез\" data-toc-modified-id=\"Приоритизация-гипотез-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Приоритизация гипотез</a></span><ul class=\"toc-item\"><li><span><a href=\"#Чтение-данных\" data-toc-modified-id=\"Чтение-данных-1.1\"><span class=\"toc-item-num\">1.1&nbsp;&nbsp;</span>Чтение данных</a></span></li><li><span><a href=\"#Фреймворк-ICE\" data-toc-modified-id=\"Фреймворк-ICE-1.2\"><span class=\"toc-item-num\">1.2&nbsp;&nbsp;</span>Фреймворк ICE</a></span></li><li><span><a href=\"#Фреймворк-RICE\" data-toc-modified-id=\"Фреймворк-RICE-1.3\"><span class=\"toc-item-num\">1.3&nbsp;&nbsp;</span>Фреймворк RICE</a></span></li><li><span><a href=\"#Вывод\" data-toc-modified-id=\"Вывод-1.4\"><span class=\"toc-item-num\">1.4&nbsp;&nbsp;</span>Вывод</a></span></li></ul></li><li><span><a href=\"#Анализ-A/B-теста\" data-toc-modified-id=\"Анализ-A/B-теста-2\"><span class=\"toc-item-num\">2&nbsp;&nbsp;</span>Анализ A/B-теста</a></span><ul class=\"toc-item\"><li><span><a href=\"#Чтение-данных\" data-toc-modified-id=\"Чтение-данных-2.1\"><span class=\"toc-item-num\">2.1&nbsp;&nbsp;</span>Чтение данных</a></span></li><li><span><a href=\"#Предобработка-данных\" data-toc-modified-id=\"Предобработка-данных-2.2\"><span class=\"toc-item-num\">2.2&nbsp;&nbsp;</span>Предобработка данных</a></span></li><li><span><a href=\"#Анализ-данных\" data-toc-modified-id=\"Анализ-данных-2.3\"><span class=\"toc-item-num\">2.3&nbsp;&nbsp;</span>Анализ данных</a></span><ul class=\"toc-item\"><li><span><a href=\"#График-кумулятивной-выручки-по-группам\" data-toc-modified-id=\"График-кумулятивной-выручки-по-группам-2.3.1\"><span class=\"toc-item-num\">2.3.1&nbsp;&nbsp;</span>График кумулятивной выручки по группам</a></span></li><li><span><a href=\"#График-кумулятивного-среднего-чека-по-группам\" data-toc-modified-id=\"График-кумулятивного-среднего-чека-по-группам-2.3.2\"><span class=\"toc-item-num\">2.3.2&nbsp;&nbsp;</span>График кумулятивного среднего чека по группам</a></span></li><li><span><a href=\"#График-относительного-изменения-кумулятивного-среднего-чека-группы-B-к-группе-A\" data-toc-modified-id=\"График-относительного-изменения-кумулятивного-среднего-чека-группы-B-к-группе-A-2.3.3\"><span class=\"toc-item-num\">2.3.3&nbsp;&nbsp;</span>График относительного изменения кумулятивного среднего чека группы B к группе A</a></span></li><li><span><a href=\"#График-кумулятивного-среднего-количества-заказов-на-посетителя-по-группам\" data-toc-modified-id=\"График-кумулятивного-среднего-количества-заказов-на-посетителя-по-группам-2.3.4\"><span class=\"toc-item-num\">2.3.4&nbsp;&nbsp;</span>График кумулятивного среднего количества заказов на посетителя по группам</a></span></li><li><span><a href=\"#График-относительного-изменения-кумулятивного-среднего-количества-заказов-на-посетителя-группы-B-к-группе-A\" data-toc-modified-id=\"График-относительного-изменения-кумулятивного-среднего-количества-заказов-на-посетителя-группы-B-к-группе-A-2.3.5\"><span class=\"toc-item-num\">2.3.5&nbsp;&nbsp;</span>График относительного изменения кумулятивного среднего количества заказов на посетителя группы B к группе A</a></span></li><li><span><a href=\"#Точечный-график-количества-заказов-по-пользователям\" data-toc-modified-id=\"Точечный-график-количества-заказов-по-пользователям-2.3.6\"><span class=\"toc-item-num\">2.3.6&nbsp;&nbsp;</span>Точечный график количества заказов по пользователям</a></span></li><li><span><a href=\"#95-й-и-99-й-перцентили-количества-заказов-на-пользователя.-Граница-для-определения-аномальных-пользователей\" data-toc-modified-id=\"95-й-и-99-й-перцентили-количества-заказов-на-пользователя.-Граница-для-определения-аномальных-пользователей-2.3.7\"><span class=\"toc-item-num\">2.3.7&nbsp;&nbsp;</span>95-й и 99-й перцентили количества заказов на пользователя. Граница для определения аномальных пользователей</a></span></li><li><span><a href=\"#Точечный-график-стоимостей-заказов\" data-toc-modified-id=\"Точечный-график-стоимостей-заказов-2.3.8\"><span class=\"toc-item-num\">2.3.8&nbsp;&nbsp;</span>Точечный график стоимостей заказов</a></span></li><li><span><a href=\"#95-й-и-99-й-перцентили-стоимости-заказов.-Граница-для-определения-аномальных-заказов\" data-toc-modified-id=\"95-й-и-99-й-перцентили-стоимости-заказов.-Граница-для-определения-аномальных-заказов-2.3.9\"><span class=\"toc-item-num\">2.3.9&nbsp;&nbsp;</span>95-й и 99-й перцентили стоимости заказов. Граница для определения аномальных заказов</a></span></li><li><span><a href=\"#Статистическая-значимость-различий-в-среднем-количестве-заказов-на-посетителя-между-группами-по-&quot;сырым&quot;-данным\" data-toc-modified-id=\"Статистическая-значимость-различий-в-среднем-количестве-заказов-на-посетителя-между-группами-по-&quot;сырым&quot;-данным-2.3.10\"><span class=\"toc-item-num\">2.3.10&nbsp;&nbsp;</span>Статистическая значимость различий в среднем количестве заказов на посетителя между группами по \"сырым\" данным</a></span></li><li><span><a href=\"#Статистическая-значимость-различий-в-среднем-чеке-заказа-между-группами-по-&quot;сырым&quot;-данным\" data-toc-modified-id=\"Статистическая-значимость-различий-в-среднем-чеке-заказа-между-группами-по-&quot;сырым&quot;-данным-2.3.11\"><span class=\"toc-item-num\">2.3.11&nbsp;&nbsp;</span>Статистическая значимость различий в среднем чеке заказа между группами по \"сырым\" данным</a></span></li><li><span><a href=\"#Статистическая-значимость-различий-в-среднем-количестве-заказов-на-посетителя-между-группами-по-очищенным-данным\" data-toc-modified-id=\"Статистическая-значимость-различий-в-среднем-количестве-заказов-на-посетителя-между-группами-по-очищенным-данным-2.3.12\"><span class=\"toc-item-num\">2.3.12&nbsp;&nbsp;</span>Статистическая значимость различий в среднем количестве заказов на посетителя между группами по очищенным данным</a></span></li><li><span><a href=\"#Статистическая-значимость-различий-в-среднем-чеке-заказа-между-группами-по-очищенным-данным\" data-toc-modified-id=\"Статистическая-значимость-различий-в-среднем-чеке-заказа-между-группами-по-очищенным-данным-2.3.13\"><span class=\"toc-item-num\">2.3.13&nbsp;&nbsp;</span>Статистическая значимость различий в среднем чеке заказа между группами по очищенным данным</a></span></li></ul></li><li><span><a href=\"#Вывод\" data-toc-modified-id=\"Вывод-2.4\"><span class=\"toc-item-num\">2.4&nbsp;&nbsp;</span>Вывод</a></span></li></ul></li><li><span><a href=\"#Общий-вывод\" data-toc-modified-id=\"Общий-вывод-3\"><span class=\"toc-item-num\">3&nbsp;&nbsp;</span>Общий вывод</a></span></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7a9ac2e",
   "metadata": {},
   "source": [
    "## Приоритизация гипотез"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6feaf56",
   "metadata": {},
   "source": [
    "В файле `hypothesis.csv` подготовлено 9 гипотез по увеличению выручки интернет-магазина с указанными параметрами `Reach`, `Impact`, `Confidence`, и `Effort`.\n",
    "\n",
    "Перед проведением A/B-теста приоритизируем гипотезы, используя фреймфорки **ICE** и **RICE**."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d6ff44b6",
   "metadata": {},
   "source": [
    "### Чтение данных"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecc182b4",
   "metadata": {},
   "source": [
    "Прочитаем файл с гипотезами."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2b069c43",
   "metadata": {},
   "outputs": [],
   "source": [
    "# импорт библиотек\n",
    "\n",
    "import pandas as pd\n",
    "import scipy.stats as stats\n",
    "import datetime as dt\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from pandas.plotting import register_matplotlib_converters\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "45046764",
   "metadata": {},
   "outputs": [],
   "source": [
    "# чтение файла\n",
    "\n",
    "try:\n",
    "    hypothesis = pd.read_csv('/Users/lina/Documents/УЧЁБА/Яндекс Практикум/datasets/hypothesis.csv')\n",
    "except:\n",
    "    hypothesis = pd.read_csv('https://code.s3.yandex.net/datasets/hypothesis.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "617b5653",
   "metadata": {},
   "source": [
    "### Фреймворк ICE"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c29dd71b",
   "metadata": {},
   "source": [
    "Посчитаем оценку ICE по формуле: `ICE score = Impact * Confidence / Efforts`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6e5a1b52",
   "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>Hypothesis</th>\n",
       "      <th>ICE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Запустить акцию, дающую скидку на товар в день рождения</td>\n",
       "      <td>16.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Добавить два новых канала привлечения трафика, что позволит привлекать на 30% больше пользователей</td>\n",
       "      <td>13.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок</td>\n",
       "      <td>11.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Показать на главной странице баннеры с актуальными акциями и распродажами, чтобы увеличить конверсию</td>\n",
       "      <td>8.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Добавить блоки рекомендаций товаров на сайт интернет магазина, чтобы повысить конверсию и средний чек заказа</td>\n",
       "      <td>7.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Запустить собственную службу доставки, что сократит срок доставки заказов</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Добавить страницу отзывов клиентов о магазине, что позволит увеличить количество заказов</td>\n",
       "      <td>1.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Изменить структура категорий, что увеличит конверсию, т.к. пользователи быстрее найдут нужный товар</td>\n",
       "      <td>1.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Изменить цвет фона главной страницы, чтобы увеличить вовлеченность пользователей</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                     Hypothesis  \\\n",
       "8                                                       Запустить акцию, дающую скидку на товар в день рождения   \n",
       "0            Добавить два новых канала привлечения трафика, что позволит привлекать на 30% больше пользователей   \n",
       "7              Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок   \n",
       "6          Показать на главной странице баннеры с актуальными акциями и распродажами, чтобы увеличить конверсию   \n",
       "2  Добавить блоки рекомендаций товаров на сайт интернет магазина, чтобы повысить конверсию и средний чек заказа   \n",
       "1                                     Запустить собственную службу доставки, что сократит срок доставки заказов   \n",
       "5                      Добавить страницу отзывов клиентов о магазине, что позволит увеличить количество заказов   \n",
       "3           Изменить структура категорий, что увеличит конверсию, т.к. пользователи быстрее найдут нужный товар   \n",
       "4                              Изменить цвет фона главной страницы, чтобы увеличить вовлеченность пользователей   \n",
       "\n",
       "         ICE  \n",
       "8  16.200000  \n",
       "0  13.333333  \n",
       "7  11.200000  \n",
       "6   8.000000  \n",
       "2   7.000000  \n",
       "1   2.000000  \n",
       "5   1.333333  \n",
       "3   1.125000  \n",
       "4   1.000000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ICE score\n",
    "\n",
    "hypothesis['ICE'] = hypothesis['Impact'] * hypothesis['Confidence'] / hypothesis['Efforts']\n",
    "\n",
    "pd.set_option('max_colwidth', 150)\n",
    "hypothesis[['Hypothesis', 'ICE']].sort_values(by='ICE', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83d9b6dc",
   "metadata": {},
   "source": [
    "По фреймворку ICE наиболее приоритетной будет гипотеза:\n",
    "\n",
    "*Запустить акцию, дающую скидку на товар в день рождения*."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2092b80e",
   "metadata": {},
   "source": [
    "### Фреймворк RICE"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4990de1e",
   "metadata": {},
   "source": [
    "Посчитаем оценку RICE по формуле: `RICE score = Reach * Impact * Confidence / Efforts`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c8eea143",
   "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>Hypothesis</th>\n",
       "      <th>RICE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок</td>\n",
       "      <td>112.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Добавить блоки рекомендаций товаров на сайт интернет магазина, чтобы повысить конверсию и средний чек заказа</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Добавить два новых канала привлечения трафика, что позволит привлекать на 30% больше пользователей</td>\n",
       "      <td>40.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Показать на главной странице баннеры с актуальными акциями и распродажами, чтобы увеличить конверсию</td>\n",
       "      <td>40.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Запустить акцию, дающую скидку на товар в день рождения</td>\n",
       "      <td>16.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Изменить структура категорий, что увеличит конверсию, т.к. пользователи быстрее найдут нужный товар</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Запустить собственную службу доставки, что сократит срок доставки заказов</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Добавить страницу отзывов клиентов о магазине, что позволит увеличить количество заказов</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Изменить цвет фона главной страницы, чтобы увеличить вовлеченность пользователей</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                     Hypothesis  \\\n",
       "7              Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок   \n",
       "2  Добавить блоки рекомендаций товаров на сайт интернет магазина, чтобы повысить конверсию и средний чек заказа   \n",
       "0            Добавить два новых канала привлечения трафика, что позволит привлекать на 30% больше пользователей   \n",
       "6          Показать на главной странице баннеры с актуальными акциями и распродажами, чтобы увеличить конверсию   \n",
       "8                                                       Запустить акцию, дающую скидку на товар в день рождения   \n",
       "3           Изменить структура категорий, что увеличит конверсию, т.к. пользователи быстрее найдут нужный товар   \n",
       "1                                     Запустить собственную службу доставки, что сократит срок доставки заказов   \n",
       "5                      Добавить страницу отзывов клиентов о магазине, что позволит увеличить количество заказов   \n",
       "4                              Изменить цвет фона главной страницы, чтобы увеличить вовлеченность пользователей   \n",
       "\n",
       "    RICE  \n",
       "7  112.0  \n",
       "2   56.0  \n",
       "0   40.0  \n",
       "6   40.0  \n",
       "8   16.2  \n",
       "3    9.0  \n",
       "1    4.0  \n",
       "5    4.0  \n",
       "4    3.0  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# RICE score\n",
    "\n",
    "hypothesis['RICE'] = hypothesis['Reach'] * hypothesis['Impact'] * hypothesis['Confidence']\\\n",
    "                     / hypothesis['Efforts']\n",
    "\n",
    "hypothesis[['Hypothesis', 'RICE']].sort_values(by='RICE', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3417723",
   "metadata": {},
   "source": [
    "По фреймворку RICE наиболее приоритетной будет гипотеза: \n",
    "\n",
    "*Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок*.\n",
    "\n",
    "Визуализируем, как изменились приоритеты при учёте параметра охвата пользователей (*Reach*) в фреймворке RICE, в единой таблице с заливкой."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ed76c370",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "#T_0ee8f_row0_col1{\n",
       "            background-color:  #127c39;\n",
       "            color:  #f1f1f1;\n",
       "        }#T_0ee8f_row0_col2,#T_0ee8f_row6_col2{\n",
       "            background-color:  #acdea6;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row1_col1{\n",
       "            background-color:  #eef8ea;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row1_col2,#T_0ee8f_row3_col1,#T_0ee8f_row5_col2{\n",
       "            background-color:  #f6fcf4;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row2_col1{\n",
       "            background-color:  #99d595;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row2_col2{\n",
       "            background-color:  #79c67a;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row3_col2{\n",
       "            background-color:  #eff9ec;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row4_col1,#T_0ee8f_row4_col2{\n",
       "            background-color:  #f7fcf5;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row5_col1{\n",
       "            background-color:  #f4fbf2;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row6_col1{\n",
       "            background-color:  #83cb82;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row7_col1{\n",
       "            background-color:  #369f54;\n",
       "            color:  #000000;\n",
       "        }#T_0ee8f_row7_col2,#T_0ee8f_row8_col1{\n",
       "            background-color:  #00441b;\n",
       "            color:  #f1f1f1;\n",
       "        }#T_0ee8f_row8_col2{\n",
       "            background-color:  #e5f5e1;\n",
       "            color:  #000000;\n",
       "        }</style><table id=\"T_0ee8f_\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" >Hypothesis</th>        <th class=\"col_heading level0 col1\" >ICE</th>        <th class=\"col_heading level0 col2\" >RICE</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "                        <td id=\"T_0ee8f_row0_col0\" class=\"data row0 col0\" >Добавить два новых канала привлечения трафика, что позволит привлекать на 30% больше пользователей</td>\n",
       "                        <td id=\"T_0ee8f_row0_col1\" class=\"data row0 col1\" >13.333333</td>\n",
       "                        <td id=\"T_0ee8f_row0_col2\" class=\"data row0 col2\" >40.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "                        <td id=\"T_0ee8f_row1_col0\" class=\"data row1 col0\" >Запустить собственную службу доставки, что сократит срок доставки заказов</td>\n",
       "                        <td id=\"T_0ee8f_row1_col1\" class=\"data row1 col1\" >2.000000</td>\n",
       "                        <td id=\"T_0ee8f_row1_col2\" class=\"data row1 col2\" >4.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "                        <td id=\"T_0ee8f_row2_col0\" class=\"data row2 col0\" >Добавить блоки рекомендаций товаров на сайт интернет магазина, чтобы повысить конверсию и средний чек заказа</td>\n",
       "                        <td id=\"T_0ee8f_row2_col1\" class=\"data row2 col1\" >7.000000</td>\n",
       "                        <td id=\"T_0ee8f_row2_col2\" class=\"data row2 col2\" >56.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "                        <td id=\"T_0ee8f_row3_col0\" class=\"data row3 col0\" >Изменить структура категорий, что увеличит конверсию, т.к. пользователи быстрее найдут нужный товар</td>\n",
       "                        <td id=\"T_0ee8f_row3_col1\" class=\"data row3 col1\" >1.125000</td>\n",
       "                        <td id=\"T_0ee8f_row3_col2\" class=\"data row3 col2\" >9.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "                        <td id=\"T_0ee8f_row4_col0\" class=\"data row4 col0\" >Изменить цвет фона главной страницы, чтобы увеличить вовлеченность пользователей</td>\n",
       "                        <td id=\"T_0ee8f_row4_col1\" class=\"data row4 col1\" >1.000000</td>\n",
       "                        <td id=\"T_0ee8f_row4_col2\" class=\"data row4 col2\" >3.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "                        <td id=\"T_0ee8f_row5_col0\" class=\"data row5 col0\" >Добавить страницу отзывов клиентов о магазине, что позволит увеличить количество заказов</td>\n",
       "                        <td id=\"T_0ee8f_row5_col1\" class=\"data row5 col1\" >1.333333</td>\n",
       "                        <td id=\"T_0ee8f_row5_col2\" class=\"data row5 col2\" >4.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "                        <td id=\"T_0ee8f_row6_col0\" class=\"data row6 col0\" >Показать на главной странице баннеры с актуальными акциями и распродажами, чтобы увеличить конверсию</td>\n",
       "                        <td id=\"T_0ee8f_row6_col1\" class=\"data row6 col1\" >8.000000</td>\n",
       "                        <td id=\"T_0ee8f_row6_col2\" class=\"data row6 col2\" >40.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "                        <td id=\"T_0ee8f_row7_col0\" class=\"data row7 col0\" >Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок</td>\n",
       "                        <td id=\"T_0ee8f_row7_col1\" class=\"data row7 col1\" >11.200000</td>\n",
       "                        <td id=\"T_0ee8f_row7_col2\" class=\"data row7 col2\" >112.000000</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_0ee8f_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "                        <td id=\"T_0ee8f_row8_col0\" class=\"data row8 col0\" >Запустить акцию, дающую скидку на товар в день рождения</td>\n",
       "                        <td id=\"T_0ee8f_row8_col1\" class=\"data row8 col1\" >16.200000</td>\n",
       "                        <td id=\"T_0ee8f_row8_col2\" class=\"data row8 col2\" >16.200000</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7fbbf0147760>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# выведем гипотезы с заливкой самых высоких баллов\n",
    "\n",
    "(\n",
    "    hypothesis[['Hypothesis', 'ICE', 'RICE']]\n",
    "    .style.background_gradient(subset=['ICE', 'RICE'], cmap='Greens')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b79fbdd",
   "metadata": {},
   "source": [
    "Также можем построить slope graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3131428f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# добавим столбцы с номером приоритета для визуализации изменений\n",
    "\n",
    "hypothesis['ICE_rank'] = hypothesis['ICE'].rank(method='min', ascending=False)\n",
    "hypothesis['RICE_rank'] = hypothesis['RICE'].rank(method='min', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "82694426",
   "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": [
    "# создаём список со списками-парами значений ICE-RICE\n",
    "\n",
    "list_of_values = []\n",
    "for index, row in hypothesis.iterrows():\n",
    "    temp_list = []\n",
    "    temp_list.append(row['ICE_rank'])\n",
    "    temp_list.append(row['RICE_rank'])\n",
    "    list_of_values.append(temp_list)\n",
    "\n",
    "# регулярное выражение, чтобы в лейблах сократить по нему все, что после \",\" в названии гипотезы\n",
    "\n",
    "regex = re.compile(r',.*')\n",
    "\n",
    "# строим графики\n",
    "\n",
    "fig, ax = plt.subplots()    \n",
    "    \n",
    "for i in range(0, len(list_of_values)):\n",
    "    h_label = regex.sub('', hypothesis.loc[i, 'Hypothesis'])\n",
    "    ax.plot(['ICE', 'RICE'], list_of_values[i], marker='o', label=h_label)\n",
    "\n",
    "legend = fig.legend(loc='right', bbox_to_anchor=(1.8, 0.7))\n",
    "plt.subplots_adjust(right=0.7)\n",
    "\n",
    "plt.xlabel('фреймворк')\n",
    "plt.ylabel('приоритет')\n",
    "plt.title('Слоупграф изменений приоритета гипотез') \n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbaae5ba",
   "metadata": {},
   "source": [
    "### Вывод"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4343a795",
   "metadata": {},
   "source": [
    "Мы применили фреймворки ICE и RICE для приоритизации 9 гипотез по увеличению выручки интернет-магазина.\n",
    "\n",
    "Результаты оценок различаются, так как при учёте параметра охвата пользователей (*Reach*) у некоторых гипотез получается более высокий итоговый балл."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4529fd1c",
   "metadata": {},
   "source": [
    "## Анализ A/B-теста"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6f0a10e",
   "metadata": {},
   "source": [
    "Мы провели A/B-тест и получили результаты, которые описаны в файлах `orders.csv` и `visitors.csv`.\n",
    "\n",
    "В этом разделе мы проведём анализ A/B-теста."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83ef7f74",
   "metadata": {},
   "source": [
    "### Чтение данных"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27783bd9",
   "metadata": {},
   "source": [
    "Прочитаем данные, а также выведем общую информацию о датафреймах."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ba83a07c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# чтение файлов\n",
    "\n",
    "try:\n",
    "    orders = pd.read_csv('/Users/lina/Documents/УЧЁБА/Яндекс Практикум/datasets/orders.csv')\n",
    "except:\n",
    "    orders = pd.read_csv('https://code.s3.yandex.net/datasets/orders.csv')\n",
    "    \n",
    "try:\n",
    "    visitors = pd.read_csv('/Users/lina/Documents/УЧЁБА/Яндекс Практикум/datasets/visitors.csv')\n",
    "except:\n",
    "    visitors = pd.read_csv('https://code.s3.yandex.net/datasets/visitors.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "22478799",
   "metadata": {},
   "outputs": [],
   "source": [
    "# преобразование дат в формат datetime\n",
    "\n",
    "orders['date'] = orders['date'].map(\n",
    "    lambda x: dt.datetime.strptime(x, '%Y-%m-%d')\n",
    ")\n",
    "\n",
    "visitors['date'] = visitors['date'].map(\n",
    "    lambda x: dt.datetime.strptime(x, '%Y-%m-%d')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "55d7caf8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1197 entries, 0 to 1196\n",
      "Data columns (total 5 columns):\n",
      " #   Column         Non-Null Count  Dtype         \n",
      "---  ------         --------------  -----         \n",
      " 0   transactionId  1197 non-null   int64         \n",
      " 1   visitorId      1197 non-null   int64         \n",
      " 2   date           1197 non-null   datetime64[ns]\n",
      " 3   revenue        1197 non-null   int64         \n",
      " 4   group          1197 non-null   object        \n",
      "dtypes: datetime64[ns](1), int64(3), object(1)\n",
      "memory usage: 46.9+ KB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "None"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 62 entries, 0 to 61\n",
      "Data columns (total 3 columns):\n",
      " #   Column    Non-Null Count  Dtype         \n",
      "---  ------    --------------  -----         \n",
      " 0   date      62 non-null     datetime64[ns]\n",
      " 1   group     62 non-null     object        \n",
      " 2   visitors  62 non-null     int64         \n",
      "dtypes: datetime64[ns](1), int64(1), object(1)\n",
      "memory usage: 1.6+ KB\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "None"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# общая информация о датафреймах\n",
    "\n",
    "display(orders.info())\n",
    "display(visitors.info())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "523c2c0f",
   "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>transactionId</th>\n",
       "      <th>visitorId</th>\n",
       "      <th>date</th>\n",
       "      <th>revenue</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3667963787</td>\n",
       "      <td>3312258926</td>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>1650</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2804400009</td>\n",
       "      <td>3642806036</td>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>730</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2961555356</td>\n",
       "      <td>4069496402</td>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>400</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3797467345</td>\n",
       "      <td>1196621759</td>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>9759</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2282983706</td>\n",
       "      <td>2322279887</td>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>2308</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   transactionId   visitorId       date  revenue group\n",
       "0     3667963787  3312258926 2019-08-15     1650     B\n",
       "1     2804400009  3642806036 2019-08-15      730     B\n",
       "2     2961555356  4069496402 2019-08-15      400     A\n",
       "3     3797467345  1196621759 2019-08-15     9759     B\n",
       "4     2282983706  2322279887 2019-08-15     2308     B"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>visitors</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "      <td>719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>A</td>\n",
       "      <td>619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>A</td>\n",
       "      <td>507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>A</td>\n",
       "      <td>717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>A</td>\n",
       "      <td>756</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date group  visitors\n",
       "0 2019-08-01     A       719\n",
       "1 2019-08-02     A       619\n",
       "2 2019-08-03     A       507\n",
       "3 2019-08-04     A       717\n",
       "4 2019-08-05     A       756"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# первые 5 строк датафреймов\n",
    "\n",
    "display(orders.head())\n",
    "display(visitors.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71f4dc04",
   "metadata": {},
   "source": [
    "В данных нет пропусков, столбцам присвоены корректные типы."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac71d939",
   "metadata": {},
   "source": [
    "### Предобработка данных"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45e74ae7",
   "metadata": {},
   "source": [
    "Проверим данные на дубликаты, а также убедимся, что инструмент \"деления\" трафика работал безошибочно."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e53511bf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# количество строк - полных дубликатов в датафреймах\n",
    "\n",
    "display(orders.duplicated().sum())\n",
    "display(visitors.duplicated().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a26113a",
   "metadata": {},
   "source": [
    "Дубликаты не обнаружены."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "11f6c209",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Количество id, попавших в обе группы: 58\n"
     ]
    }
   ],
   "source": [
    "# проверка id, попавших в обе группы теста\n",
    "\n",
    "group_a = orders[orders['group'] == 'A']['visitorId'].unique()\n",
    "group_b = orders[orders['group'] == 'B']['visitorId'].unique()\n",
    "\n",
    "common_ids = []\n",
    "for i in group_a:\n",
    "    if i in group_b:\n",
    "        common_ids.append(i)\n",
    "\n",
    "print('Количество id, попавших в обе группы:', len(common_ids))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecd59be5",
   "metadata": {},
   "source": [
    "❗️ 58 пользователей ошибочно попали в обе группы A/B-теста, что может повлиять на результаты. Будем учитывать этот факт при статистическом анализе."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b909684",
   "metadata": {},
   "source": [
    "### Анализ данных"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a76c6e1",
   "metadata": {},
   "source": [
    "Построим графики и проведём статистический анализ полученных данных."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65ee3e9c",
   "metadata": {},
   "source": [
    "#### График кумулятивной выручки по группам"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c89ad8f9",
   "metadata": {},
   "source": [
    "Создадим датафрейм `cumulativeData`, содержащий:\n",
    "- `date` — дату;\n",
    "- `group` — группу A/B-теста;\n",
    "- `orders` — число уникальных заказов в группе теста по указанную дату включительно;\n",
    "- `buyers` — число уникальных пользователей, совершивших хотя бы 1 заказ в группе теста по указанную дату включительно;\n",
    "- `revenue` — суммарную выручку заказов в группе теста по указанную дату включительно;\n",
    "- `visitors` — количество уникальных посетителей в группе теста по указанную дату включительно."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "c2a7d6f1",
   "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>date</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-15</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>2019-08-16</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>2019-08-16</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date group\n",
       "0  2019-08-15     B\n",
       "2  2019-08-15     A\n",
       "45 2019-08-16     A\n",
       "47 2019-08-16     B\n",
       "55 2019-08-01     A"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# датафрейм с уникальными парами значений 'date' и 'group' таблицы orders\n",
    "\n",
    "datesGroups = orders[['date', 'group']].drop_duplicates()\n",
    "\n",
    "datesGroups.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "195bf37c",
   "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>transactionId</th>\n",
       "      <th>visitorId</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "      <td>24</td>\n",
       "      <td>20</td>\n",
       "      <td>148579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>B</td>\n",
       "      <td>21</td>\n",
       "      <td>20</td>\n",
       "      <td>101217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>A</td>\n",
       "      <td>44</td>\n",
       "      <td>38</td>\n",
       "      <td>242401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>B</td>\n",
       "      <td>45</td>\n",
       "      <td>43</td>\n",
       "      <td>266748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>A</td>\n",
       "      <td>68</td>\n",
       "      <td>62</td>\n",
       "      <td>354874</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          date group  transactionId  visitorId  revenue\n",
       "55  2019-08-01     A             24         20   148579\n",
       "66  2019-08-01     B             21         20   101217\n",
       "175 2019-08-02     A             44         38   242401\n",
       "173 2019-08-02     B             45         43   266748\n",
       "291 2019-08-03     A             68         62   354874"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''\n",
    "переменная, содержащая:\n",
    "- дату;\n",
    "- группу A/B-теста;\n",
    "- число уникальных заказов в группе теста по указанную дату включительно;\n",
    "- число уникальных пользователей, совершивших хотя бы 1 заказ в группе теста по указанную дату включительно;\n",
    "- суммарную выручку заказов в группе теста по указанную дату включительно\n",
    "'''\n",
    "\n",
    "ordersAggregated = datesGroups.apply(\n",
    "    lambda x: orders[\n",
    "        np.logical_and(\n",
    "            orders['date'] <= x['date'], orders['group'] == x['group']\n",
    "        )\n",
    "    ].agg(\n",
    "        {\n",
    "            'date': 'max',\n",
    "            'group': 'max',\n",
    "            'transactionId': 'nunique',\n",
    "            'visitorId': 'nunique',\n",
    "            'revenue': 'sum',\n",
    "        }\n",
    "    ),\n",
    "    axis=1,\n",
    ").sort_values(by=['date', 'group'])\n",
    "\n",
    "ordersAggregated.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "af8c8353",
   "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>visitors</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "      <td>719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>B</td>\n",
       "      <td>713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>A</td>\n",
       "      <td>1338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>B</td>\n",
       "      <td>1294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>291</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>A</td>\n",
       "      <td>1845</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          date group  visitors\n",
       "55  2019-08-01     A       719\n",
       "66  2019-08-01     B       713\n",
       "175 2019-08-02     A      1338\n",
       "173 2019-08-02     B      1294\n",
       "291 2019-08-03     A      1845"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'''\n",
    "переменная, содержащая:\n",
    "- дату;\n",
    "- группу A/B-теста;\n",
    "- количество уникальных посетителей в группе теста по указанную дату включительно\n",
    "'''\n",
    "\n",
    "visitorsAggregated = datesGroups.apply(\n",
    "    lambda x: visitors[\n",
    "        np.logical_and(\n",
    "            visitors['date'] <= x['date'], visitors['group'] == x['group']\n",
    "        )\n",
    "    ].agg(\n",
    "        {\n",
    "            'date': 'max', \n",
    "            'group': 'max', \n",
    "            'visitors': 'sum'\n",
    "        }\n",
    "    ),\n",
    "    axis=1,\n",
    ").sort_values(by=['date', 'group'])\n",
    "\n",
    "visitorsAggregated.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a4648a0f",
   "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>orders</th>\n",
       "      <th>buyers</th>\n",
       "      <th>revenue</th>\n",
       "      <th>visitors</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "      <td>24</td>\n",
       "      <td>20</td>\n",
       "      <td>148579</td>\n",
       "      <td>719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>B</td>\n",
       "      <td>21</td>\n",
       "      <td>20</td>\n",
       "      <td>101217</td>\n",
       "      <td>713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>A</td>\n",
       "      <td>44</td>\n",
       "      <td>38</td>\n",
       "      <td>242401</td>\n",
       "      <td>1338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>B</td>\n",
       "      <td>45</td>\n",
       "      <td>43</td>\n",
       "      <td>266748</td>\n",
       "      <td>1294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>A</td>\n",
       "      <td>68</td>\n",
       "      <td>62</td>\n",
       "      <td>354874</td>\n",
       "      <td>1845</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date group  orders  buyers  revenue  visitors\n",
       "0 2019-08-01     A      24      20   148579       719\n",
       "1 2019-08-01     B      21      20   101217       713\n",
       "2 2019-08-02     A      44      38   242401      1338\n",
       "3 2019-08-02     B      45      43   266748      1294\n",
       "4 2019-08-03     A      68      62   354874      1845"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# объединяем ordersAggregated и visitorsAggregated по колонкам 'date' и 'group'\n",
    "\n",
    "cumulativeData = ordersAggregated.merge(\n",
    "    visitorsAggregated, left_on=['date', 'group'], right_on=['date', 'group']\n",
    ")\n",
    "cumulativeData.columns = [\n",
    "    'date',\n",
    "    'group',\n",
    "    'orders',\n",
    "    'buyers',\n",
    "    'revenue',\n",
    "    'visitors',\n",
    "]\n",
    "\n",
    "cumulativeData.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39d7e265",
   "metadata": {},
   "source": [
    "На основе `cumulativeData` с помощью среза получаем данные о датах, выручке и числе заказов в группах A и B."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "8241eb67",
   "metadata": {},
   "outputs": [],
   "source": [
    "# датафрейм с кумулятивным количеством заказов и кумулятивной выручкой по дням в группе А\n",
    "\n",
    "cumulativeRevenueA = cumulativeData[cumulativeData['group']=='A'][['date','revenue', 'orders']]\n",
    "\n",
    "# датафрейм с кумулятивным количеством заказов и кумулятивной выручкой по дням в группе B\n",
    "\n",
    "cumulativeRevenueB = cumulativeData[cumulativeData['group']=='B'][['date','revenue', 'orders']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "7e80cbc6",
   "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>date</th>\n",
       "      <th>revenue</th>\n",
       "      <th>orders</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>148579</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>242401</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>354874</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>425699</td>\n",
       "      <td>84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>549917</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  revenue  orders\n",
       "0 2019-08-01   148579      24\n",
       "2 2019-08-02   242401      44\n",
       "4 2019-08-03   354874      68\n",
       "6 2019-08-04   425699      84\n",
       "8 2019-08-05   549917     109"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>date</th>\n",
       "      <th>revenue</th>\n",
       "      <th>orders</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>101217</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>266748</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>380996</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>489567</td>\n",
       "      <td>78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>581995</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  revenue  orders\n",
       "1 2019-08-01   101217      21\n",
       "3 2019-08-02   266748      45\n",
       "5 2019-08-03   380996      61\n",
       "7 2019-08-04   489567      78\n",
       "9 2019-08-05   581995     101"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# первые 5 строк датафреймов\n",
    "\n",
    "display(cumulativeRevenueA.head())\n",
    "display(cumulativeRevenueB.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a52267b",
   "metadata": {},
   "source": [
    "Теперь можем построить график кумулятивной выручки по группам."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "1c01170f",
   "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": [
    "# график выручки группы А\n",
    "\n",
    "plt.plot(cumulativeRevenueA['date'], cumulativeRevenueA['revenue'], label='A')\n",
    "\n",
    "# график выручки группы B\n",
    "\n",
    "plt.plot(cumulativeRevenueB['date'], cumulativeRevenueB['revenue'], label='B')\n",
    "\n",
    "plt.xlabel('дата')\n",
    "plt.xticks(rotation=90)\n",
    "plt.ylabel('кумулятивная выручка, у. е.')\n",
    "plt.title('График кумулятивной выручки по группам A/B-теста') \n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d28f20d6",
   "metadata": {},
   "source": [
    "До середины наблюдаемого периода выручка росла равномерно в обеих группах, но затем в группе B произошёл скачок вверх. Он может быть вызван как ростом числа заказов, так и аномалиями в данных."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "faab819a",
   "metadata": {},
   "source": [
    "#### График кумулятивного среднего чека по группам"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ddf5dd53",
   "metadata": {},
   "source": [
    "Построим график кумулятивного среднего чека по группам."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "4c80cfc7",
   "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": [
    "# график кумулятивного среднего чека группы А\n",
    "\n",
    "plt.plot(cumulativeRevenueA['date'], cumulativeRevenueA['revenue'] / cumulativeRevenueA['orders'], label='A')\n",
    "\n",
    "# график кумулятивного среднего чека группы B\n",
    "\n",
    "plt.plot(cumulativeRevenueB['date'], cumulativeRevenueB['revenue'] / cumulativeRevenueB['orders'], label='B')\n",
    "\n",
    "\n",
    "plt.xlabel('дата')\n",
    "plt.xticks(rotation=90)\n",
    "plt.ylabel('кумулятивный средний чек, у. е.')\n",
    "plt.title('График кумулятивного среднего чека по группам A/B-теста') \n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e1f55e5",
   "metadata": {},
   "source": [
    "До середины наблюдаемого периода средний чек колебался в обеих группах. Как и на предыдущем графике, в группе B видим скачок вверх, после которого средний чек постепенно снижается."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "12e3252d",
   "metadata": {},
   "source": [
    "#### График относительного изменения кумулятивного среднего чека группы B к группе A"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14f451a0",
   "metadata": {},
   "source": [
    "Построим график относительного изменения кумулятивного среднего чека группы B к группе A. \n",
    "\n",
    "Для этого создадим переменную `mergedCumulativeRevenue`, объединяющую датафреймы `cumulativeRevenueA` и `cumulativeRevenueB`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "cc8f46cd",
   "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>date</th>\n",
       "      <th>revenueA</th>\n",
       "      <th>ordersA</th>\n",
       "      <th>revenueB</th>\n",
       "      <th>ordersB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>148579</td>\n",
       "      <td>24</td>\n",
       "      <td>101217</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>242401</td>\n",
       "      <td>44</td>\n",
       "      <td>266748</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>354874</td>\n",
       "      <td>68</td>\n",
       "      <td>380996</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>425699</td>\n",
       "      <td>84</td>\n",
       "      <td>489567</td>\n",
       "      <td>78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>549917</td>\n",
       "      <td>109</td>\n",
       "      <td>581995</td>\n",
       "      <td>101</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  revenueA  ordersA  revenueB  ordersB\n",
       "0 2019-08-01    148579       24    101217       21\n",
       "1 2019-08-02    242401       44    266748       45\n",
       "2 2019-08-03    354874       68    380996       61\n",
       "3 2019-08-04    425699       84    489567       78\n",
       "4 2019-08-05    549917      109    581995      101"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# собираем данные в одном датафрейме\n",
    "\n",
    "mergedCumulativeRevenue = cumulativeRevenueA.merge(\n",
    "    cumulativeRevenueB, \n",
    "    left_on='date', \n",
    "    right_on='date', \n",
    "    how='left', \n",
    "    suffixes=['A', 'B']\n",
    ")\n",
    "\n",
    "mergedCumulativeRevenue.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "64f1a885",
   "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": [
    "# строим отношение средних чеков\n",
    "\n",
    "plt.plot(\n",
    "    mergedCumulativeRevenue['date'], \n",
    "    (mergedCumulativeRevenue['revenueB'] / mergedCumulativeRevenue['ordersB']) /\n",
    "    (mergedCumulativeRevenue['revenueA'] / mergedCumulativeRevenue['ordersA']) - 1\n",
    ")\n",
    "\n",
    "# добавляем ось X\n",
    "\n",
    "plt.axhline(y=0, color='black', linestyle='--') \n",
    "\n",
    "plt.xlabel('дата')\n",
    "plt.xticks(rotation=90)\n",
    "plt.ylabel('относительное изменение')\n",
    "plt.title('График относительного изменения кумулятивного среднего чека группы B к группе A') \n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e062254",
   "metadata": {},
   "source": [
    "В нескольких точках график различия между сегментами резко \"скачет\", что может говорить о выбросах или особо крупных заказах."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad515708",
   "metadata": {},
   "source": [
    "#### График кумулятивного среднего количества заказов на посетителя по группам"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6326e57",
   "metadata": {},
   "source": [
    "Построим график кумулятивного среднего количества заказов на посетителя по группам.\n",
    "\n",
    "Для этого добавим в `cumulativeData` столбец `'conversion'` с отношением числа заказов к количеству пользователей в указанной группе в указанный день."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "de654beb",
   "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>orders</th>\n",
       "      <th>buyers</th>\n",
       "      <th>revenue</th>\n",
       "      <th>visitors</th>\n",
       "      <th>conversion</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "      <td>24</td>\n",
       "      <td>20</td>\n",
       "      <td>148579</td>\n",
       "      <td>719</td>\n",
       "      <td>0.033380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>B</td>\n",
       "      <td>21</td>\n",
       "      <td>20</td>\n",
       "      <td>101217</td>\n",
       "      <td>713</td>\n",
       "      <td>0.029453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>A</td>\n",
       "      <td>44</td>\n",
       "      <td>38</td>\n",
       "      <td>242401</td>\n",
       "      <td>1338</td>\n",
       "      <td>0.032885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>B</td>\n",
       "      <td>45</td>\n",
       "      <td>43</td>\n",
       "      <td>266748</td>\n",
       "      <td>1294</td>\n",
       "      <td>0.034776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>A</td>\n",
       "      <td>68</td>\n",
       "      <td>62</td>\n",
       "      <td>354874</td>\n",
       "      <td>1845</td>\n",
       "      <td>0.036856</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date group  orders  buyers  revenue  visitors  conversion\n",
       "0 2019-08-01     A      24      20   148579       719    0.033380\n",
       "1 2019-08-01     B      21      20   101217       713    0.029453\n",
       "2 2019-08-02     A      44      38   242401      1338    0.032885\n",
       "3 2019-08-02     B      45      43   266748      1294    0.034776\n",
       "4 2019-08-03     A      68      62   354874      1845    0.036856"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# считаем кумулятивную конверсию\n",
    "\n",
    "cumulativeData['conversion'] = cumulativeData['orders'] / cumulativeData['visitors']\n",
    "\n",
    "cumulativeData.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "751b01fe",
   "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>orders</th>\n",
       "      <th>buyers</th>\n",
       "      <th>revenue</th>\n",
       "      <th>visitors</th>\n",
       "      <th>conversion</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>A</td>\n",
       "      <td>24</td>\n",
       "      <td>20</td>\n",
       "      <td>148579</td>\n",
       "      <td>719</td>\n",
       "      <td>0.033380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>A</td>\n",
       "      <td>44</td>\n",
       "      <td>38</td>\n",
       "      <td>242401</td>\n",
       "      <td>1338</td>\n",
       "      <td>0.032885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>A</td>\n",
       "      <td>68</td>\n",
       "      <td>62</td>\n",
       "      <td>354874</td>\n",
       "      <td>1845</td>\n",
       "      <td>0.036856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>A</td>\n",
       "      <td>84</td>\n",
       "      <td>77</td>\n",
       "      <td>425699</td>\n",
       "      <td>2562</td>\n",
       "      <td>0.032787</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>A</td>\n",
       "      <td>109</td>\n",
       "      <td>100</td>\n",
       "      <td>549917</td>\n",
       "      <td>3318</td>\n",
       "      <td>0.032851</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date group  orders  buyers  revenue  visitors  conversion\n",
       "0 2019-08-01     A      24      20   148579       719    0.033380\n",
       "2 2019-08-02     A      44      38   242401      1338    0.032885\n",
       "4 2019-08-03     A      68      62   354874      1845    0.036856\n",
       "6 2019-08-04     A      84      77   425699      2562    0.032787\n",
       "8 2019-08-05     A     109     100   549917      3318    0.032851"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>date</th>\n",
       "      <th>group</th>\n",
       "      <th>orders</th>\n",
       "      <th>buyers</th>\n",
       "      <th>revenue</th>\n",
       "      <th>visitors</th>\n",
       "      <th>conversion</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>B</td>\n",
       "      <td>21</td>\n",
       "      <td>20</td>\n",
       "      <td>101217</td>\n",
       "      <td>713</td>\n",
       "      <td>0.029453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>B</td>\n",
       "      <td>45</td>\n",
       "      <td>43</td>\n",
       "      <td>266748</td>\n",
       "      <td>1294</td>\n",
       "      <td>0.034776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>B</td>\n",
       "      <td>61</td>\n",
       "      <td>59</td>\n",
       "      <td>380996</td>\n",
       "      <td>1803</td>\n",
       "      <td>0.033833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>B</td>\n",
       "      <td>78</td>\n",
       "      <td>75</td>\n",
       "      <td>489567</td>\n",
       "      <td>2573</td>\n",
       "      <td>0.030315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>B</td>\n",
       "      <td>101</td>\n",
       "      <td>98</td>\n",
       "      <td>581995</td>\n",
       "      <td>3280</td>\n",
       "      <td>0.030793</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date group  orders  buyers  revenue  visitors  conversion\n",
       "1 2019-08-01     B      21      20   101217       713    0.029453\n",
       "3 2019-08-02     B      45      43   266748      1294    0.034776\n",
       "5 2019-08-03     B      61      59   380996      1803    0.033833\n",
       "7 2019-08-04     B      78      75   489567      2573    0.030315\n",
       "9 2019-08-05     B     101      98   581995      3280    0.030793"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# отделяем данные по группе A\n",
    "\n",
    "cumulativeDataA = cumulativeData[cumulativeData['group']=='A']\n",
    "\n",
    "# отделяем данные по группе B\n",
    "\n",
    "cumulativeDataB = cumulativeData[cumulativeData['group']=='B']\n",
    "\n",
    "display(cumulativeDataA.head())\n",
    "display(cumulativeDataB.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "356ff808",
   "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": [
    "# график кумулятивного среднего количества заказов группы А\n",
    "\n",
    "plt.plot(cumulativeDataA['date'], cumulativeDataA['conversion'], label='A')\n",
    "\n",
    "# график кумулятивного среднего количества заказов группы B\n",
    "\n",
    "plt.plot(cumulativeDataB['date'], cumulativeDataB['conversion'], label='B')\n",
    "\n",
    "plt.xlabel('дата')\n",
    "plt.xticks(rotation=90)\n",
    "plt.ylabel('кумулятивное среднее')\n",
    "plt.title('График кумулятивного среднего количества заказов на посетителя по группам A/B-теста') \n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e32684b6",
   "metadata": {},
   "source": [
    "В начале A/B-теста кумулятивное среднее количество заказов на посетителя колебалось, но к концу наблюдаемого периода выровнялось. По графику видно, что в группе B среднее количество заказов на посетителя выше."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b059cc0",
   "metadata": {},
   "source": [
    "#### График относительного изменения кумулятивного среднего количества заказов на посетителя группы B к группе A"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4af0558",
   "metadata": {},
   "source": [
    "Построим график относительного изменения кумулятивного среднего количества заказов на посетителя группы B к группе A.\n",
    "\n",
    "Для этого создадим переменную `mergedCumulativeConversions`, объединяющую датафреймы `cumulativeDataA` и `cumulativeDataB`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4731b52a",
   "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>date</th>\n",
       "      <th>conversionA</th>\n",
       "      <th>conversionB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>0.033380</td>\n",
       "      <td>0.029453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>0.032885</td>\n",
       "      <td>0.034776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>0.036856</td>\n",
       "      <td>0.033833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>0.032787</td>\n",
       "      <td>0.030315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>0.032851</td>\n",
       "      <td>0.030793</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  conversionA  conversionB\n",
       "0 2019-08-01     0.033380     0.029453\n",
       "1 2019-08-02     0.032885     0.034776\n",
       "2 2019-08-03     0.036856     0.033833\n",
       "3 2019-08-04     0.032787     0.030315\n",
       "4 2019-08-05     0.032851     0.030793"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# собираем данные в одном датафрейме\n",
    "\n",
    "mergedCumulativeConversions = cumulativeDataA[['date','conversion']].merge(\n",
    "    cumulativeDataB[['date','conversion']], \n",
    "    left_on='date', \n",
    "    right_on='date', \n",
    "    how='left', \n",
    "    suffixes=['A', 'B']\n",
    ")\n",
    "\n",
    "mergedCumulativeConversions.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "9fc47958",
   "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": [
    "# конвертеры, которые позволяют использовать типы pandas в matplotlib  \n",
    "\n",
    "register_matplotlib_converters()\n",
    "\n",
    "# строим график\n",
    "\n",
    "plt.plot(\n",
    "    mergedCumulativeConversions['date'], \n",
    "    mergedCumulativeConversions['conversionB'] / mergedCumulativeConversions['conversionA'] - 1\n",
    ")\n",
    "\n",
    "# добавляем ось X\n",
    "\n",
    "plt.axhline(y=0, color='black', linestyle='--')\n",
    "\n",
    "plt.xlabel('дата')\n",
    "plt.xticks(rotation=90)\n",
    "plt.ylabel('относительное изменение')\n",
    "plt.title('График относительного изменения кумулятивного среднего количества заказов на посетителя группы B к группе A') \n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4ac494c",
   "metadata": {},
   "source": [
    "В начале наблюдений график колеблется, а после резкого скачка вверх в середине теста постепенно снижается. Тем не менее, группа B показывает стабильно более высокий результат относительно группы A по среднему количеству заказов."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6b0369a",
   "metadata": {},
   "source": [
    "#### Точечный график количества заказов по пользователям"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13bcbd61",
   "metadata": {},
   "source": [
    "Посмотрим, сколько заказов приходилось на каждого пользователя.\n",
    "\n",
    "Создадим датафрейм `ordersByUsers` с двумя столбцами: `'userId'` и `'orders'`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "4f643510",
   "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>userId</th>\n",
       "      <th>orders</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1023</th>\n",
       "      <td>4256040402</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>591</th>\n",
       "      <td>2458001652</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>569</th>\n",
       "      <td>2378935119</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>487</th>\n",
       "      <td>2038680547</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>199603092</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>744</th>\n",
       "      <td>3062433592</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>237748145</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>917</th>\n",
       "      <td>3803269165</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>299</th>\n",
       "      <td>1230306981</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>897</th>\n",
       "      <td>3717692402</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          userId  orders\n",
       "1023  4256040402      11\n",
       "591   2458001652      11\n",
       "569   2378935119       9\n",
       "487   2038680547       8\n",
       "44     199603092       5\n",
       "744   3062433592       5\n",
       "55     237748145       5\n",
       "917   3803269165       5\n",
       "299   1230306981       5\n",
       "897   3717692402       5"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# создадим датафрейм\n",
    "\n",
    "ordersByUsers = (\n",
    "    orders.groupby('visitorId', as_index=False)\n",
    "    .agg({'transactionId': 'nunique'})\n",
    ")\n",
    "ordersByUsers.columns = ['userId', 'orders']\n",
    "\n",
    "ordersByUsers.sort_values(by='orders', ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25399911",
   "metadata": {},
   "source": [
    "Наибольшее число заказов — 11 — сделали пользователи с id *4256040402* и *2458001652*.\n",
    "\n",
    "Построим гистограмму распределения количества заказов по пользователям."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "660fd421",
   "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": [
    "# гистограмма\n",
    "\n",
    "plt.hist(ordersByUsers['orders'])\n",
    "\n",
    "plt.xlabel('количество заказов')\n",
    "plt.ylabel('частота')\n",
    "plt.title('Гистограмма распределения количества заказов по пользователям') \n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f90c5512",
   "metadata": {},
   "source": [
    "Значения больше 4 встречаются редко и похожи на выбросы.\n",
    "\n",
    "Теперь построим точечный график количества заказов по пользователям. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a60c6e27",
   "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": [
    "# серия из чисел от 0 до количества наблюдений в ordersByUsers\n",
    "\n",
    "x_values = pd.Series(range(0, len(ordersByUsers)))\n",
    "\n",
    "# график\n",
    "\n",
    "plt.scatter(x_values, ordersByUsers['orders'])\n",
    "\n",
    "plt.xlabel('номер наблюдения')\n",
    "plt.ylabel('количество заказов')\n",
    "plt.title('Точечная диаграмма количества заказов по пользователям')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "adaec921",
   "metadata": {},
   "source": [
    "На графике видим выбивающиеся значения у 4 пользователей в 8, 9 и 11 заказов."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5164717c",
   "metadata": {},
   "source": [
    "#### 95-й и 99-й перцентили количества заказов на пользователя. Граница для определения аномальных пользователей"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27e195da",
   "metadata": {},
   "source": [
    "Посчитаем 95-й и 99-й перцентили количества заказов на пользователя."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "bf573850",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "95-й перцентиль количества заказов на пользователя: 2.0\n",
      "99-й перцентиль количества заказов на пользователя: 4.0\n"
     ]
    }
   ],
   "source": [
    "# np.percentile\n",
    "\n",
    "print('95-й перцентиль количества заказов на пользователя:', np.percentile(ordersByUsers['orders'], 95))\n",
    "print('99-й перцентиль количества заказов на пользователя:', np.percentile(ordersByUsers['orders'], 99))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "63f116a5",
   "metadata": {},
   "source": [
    "Если в качестве границы для определения аномальных пользователей мы возьмём 95-й перцентиль, то потеряем много данных, что видно на предыдущем графике.\n",
    "\n",
    "Можно также проверить себя без визуализации."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c1341a2e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Всего\n",
      "количество наблюдений 1031\n",
      "\n",
      "Граница — 95-й перцентиль\n",
      "количество наблюдений 1001\n",
      "потеря данных составит 2.910%\n",
      "\n",
      "Граница — 99-й перцентиль\n",
      "количество наблюдений 1021\n",
      "потеря данных составит 0.970%\n"
     ]
    }
   ],
   "source": [
    "# считаем долю потерянных данных\n",
    "\n",
    "total_orders = ordersByUsers.shape[0]\n",
    "orders_95 = ordersByUsers[ordersByUsers['orders'] > 2].shape[0]\n",
    "orders_99 = ordersByUsers[ordersByUsers['orders'] > 4].shape[0]\n",
    "\n",
    "print('Всего')\n",
    "print('количество наблюдений', total_orders)\n",
    "print()\n",
    "print('Граница — 95-й перцентиль')\n",
    "print('количество наблюдений', total_orders - orders_95)\n",
    "print('потеря данных составит', '{:.3%}'.format(orders_95 / total_orders))\n",
    "print()\n",
    "print('Граница — 99-й перцентиль')\n",
    "print('количество наблюдений', total_orders - orders_99)\n",
    "print('потеря данных составит', '{:.3%}'.format(orders_99 / total_orders))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e9cbe4d",
   "metadata": {},
   "source": [
    "Таким образом, будем определять аномальных пользователей с наибольшим числом заказов по **99-му перцентилю**."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "abe18c22",
   "metadata": {},
   "source": [
    "#### Точечный график стоимостей заказов"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9974eb7",
   "metadata": {},
   "source": [
    "Построим гистограмму распределения стоимостей заказов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "8ec4ee2a",
   "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": [
    "# гистограмма\n",
    "\n",
    "plt.hist(orders['revenue'])\n",
    "\n",
    "plt.xlabel('стоимость заказа')\n",
    "plt.ylabel('частота')\n",
    "plt.title('Гистограмма распределения стоимостей заказов')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e5be02f",
   "metadata": {},
   "source": [
    "Большинство заказов стоят менее 20000 у. е., есть выбросы.\n",
    "\n",
    "Построим точечный график стоимостей заказов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "a341df63",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAAAoSElEQVR4nO3deZgdZZn+8e+dpAPNloC0DgmBRAUUUURbfjg4CooSUAEVBMQNQUYUdRyMhlEBcUPirigimysElMlEQKOy6MxgMB0TloDBgEDS4NAsIUCi2Z7fH/WetvrknO7qztk65/5cV19d+3lqOfXU+751qhQRmJlZ+xrT7ADMzKy5nAjMzNqcE4GZWZtzIjAza3NOBGZmbc6JwMyszTkRmFnDSJok6ThJ4yQdJmnvZsdkTgSbRdJTub+Nktbk+k9odnxmLegx4J1AH3Bm+m9NJv+grDYk3QecHBG/aXYsZmbD4RJBHUnaStLXJD2Y/r4maas07mxJP8pNW95/gKSbJa2UdKukg3LjQtJzK/VLukzSZ3Pjnispcv0nSrpL0pOS7pX0r2UxnyXpoVSqWZMS3GDruCJXElpbtg73STokdW8n6f8k/U9u/NclLZe0StJCSf9Stj3WlZW6XihpalrfU9I2fUjSR3Pz7S/p92m7PSTpW5LGl22rW3P9YyX1SlpRFndf2Xx/SPOOK7IdK2yn9+amv1PSS1JspXULSU+n7l+keSZJmivpMUnLJL23bPvkt/W3KxwHIWm/3DTnpWGlfVL1+Ezjj5S0OO2feyRNl/QxVS4FL8kt80uSHkj7+wJJnWncQWXb+a0pnpOrbLMPS/prWv5tGvgdGOrY+VHq3lrSbyV9MTf+qrTcJyT9TtILyrbRI2k/zZe0T27cEZKWpGPrJknPz427L7cteiWdNtjx0GqcCOrrE8ABwIuBfYH9gU+mcRupsv0lTQauBT4L7AR8FPiZpK40SVSbt4CHgTcAOwAnAl+V9JL0uc8DzgAOjojtgDcWWJ6A6Wn6zw8y3QxgXdmwBWTbZifgJ8BVkrbOjZ8dEdvl/m7PjTsY2AN4HfDx0skN2AB8BNgZeDnwGuD9ZZ87XtLLUvfrgScqxPsIcCSApBcC25WNr7ody0k6BjibrEpkB+AI4NGIOK20bmnSfVP/Yan/CmAFMAk4Gvi8pFdXWP6ewGHlw4E/ASenaTrI9uf/5cZXPT4l7Q/8gGy/TQReCdwXEeflYn4AeGPqL51MzwX2TMt8LjCZrAqoPOYO4DPAQ5W2WfJzYC9ge+DbwJdz44Y6dkhJ+0rg7oj4eG7UL8iOnWcCfwR+nBt3MbBbWucbgE+nZe0JXA78G9AFXAf8PH+xUNoWwNuAb0jaYZB1aymjMhFIukTSw5LuKDj9W9NV2BJJP6l3fDknAOdExMMR0Ud2UL0jjXsAeJmkiRXmeztwXURcFxEbI+LXQA9weG7e144koIi4NiLuicxvgV8B/1I22bhhLLITWDvYBJL+CTgJ+EpZLD+KiEcjYn1EfBnYiuyLX8SnI+LplBwuBY5Py1wYEfPTMu8Dvgu8qmzei0knyPT/4grLvzjFDPDe8mkKbseSk4HzImJBmn5ZRNw/2MpJmgIcCHw8Iv4WEYuBi8iSSbnPk51Uy80FXpuuyN8I/Ab4W278YMfnScAlEfHrdAz2RsSfhohZwCnARyLisYh4MsV2XIXJ/xW4Bbi72vIi4t6IKCVpkZ20S+OGOnYEXEKWwN9XttxLIuLJiPg7WYLeV9KENG5pRKxO85P7zGOBa9P2WAd8iezY/+cKoY8DVjHE96KVjMpEAFwGTC8yoaQ9yK5yD0xXLf9Wv7A2MQnIf+HvT8Mgu9q7FfiLpJXAzNx0uwPHpCLoyjT+FcAuafxpwOmpaLuywud+NDffH/MjlN2pMT9VN6wkSy47A6Qv+ieB/5b0FDBnsJVL1QgTGbrB7yzgm2QNhfn5P6qsuqS0HhNKsRSwPNfdv10l7SnpmlT0X0V2Iipf5jXAQcqqUXYBFlZY/mJgR0l7kSXduWWxV92OFUwB7im4XiWTgNLJtOR+sivsfBwHkJ0Av19hGetS3EeTndgvqvAZ1Y7PkcTcBWwDLMwdf79Mw/Mxbw98DPjUUAuUNBNYTZborskNH+rYeRPwfOAF+c9XVhV4rrKqrlXAfWnUzrlpvg08TXZlf2MaPGBbRcRGsmMwvz/mpGX+Cvh8ROSTbksblYkgIn7HpieV50j6Zaov/O9UzQHZ1dz5EfF4mvfhBob6INlJvWS3NIx0lXd0ROwYERPJitQly4EfRsTE3N+2EXFumveaiHh2RExI85b7Umk+oL+6Ip24f0Z2NfOsNP46/nH1A1lR+u/APsBRQ6zfi4Engb8MMs2ewKHA1/MDU53ux4C3AqVt8ERZLIOZkuvu367Ad8iqRPaIiB2A/6iwzPXAfwI/JbuoqOZSYDbZCai/WqvgdsxbDjxnqBUq8yCwUzppluwG9JZNdx5wRkRsqLKci8i28zMi4taycVWPzxHG/AiwBnhB7ridkKv6KpkBXDlUqQggHfPbAO8GrpQ0seCxcy9Z9eHFZNVKJW8jq/I7hCx5TE3D++eNiPenz5xFdpxA2bZKpZ8pDNwfR6Vjbjfgw5JePtT6tYpRmQiquBD4YES8lKxOvbTz9wT2lPS/6QquUEmiRi4HPimpS9LOZHWlPxpiHtI0b5R0aLqC2VpZQ9uumxnPeLIidB+wXtJhZHXsed8BZqVqlaokjQE+CFw1yEkIshLGORWujrYnOyH3AeMknUlWf17UpyRtkxr6TiQ7YZeWuwp4Kl0MnFpl/guBuxhYP1zuJ2maC8uGF9mOeReRldJeqsxzJe0+yPRExHLgZuALaf+/iOyqPn/8vBrYGBHXVFpGWs6fyK7Kv1Bh9GDH58XAiZJeI2mMpMm5i6tqn7UR+B5Ze8kzIWvvknRobrLtyfbX5wZbVpp371TPD1k1zEayqq0ix87iiHiKrLrreZKOzX3+34FHyU72A9q1JO2Tjm2R7eM1adSVwOvT9ugATk/LublC6KXvQ1eFcS1pi0gEkrYjq6u7StJisnrhUjXKOLKGoYPI6pG/V6Vevh4+S1a3fxtwO1k1zWcHnYP+k8CRZFezfWRXZzPYzP2Vqhk+RHZQP052ddRf5SHpbcA/AV8rsLgLyOqY3650F0mK91gN/A3FI2SNjuXmkZ2g7iYrcv+NgdU9Q/ktsAy4nqwE9Ks0/KNk6/Uk2UlpdqWZU/3z8RGxstoHRMSqNM2fy4YPuh0rLOcqshPfT1Jcc8gaOYdyPNkV64NkV6ZnxcDbk3chuzIeVETMiIj/qjCq6vEZEX8gNYKTXW3/loGlh2o+TrZf5qdqkt8wsO5+B+AbpRL6ED5I1ij/BFnD9lvTBUXhYye1A5wIfC0lux+keXqBO4H5ZbN8gWyfPkpWpVZqe1pK1nb3TbJj+o1kjcP5doCfp+/BbcDVZDd8jAqj9ncEkqYC10TEPspa55dGxC4VprsAuCUiLk391wMzI2JBQwPewki6DLgsIm4qG/52YFxEXFanz51KVhXVERHr6/EZZu1miygRRMQqskbXYyCrv5O0bxo9h6w0QLoi2JOs/tA2z2NkReNyT6c/MxslRmWJQNLlZCf3ncnuiz6L7J7f75AVlzuAKyLinNSo82Wyu4w2AJ+LiCuaEbdtPpcIzGpvVCYCMzOrnS2iasjMzEZuOL8gbQk777xzTJ06tdlhmJmNKgsXLnwkIire0jrqEsHUqVPp6elpdhhmZqOKpKo/4HPVkJlZm3MiMDNrc04EZmZtzonAzKzNORGYmbW5UXfXkLWHOYt6mTVvKQ+uXMOkiZ3MOHQvjtpv8tAzmtmwORFYy5mzqJczrr6dNeuyp/n2rlzDGVdnb6l0MjCrPVcNWcuZNW9pfxIoWbNuA7PmLW1SRGZbNicCazkPrlwzrOFmtnmcCKzlTJrYOazhZrZ5nAis5cw4dC86O8YOGNbZMZYZh+5VZQ4z2xxuLLaWU2oQ9l1DZo3hRGAt6aj9JvvEb9YgrhoyM2tzTgRmZm3OicDMrM05EZiZtTknAjOzNudEYGbW5pwIzMzaXN0SgaRLJD0s6Y4q40+QdJuk2yXdLGnfesViZmbV1bNEcBkwfZDxfwFeFREvBD4DXFjHWMzMrIq6/bI4In4naeog42/O9c4Hdq1XLGZmVl2rtBGcBPyi2khJp0jqkdTT19fXwLDMzLZ8TU8Ekg4mSwQfrzZNRFwYEd0R0d3V1dW44MzM2kBTHzon6UXARcBhEfFoM2MxM2tXTSsRSNoNuBp4R0Tc3aw4zMzaXd1KBJIuBw4Cdpa0AjgL6ACIiAuAM4FnAN+WBLA+IrrrFY+ZmVVWz7uGjh9i/MnAyfX6fDMzK6bpjcVmZtZcTgRmZm3OicDMrM05EZiZtTknAjOzNudEYGbW5pwIzMzanBOBmVmbcyIwM2tzTgRmZm3OicDMrM05EZiZtTknAjOzNudEYGbW5pwIzMzanBOBmVmbcyIwM2tzTgRmZm3OicDMrM05EZiZtTknAjOzNle3RCDpEkkPS7qjynhJ+oakZZJuk/SSesViZmbV1bNEcBkwfZDxhwF7pL9TgO/UMRYzM6uibokgIn4HPDbIJEcCP4jMfGCipF3qFY+ZmVXWzDaCycDyXP+KNGwTkk6R1COpp6+vryHBmZm1i1HRWBwRF0ZEd0R0d3V1NTscM7MtSjMTQS8wJde/axpmZmYN1MxEMBd4Z7p76ADgiYh4qInxmJm1pXH1WrCky4GDgJ0lrQDOAjoAIuIC4DrgcGAZsBo4sV6xmJlZdXVLBBFx/BDjA/hAvT7fzMyKGRWNxWZmVj9OBGZmbc6JwMyszTkRmJm1uSETgaQDJC2Q9JSktZI2SFrViODMzKz+ipQIvgUcD/wZ6AROBs6vZ1BmZtY4haqGImIZMDYiNkTEpQz+VFEzMxtFivyOYLWk8cBiSecBD+G2BTOzLUaRE/o70nSnAU+TPR/oLfUMyszMGmfIEkFE3C9pfESskvRl4BkRcX8DYjMzswYoctfQZ4A+SecCNwK/k/SpukdmZmYNUaSN4C3A7mQvjpkCbAAWAJ+pY1xmZtYghRqLI2KlpBsi4nEASWvqHJeZmTVIkcbimwEi4ggASROAh+sZlJmZNU6RxuIPlfU/AbyubhGZmVlDFXofgaTXAy8Ati4Ni4hz6hWUmZk1TpG7hi4AjgU+CAg4hqzx2MzMtgBF2gj+OSLeCTweEZ8GXg7sWd+wzMysUYokgtIdQqslTQLWAbvULyQzM2ukIm0E10iaCMwC/ggEcFE9gzIzs8YpctdQ6YdjP5N0DbB1unPIzMy2AEUaiw8pdUfE34Hxkq6oa1RmZtYwRdoIzpZ0HICkE4HfAnOKLFzSdElLJS2TNLPC+N0k3ShpkaTbJB0+nODNzGzzFUkE04F3Svoj8ErgFRExZIlA0liyN5kdBuwNHC9p77LJPglcGRH7AccB3x5O8GZmtvmKJILxwHuAXuAxICTtVGC+/YFlEXFvRKwFrgCOLJsmgB1S9wTgwUJRm5lZzRS5a2gh2QlbwD7Am1P/s4eYbzKwPNe/Avh/ZdOcDfxK0geBbYFDqEDSKcApALvttluBkM3MrKghSwQRMS0inp3+l/6GSgJFHQ9cFhG7AocDP5S0SUwRcWFEdEdEd1dXV40+2szMoPizhvYhq+fPP2voB0PM1kv2/oKSXdOwvJPI2iCIiN9L2hrYGT/d1MysYYrcPnoW8M30dzBwHnBEgWUvAPaQNE3SeLLG4Lll0zwAvCZ9zvPJEk1f4ejNzGyzFWksPprsZP3XiDgR2JesYXdQEbGe7IX384C7yO4OWiLpHEmlRHI68F5JtwKXA++OiBjBepiZ2QgVqRpaExEbJa2XtANZtc2UoWYCiIjrgOvKhp2Z674TOHAY8ZqZWY0VSQQ96VlD3yO7g+gp4Pf1DMrMzBqnyLOG3p86L5D0S2CHiLitvmGZmVmjFGks7pL0XkldQDfwptRtZmZbgCKNxVcCrwJ+TXbX0I7Aj+oZlJmZNU6RNoKJEXGwpBUR8QEASXfUOS4zM2uQIiWCDen/WRWGmZnZKFckEfwHQERcDCBpAnBhPYMyM7PGKXLX0K/K+p8ge7y0mZltAYqUCMzMbAvmRGBm1uacCMzM2lzhRCDp9ZJWSHpI0gn1DMrMzBpnOCWCM4GDyN5SdnpdojEzs4Yr9GKapCMilgFIeqpO8ZiZWYMNmQgkfZPsHcW7SvoG2buLa/WqSjMza7JCj6FO/xdWGGZmZqNckUSwVUT4l8RmZluoIo3F76t7FGZm1jSFnj4q6c3lAyPi6jrEY2ZmDVYkEUwA3kDWSFwSgBOBmdkWoEgieCAi3lP3SMzMrCmKtBEsGenCJU2XtFTSMkkzq0zzVkl3Sloi6Scj/SwzMxuZIiWCT0naOiL+BiCpE3hWRNw32EySxpI9rvq1wApggaS5EXFnbpo9gDOAAyPicUnPHOF6mJnZCBUpEVwFbMz1b0jDhrI/sCwi7o2ItcAVwJFl07wXOD8iHgeIiIcLLNfMzGqoSCIYl07kAKTu8QXmmwwsz/WvSMPy9gT2lPS/kuZLml5pQZJOkdQjqaevr6/AR5uZWVFFEkGfpCNKPZKOBB6p0eePA/Yge5jd8cD3JE0snygiLoyI7ojo7urqqtFHm5kZFGsjeB/wY0ml11MuB95RYL5eYEquf9c0LG8FcEtErAP+IulussSwoMDyzcysBoq8s/ge4ABJ26X+ok8eXQDsIWkaWQI4Dnhb2TRzyEoCl0ramayq6N6CyzczsxoYsmpI0gRJXwFuAm6S9GVJE4aaLyLWA6cB84C7gCsjYomkc3JVTfOARyXdCdwIzIiIR0e4LmZmNgKKiMEnkH4G3AF8Pw16B7BvRGzy2IlG6O7ujp4eP/zUzGw4JC2MiO5K44q0ETwnIt6S6/+0pMU1iczMzJquyF1DayS9otQj6UBgTf1CMjOzRipSIjgV+H5qFxDwGPDuegZlZmaNU+SuocXAvpJ2SP2r6h2UmZk1TpF3Fp9Z1g9ARJxTp5jMzKyBirQRzASmA2uBp3N/Zma2BSjSRjAJOAF4I/An4JKIuK2uUZmZWcMMWSKIiJURcT7ZL4C3AS6qe1RmZtYwRdoIXge8E9gK+Anw/noHZWZmjVOkjeCXwPOArYETgaslza1rVGZm1jBF2gheTfayejMz2wIV+R3BTQ2Iw8zMmqRIG8GT/KNEoPQ/ImKHukVlZmYNU6SN4OtkTx89PiK2T39OAmZmW4git49+kuyl84dKuj49dM7MzLYQRaqGXpI6LwOmAd+WtDwi3lDPwMzMrDGK3DX05bL+x4Bt6xCLmZk1QZG7hg5uRCBmZtYchd9ZLKkn/RV6Z7GZmY0ORe4augR4Enhr+lsFXFrPoMzMrHH8zmIzszbndxabmbW5IongVOB8SfdJug/4FvCvRRYuabqkpZKWSZo5yHRvkRSSugtFbWZmNVOkauivETHsdxZLGgucD7wWWAEskDQ3Iu4sm2574MPALcOK3MzMaqJIieA6yBLAMF9cvz+wLCLujYi1wBVkv1Au9xngi8DfhrFsMzOrkSKJYKQmA8tz/SvSsH7pV8tTIuLawRYk6ZTS7at9fX21j9TMrI0VSQQvkrQq9/ekpOGUDCqSNAb4CnD6UNNGxIUR0R0R3V1dXZv70WZmllOkjeD2iNhvBMvuBabk+ndNw0q2B/YBbpIE8E/AXElHRETPCD7PzMxGoJ5VQwuAPSRNkzQeOA7of8VlRDwRETtHxNSImArMB5wEzMwarEgieMvQk2wqItYDpwHzgLuAKyNiiaRzJB0xkmWamVntFakaOkvShyNiJYCkHYEvR8R7hpoxIq4j3XWUG3ZmlWkPKhCLmZnVWKHG4lISAIiIx4GRtBmYmVkLKpIIxqRSAACSdqJYScLMzEaBoi+m+b2kq1L/McDn6heSmZk1UpEX0/xAUg/w6jTozeWPiTAzs9GrUBVPOvH75G9mtgWq5+8IzMxsFHAiMDNrc04EZmZtzonAzKzNORGYmbU5JwIzszbnRGBm1uacCMzM2pwTgZlZm3MiMDNrc04EZmZtzonAzKzNORGYmbU5JwIzszbnRGBm1uacCMzM2lxdE4Gk6ZKWSlomaWaF8f8u6U5Jt0m6XtLu9YzHzGw0mrOolwPPvYFpM6/lwHNvYM6i3pouv26JQNJY4HzgMGBv4HhJe5dNtgjojogXAT8FzqtXPGZmo9GcRb2ccfXt9K5cQwC9K9dwxtW31zQZ1LNEsD+wLCLujYi1wBXAkfkJIuLGiFideucDu9YxHjOzUWfWvKWsWbdhwLA16zYwa97Smn1GPRPBZGB5rn9FGlbNScAvKo2QdIqkHkk9fX19NQzRzKy1PbhyzbCGj0RLNBZLejvQDcyqND4iLoyI7ojo7urqamxwZmZNNGli57CGj0Q9E0EvMCXXv2saNoCkQ4BPAEdExN/rGI+Z2agz49C96OwYO2BYZ8dYZhy6V80+Y1zNlrSpBcAekqaRJYDjgLflJ5C0H/BdYHpEPFzHWMzMRqWj9stq1GfNW8qDK9cwaWInMw7dq394LdQtEUTEekmnAfOAscAlEbFE0jlAT0TMJasK2g64ShLAAxFxRL1iMjMbjY7ab3JNT/zl6lkiICKuA64rG3ZmrvuQen6+mZkNrSUai83MrHmcCMzM2pwTgZlZm3MiMDNrc04EZmZtzonAzKzNORGYmbU5JwIzszZX1x+UmVl7mbOot66PQhjNWnnbOBGYbaEafeIpvUCl9Oz80gtUgJY54TVLq28bJwKzFlDppA0jf9BYM048g71ApRVOds3U6tvGicCsySqdtGdcdSsI1m2I/mHDOZE348TTiBeojFatvm2cCMyarNJJe93G2GS64ZzIm3HimTSxk94Ky6/lC1Ra0ZxFvZw9dwkr16wDYMdtOjjrjS8YsJ9afds4ERTQyo08NvoN5+RcdNqJ23Tw+Op1FYfXy4xD9xpQsgHoGCNWr13PtJnXjvi7U/r+9a5cw1iJDRFMrsH3sBbf6zmLeplx1a0DEvfjq9cx46e3Av8ovVXaNrV+uczmcCIYQi3rWls5obRybKPNcLdltavFatMWEZsWKAYdXgvlL1CZ0NnB02vX9yekkXx3yr9/G2L4VWXV2l9q8b2eNW9pxdLbug0xoPR21H6T6bn/MX48/wFKU49R4Y+pOyeCIdSqrrWV7xpo5djqrdYJsOi2zH/uhM4OOsaqvz0AsivpDRHkzzEdY9R/Ehsq7lI1RbmVa9Zx4Lk31C3R51+gcuC5N2wSx3C/O5W+f8NZVrX9sXXHmJp8rwcroeXHzVnUy+w/LCefMp5eu2GTkkOzOBEModqO7l25ZljF3eEklEZfnW9OsqvHibRR616PBFhtW55+5a18ZPZiJk3s5ODndfGzhb39061cs46OMWLHbTpYuXpd/zSzFyxnYy45oOJxl6pQKmlEop+zqLdqKaeWVWFDja+2P6oll6KxlY7TwQpY+dJb0ZJDszgRDGGwYnuQ7vAokNWLNt6N5IqyyAlzsNsTR/qFrfWJtNLyPjJ7MT33P8Znj3ph4WXk1/Pg53Vx45/6Km6naieJs+cuGXC1vnb9Blav29g/TaXGwJJq2yxfpZGvHihZtzFYtWY9Xz32xRy132QOPPeGASUEyE4ap195a8UTfHncQ9UA1fMOotJ+rGaMNOhFVHnj62CGqiobbuN4kaq38uO0ko6xGlD/X7TkUOmzGnFh5ESQVNvglRp5yq3bEPz7lYv7r/gq7axqCWVC58DGuyJX54OdgEvLGKo+tPz2xErGSMxZ1Fv1wKtltVm1L34AP57/AN277zRkyan8Srt35Rp+NP+B/unLE1W1BLhyzbr+WCrF9PjqdZx+VeXkX6S+v9oW3xDRH1+1ZVS7yi+Pu4ha3EFU6XszWHUOVK/nH04CKOlduWbQqq5q+2NiZwd/X79xQJwqsDwYvLoKYNvxY/ncm7ILlwPPvWHI46FS8qm0LepZklPUs/WoDrq7u6Onp2ezl1NeR/v02vWbnBS3HT+W1Ws3MKGzA4mKd2EMJn9nQ6W7CyC7cph19L7A4FfnkB2okyZ2sjrXAFc+flxZXXNnx1i27hgz7NjzStuhPMlNm3lt1ZPa5GE0gBY1RrAxsmUf/LwuZv9hecXidhGTJ3YWunIeSvldLMCQFw6tYmJnB4vPet2Qd+VUOintuE0Hr3/RLgMSL2TH4Ei36ebM2zFGzDpm34oXC5X2x47bdLD3Ltsz/97HKybXjjHi2P2ncM2tD21yW+hHZi8eNM6JnR2cfcQLCh0Hpe9/eWl/sHlL+224JC2MiO6K49olEYzkaiOvY4xGfNIpncC2FKUvxKd/vmSzEow13zYdYwZUe+V1jBXHvmzKZiXcZhnqRL85avV9LiW+yWVVmGMGad8p+VqqQhzW57V7Iqh2NW5mNhqNpFQwWCKo62OoJU2XtFTSMkkzK4zfStLsNP4WSVPrEcenf77EScDMthgjrdmopm6JQNJY4HzgMGBv4HhJe5dNdhLweEQ8F/gq8MVaxzFnUa+rL8xsizNnUW/NllXPEsH+wLKIuDci1gJXAEeWTXMk8P3U/VPgNZJq+nu7WfOW1nJxZmYtoZbntnomgsnA8lz/ijSs4jQRsR54AnhG+YIknSKpR1JPX1/fsIJolaf7mZnVUi3PbaPiVZURcWFEdEdEd1dX17DmbZWn+5mZ1VItz231TAS9wJRc/65pWMVpJI0DJgCP1jKIVnm6n5lZrdT6yaX1TAQLgD0kTZM0HjgOmFs2zVzgXan7aOCGqPH9rEftN5mvHftiOlqk7LNNx5jRUQyzuhCwxzO3ZWxtm8KsDZSOmMkTO/nCm19Y018X1+0RExGxXtJpwDxgLHBJRCyRdA7QExFzgYuBH0paBjxGlixqLv9ERDMzG6iuzxqKiOuA68qGnZnr/htwTD1jMDOzwbmWwsyszTkRmJm1OScCM7M250RgZtbmRt3TRyX1AfePcPadgUdqGE4zeV1ak9elNXldYPeIqPiL3FGXCDaHpJ5qj2Edbbwurcnr0pq8LoNz1ZCZWZtzIjAza3PtlggubHYANeR1aU1el9bkdRlEW7URmJnZptqtRGBmZmWcCMzM2lzbJAJJ0yUtlbRM0sxmxzMUSVMk3SjpTklLJH04Dd9J0q8l/Tn93zENl6RvpPW7TdJLmrsGA0kaK2mRpGtS/zRJt6R4Z6dHlSNpq9S/LI2f2tTAy0iaKOmnkv4k6S5JLx/F++Qj6di6Q9LlkrYeLftF0iWSHpZ0R27YsPeDpHel6f8s6V2VPqtJ6zIrHWO3SfpPSRNz485I67JU0qG54SM/x0XEFv9H9hjse4BnA+OBW4G9mx3XEDHvArwkdW8P3A3sDZwHzEzDZwJfTN2HA78ge2z5AcAtzV6HsvX5d+AnwDWp/0rguNR9AXBq6n4/cEHqPg6Y3ezYy9bj+8DJqXs8MHE07hOy18T+BejM7Y93j5b9ArwSeAlwR27YsPYDsBNwb/q/Y+resUXW5XXAuNT9xdy67J3OX1sB09J5bezmnuOafkA2aEO/HJiX6z8DOKPZcQ1zHf4LeC2wFNglDdsFWJq6vwscn5u+f7pm/5G9ne564NXANekL+UjuQO/fP2Tvr3h56h6XplOz1yHFMyGdPFU2fDTuk9L7wndK2/ka4NDRtF+AqWUnz2HtB+B44Lu54QOma+a6lI17E/Dj1D3g3FXaL5t7jmuXqqHSQV+yIg0bFVIxfD/gFuBZEfFQGvVX4Fmpu5XX8WvAx4CNqf8ZwMqIWJ/687H2r0ca/0SavhVMA/qAS1M110WStmUU7pOI6AW+BDwAPES2nRcyOvdLyXD3Q8vunzLvISvRQJ3WpV0SwaglaTvgZ8C/RcSq/LjIUn9L3/8r6Q3AwxGxsNmx1MA4siL8dyJiP+BpsiqIfqNhnwCk+vMjyZLbJGBbYHpTg6qh0bIfhiLpE8B64Mf1/Jx2SQS9wJRc/65pWEuT1EGWBH4cEVenwf8naZc0fhfg4TS8VdfxQOAISfcBV5BVD30dmCip9Ia8fKz965HGTwAebWTAg1gBrIiIW1L/T8kSw2jbJwCHAH+JiL6IWAdcTbavRuN+KRnufmjl/YOkdwNvAE5IiQ3qtC7tkggWAHukOyLGkzV2zW1yTIOSJLJ3Ot8VEV/JjZoLlO5ueBdZ20Fp+DvTHRIHAE/kislNExFnRMSuETGVbLvfEBEnADcCR6fJytejtH5Hp+lb4souIv4KLJe0Vxr0GuBORtk+SR4ADpC0TTrWSusy6vZLznD3wzzgdZJ2TCWk16VhTSdpOll16hERsTo3ai5wXLqLaxqwB/AHNvcc18zGngY3xhxOdufNPcAnmh1PgXhfQVa0vQ1YnP4OJ6uXvR74M/AbYKc0vYDz0/rdDnQ3ex0qrNNB/OOuoWenA3gZcBWwVRq+depflsY/u9lxl63Di4GetF/mkN1tMir3CfBp4E/AHcAPye5EGRX7BbicrG1jHVlJ7aSR7Aey+vdl6e/EFlqXZWR1/qXv/gW56T+R1mUpcFhu+IjPcX7EhJlZm2uXqiEzM6vCicDMrM05EZiZtTknAjOzNudEYGbW5pwIzGogPfnyfEl/kHS7pH2bHZNZUb591KwGJM0je2jZf0XEhmbHYzYcLhFY00maWvYs9qMlXZYbd0N6Lvv1knZLwy+TtELS2NR/qqQoPSdf0tvT1fliSd/NTfeUpK8qew7/9ZK6KsRzmaSjc/135JY7R9LCNP8padjewO7AmcDi9Hz5rXLz35dKCXeW1lPSuyV9K3V/QtLd6XPOKovlqbQO/fMOtX65abol3ZS6z5b00dR9SNpW3an/JGXPvl8s6QlJBxXeebZFcCKwVvdN4PsR8SKyB299Izeul+zRyZA9QG0ZgKTnA8cCB0bEi4ENwAlpum2Bnoh4AfBbYMCJt4D3RMRLgW7gQ5KeAXSRPbztrRHxQrKH052am2cs8CqyX34OIOlVZL8kfRnwUmC6pEPyk6R1ODw3z2DrV8SZpG2VnAu8Mi3rv4exHNtCjBt6ErOGeI6kxal7AtlJGrLnrL85df+Q7OUj5PrfIekBsscK7JqGv4bspLoge4wOnfzjAWQbgdmp+0dkD1urZJakT5Ziyw3/kKQ3pe4pZM96EdnLTu5Ow78PfIDs8dulz/8bsEPZZxwLHAVcFRFPAEi6guxFJb9JD3dbzaYGW7/O3HbsJHt0QT9JbyF7Ls1Lc4M3kr386GGsLTkRWKu4J12Rkqpl3lBgnr8CHcAMsieaHpyGi6wUcUaBZVRrJJsRET9N8ZSqcw4ie2rnyyNidap22RpYVWUZSNoaGJOmLx89m+wdAC+qMvtUKj9BcrD1W5Pbjt1k7xwoGUu2rd5A9uTUklOBmyX1kSW3/DzWBlw1ZK3uZrInKUJW/VFedXEp8MyI+GNu2PXA0ZKeCf139Oyexo3hH0/XfBvwP8OIZQLweDqpP4/stYeQPfxrT0nPTf3v4B8lmqOB3w+yzN8Br5c0IT018ljgpjTuGLI3h5UbbP0G83bguoh4pGz4g2SvNtwXVw21JZcIrNV9kOyNYDPI3g52Yn5kRFwLXFs27M5UrfMrSWPInur4AeB+spfJ7J/GP0x24i3ql8D7JN1FdvKfnz7vaUnvBf4zXfX/AbggVSGdSvYu4Ioi4h5Js4D/JSudzI6IGyQdAXwGeEDZy33GA9MkvS8iLhhk/QbzLCD/SHNSG8c3yB53vKFCqcXagG8ftbYi6amI2K7ZcQxF2UtJiIjLcsP2AY6OiLObE5VtqVwiMGtNv60wbDkD6/bNasIlAjOzNufGYjOzNudEYGbW5pwIzMzanBOBmVmbcyIwM2tz/x9xrEmHzRuDvAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# серия из чисел от 0 до количества наблюдений в orders\n",
    "\n",
    "x_values = pd.Series(range(0, len(orders['revenue'])))\n",
    "\n",
    "# график\n",
    "\n",
    "plt.scatter(x_values, orders['revenue'])\n",
    "\n",
    "plt.xlabel('номер наблюдения')\n",
    "plt.ylabel('стоимость заказа')\n",
    "plt.title('Точечная диаграмма стоимостей заказов')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1757cbb6",
   "metadata": {},
   "source": [
    "Обнаружилось несколько дорогих заказов, например, заказ стоимостью более 1200000 у. е. Масштабируем график, взглянув на заказы до 100000 у. е."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "72d410aa",
   "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": [
    "# серия из чисел от 0 до количества наблюдений в orders\n",
    "\n",
    "x_values = pd.Series(range(0, len(orders['revenue'])))\n",
    "\n",
    "# график\n",
    "\n",
    "plt.scatter(x_values, orders['revenue'])\n",
    "\n",
    "plt.xlabel('номер наблюдения')\n",
    "plt.ylabel('стоимость заказа')\n",
    "plt.ylim(0, 100000)\n",
    "plt.title('Точечная диаграмма стоимостей заказов')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1eea730b",
   "metadata": {},
   "source": [
    "По графику можно выбрать границу аномальных заказов между 20000 и 40000 у. е."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d36e451e",
   "metadata": {},
   "source": [
    "#### 95-й и 99-й перцентили стоимости заказов. Граница для определения аномальных заказов"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed0ddfda",
   "metadata": {},
   "source": [
    "Посчитаем 95-й и 99-й перцентили стоимости заказов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "4f7cc12c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "95-й перцентиль стоимости заказов: 28000.0\n",
      "99-й перцентиль стоимости заказов: 58233.2\n"
     ]
    }
   ],
   "source": [
    "# np.percentile\n",
    "\n",
    "print('95-й перцентиль стоимости заказов:', round(np.percentile(orders['revenue'], 95), 2))\n",
    "print('99-й перцентиль стоимости заказов:', round(np.percentile(orders['revenue'], 99), 2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b85d440a",
   "metadata": {},
   "source": [
    "Чтобы выбрать границу определения аномальных пользователей, посчитаем долю потери данных для каждого случая."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "5a367677",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Всего\n",
      "количество наблюдений 1197\n",
      "\n",
      "Граница — 95-й перцентиль\n",
      "количество наблюдений 1137\n",
      "потеря данных составит 5.013%\n",
      "\n",
      "Граница — 99-й перцентиль\n",
      "количество наблюдений 1185\n",
      "потеря данных составит 1.003%\n"
     ]
    }
   ],
   "source": [
    "# считаем долю потерянных данных\n",
    "\n",
    "total_revenue = orders.shape[0]\n",
    "revenue_95 = orders[orders['revenue'] > 28000].shape[0]\n",
    "revenue_99 = orders[orders['revenue'] > 58233.2].shape[0]\n",
    "\n",
    "print('Всего')\n",
    "print('количество наблюдений', total_revenue)\n",
    "print()\n",
    "print('Граница — 95-й перцентиль')\n",
    "print('количество наблюдений', total_revenue - revenue_95)\n",
    "print('потеря данных составит', '{:.3%}'.format(revenue_95 / total_revenue))\n",
    "print()\n",
    "print('Граница — 99-й перцентиль')\n",
    "print('количество наблюдений', total_revenue - revenue_99)\n",
    "print('потеря данных составит', '{:.3%}'.format(revenue_99 / total_revenue))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc76b521",
   "metadata": {},
   "source": [
    "При выборе 95-го перцентиля потеря данных составит более 5%, поэтому будем определять аномальных пользователей с наибольшей стоимостью заказов по **99-му перцентилю**."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7437732c",
   "metadata": {},
   "source": [
    "#### Статистическая значимость различий в среднем количестве заказов на посетителя между группами по \"сырым\" данным"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6082755c",
   "metadata": {},
   "source": [
    "Создадим датафрейм для проверки статистических гипотез `final_data` со столбцами:\n",
    "- `date` — дата;\n",
    "- `ordersPerDateA` — количество заказов в выбранную дату в группе A;\n",
    "- `revenuePerDateA` — суммарная выручка в выбранную дату в группе A;\n",
    "- `ordersPerDateB` — количество заказов в выбранную дату в группе B;\n",
    "- `revenuePerDateB` — суммарная выручка в выбранную дату в группе B;\n",
    "- `ordersCummulativeA` — суммарное число заказов до выбранной даты включительно в группе A;\n",
    "- `revenueCummulativeA` — суммарная выручка до выбранной даты включительно в группе A;\n",
    "- `ordersCummulativeB` — суммарное количество заказов до выбранной даты включительно в группе B;\n",
    "- `revenueCummulativeB` — суммарная выручка до выбранной даты включительно в группе B;\n",
    "- `visitorsPerDateA` — количество пользователей в выбранную дату в группе A;\n",
    "- `visitorsPerDateB` — количество пользователей в выбранную дату в группе B;\n",
    "- `visitorsCummulativeA` — количество пользователей до выбранной даты включительно в группе A;\n",
    "- `visitorsCummulativeB` — количество пользователей до выбранной даты включительно в группе B."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "c60d840b",
   "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>date</th>\n",
       "      <th>visitorsPerDateA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>756</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  visitorsPerDateA\n",
       "0 2019-08-01               719\n",
       "1 2019-08-02               619\n",
       "2 2019-08-03               507\n",
       "3 2019-08-04               717\n",
       "4 2019-08-05               756"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество посетителей по дням в группе A\n",
    "\n",
    "visitorsADaily = visitors[visitors['group'] == 'A'][['date', 'visitors']]\n",
    "visitorsADaily.columns = ['date', 'visitorsPerDateA']\n",
    "\n",
    "visitorsADaily.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "ce6aaac8",
   "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>date</th>\n",
       "      <th>visitorsCummulativeA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>1338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>1845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>2562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>3318</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  visitorsCummulativeA\n",
       "0 2019-08-01                   719\n",
       "1 2019-08-02                  1338\n",
       "2 2019-08-03                  1845\n",
       "3 2019-08-04                  2562\n",
       "4 2019-08-05                  3318"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество посетителей с накоплением по дням в группе A\n",
    "\n",
    "visitorsACummulative = visitorsADaily.apply(\n",
    "    lambda x: visitorsADaily[visitorsADaily['date'] <= x['date']].agg(\n",
    "        {'date': 'max', 'visitorsPerDateA': 'sum'}\n",
    "    ),\n",
    "    axis=1,\n",
    ")\n",
    "visitorsACummulative.columns = ['date', 'visitorsCummulativeA']\n",
    "\n",
    "visitorsACummulative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "17ee4792",
   "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>date</th>\n",
       "      <th>visitorsPerDateB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>707</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date  visitorsPerDateB\n",
       "31 2019-08-01               713\n",
       "32 2019-08-02               581\n",
       "33 2019-08-03               509\n",
       "34 2019-08-04               770\n",
       "35 2019-08-05               707"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество посетителей по дням в группе B\n",
    "\n",
    "visitorsBDaily = visitors[visitors['group'] == 'B'][['date', 'visitors']]\n",
    "visitorsBDaily.columns = ['date', 'visitorsPerDateB']\n",
    "\n",
    "visitorsBDaily.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "ada77765",
   "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>date</th>\n",
       "      <th>visitorsCummulativeB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>1294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>1803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>2573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>3280</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date  visitorsCummulativeB\n",
       "31 2019-08-01                   713\n",
       "32 2019-08-02                  1294\n",
       "33 2019-08-03                  1803\n",
       "34 2019-08-04                  2573\n",
       "35 2019-08-05                  3280"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество посетителей с накоплением по дням в группе B\n",
    "\n",
    "visitorsBCummulative = visitorsBDaily.apply(\n",
    "    lambda x: visitorsBDaily[visitorsBDaily['date'] <= x['date']].agg(\n",
    "        {'date': 'max', 'visitorsPerDateB': 'sum'}\n",
    "    ),\n",
    "    axis=1,\n",
    ")\n",
    "visitorsBCummulative.columns = ['date', 'visitorsCummulativeB']\n",
    "\n",
    "visitorsBCummulative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "f80f8801",
   "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>date</th>\n",
       "      <th>ordersPerDateA</th>\n",
       "      <th>revenuePerDateA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>24</td>\n",
       "      <td>148579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>20</td>\n",
       "      <td>93822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>24</td>\n",
       "      <td>112473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>16</td>\n",
       "      <td>70825</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>25</td>\n",
       "      <td>124218</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  ordersPerDateA  revenuePerDateA\n",
       "0 2019-08-01              24           148579\n",
       "1 2019-08-02              20            93822\n",
       "2 2019-08-03              24           112473\n",
       "3 2019-08-04              16            70825\n",
       "4 2019-08-05              25           124218"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество заказов и выручка по дням в группе A\n",
    "\n",
    "ordersADaily = (\n",
    "    orders[orders['group'] == 'A'][['date', 'transactionId', 'visitorId', 'revenue']]\n",
    "    .groupby('date', as_index=False)\n",
    "    .agg({'transactionId': pd.Series.nunique, 'revenue': 'sum'})\n",
    ")\n",
    "ordersADaily.columns = ['date', 'ordersPerDateA', 'revenuePerDateA']\n",
    "\n",
    "ordersADaily.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "58857190",
   "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>date</th>\n",
       "      <th>ordersCummulativeA</th>\n",
       "      <th>revenueCummulativeA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>24</td>\n",
       "      <td>148579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>44</td>\n",
       "      <td>242401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>68</td>\n",
       "      <td>354874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>84</td>\n",
       "      <td>425699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>109</td>\n",
       "      <td>549917</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  ordersCummulativeA  revenueCummulativeA\n",
       "0 2019-08-01                  24               148579\n",
       "1 2019-08-02                  44               242401\n",
       "2 2019-08-03                  68               354874\n",
       "3 2019-08-04                  84               425699\n",
       "4 2019-08-05                 109               549917"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество заказов и выручка по дням с накоплением в группе A\n",
    "\n",
    "ordersACummulative = ordersADaily.apply(\n",
    "    lambda x: ordersADaily[ordersADaily['date'] <= x['date']].agg(\n",
    "        {'date': 'max', 'ordersPerDateA': 'sum', 'revenuePerDateA': 'sum'}\n",
    "    ),\n",
    "    axis=1,\n",
    ").sort_values(by=['date'])\n",
    "ordersACummulative.columns = [\n",
    "    'date',\n",
    "    'ordersCummulativeA',\n",
    "    'revenueCummulativeA',\n",
    "]\n",
    "\n",
    "ordersACummulative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "daf4e06e",
   "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>date</th>\n",
       "      <th>ordersPerDateB</th>\n",
       "      <th>revenuePerDateB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>21</td>\n",
       "      <td>101217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>24</td>\n",
       "      <td>165531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>16</td>\n",
       "      <td>114248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>17</td>\n",
       "      <td>108571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>23</td>\n",
       "      <td>92428</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  ordersPerDateB  revenuePerDateB\n",
       "0 2019-08-01              21           101217\n",
       "1 2019-08-02              24           165531\n",
       "2 2019-08-03              16           114248\n",
       "3 2019-08-04              17           108571\n",
       "4 2019-08-05              23            92428"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество заказов и выручка по дням в группе B\n",
    "\n",
    "ordersBDaily = (\n",
    "    orders[orders['group'] == 'B'][['date', 'transactionId', 'visitorId', 'revenue']]\n",
    "    .groupby('date', as_index=False)\n",
    "    .agg({'transactionId': pd.Series.nunique, 'revenue': 'sum'})\n",
    ")\n",
    "ordersBDaily.columns = ['date', 'ordersPerDateB', 'revenuePerDateB']\n",
    "\n",
    "ordersBDaily.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "508a4112",
   "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>date</th>\n",
       "      <th>ordersCummulativeB</th>\n",
       "      <th>revenueCummulativeB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>21</td>\n",
       "      <td>101217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>45</td>\n",
       "      <td>266748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>61</td>\n",
       "      <td>380996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>78</td>\n",
       "      <td>489567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>101</td>\n",
       "      <td>581995</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  ordersCummulativeB  revenueCummulativeB\n",
       "0 2019-08-01                  21               101217\n",
       "1 2019-08-02                  45               266748\n",
       "2 2019-08-03                  61               380996\n",
       "3 2019-08-04                  78               489567\n",
       "4 2019-08-05                 101               581995"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# количество заказов и выручка по дням с накоплением в группе B\n",
    "\n",
    "ordersBCummulative = ordersBDaily.apply(\n",
    "    lambda x: ordersBDaily[ordersBDaily['date'] <= x['date']].agg(\n",
    "        {'date': 'max', 'ordersPerDateB': 'sum', 'revenuePerDateB': 'sum'}\n",
    "    ),\n",
    "    axis=1,\n",
    ").sort_values(by=['date'])\n",
    "ordersBCummulative.columns = [\n",
    "    'date',\n",
    "    'ordersCummulativeB',\n",
    "    'revenueCummulativeB',\n",
    "]\n",
    "\n",
    "ordersBCummulative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "85f86154",
   "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>date</th>\n",
       "      <th>ordersPerDateA</th>\n",
       "      <th>revenuePerDateA</th>\n",
       "      <th>ordersPerDateB</th>\n",
       "      <th>revenuePerDateB</th>\n",
       "      <th>ordersCummulativeA</th>\n",
       "      <th>revenueCummulativeA</th>\n",
       "      <th>ordersCummulativeB</th>\n",
       "      <th>revenueCummulativeB</th>\n",
       "      <th>visitorsPerDateA</th>\n",
       "      <th>visitorsPerDateB</th>\n",
       "      <th>visitorsCummulativeA</th>\n",
       "      <th>visitorsCummulativeB</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-08-01</td>\n",
       "      <td>24</td>\n",
       "      <td>148579</td>\n",
       "      <td>21</td>\n",
       "      <td>101217</td>\n",
       "      <td>24</td>\n",
       "      <td>148579</td>\n",
       "      <td>21</td>\n",
       "      <td>101217</td>\n",
       "      <td>719</td>\n",
       "      <td>713</td>\n",
       "      <td>719</td>\n",
       "      <td>713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-08-02</td>\n",
       "      <td>20</td>\n",
       "      <td>93822</td>\n",
       "      <td>24</td>\n",
       "      <td>165531</td>\n",
       "      <td>44</td>\n",
       "      <td>242401</td>\n",
       "      <td>45</td>\n",
       "      <td>266748</td>\n",
       "      <td>619</td>\n",
       "      <td>581</td>\n",
       "      <td>1338</td>\n",
       "      <td>1294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-08-03</td>\n",
       "      <td>24</td>\n",
       "      <td>112473</td>\n",
       "      <td>16</td>\n",
       "      <td>114248</td>\n",
       "      <td>68</td>\n",
       "      <td>354874</td>\n",
       "      <td>61</td>\n",
       "      <td>380996</td>\n",
       "      <td>507</td>\n",
       "      <td>509</td>\n",
       "      <td>1845</td>\n",
       "      <td>1803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-08-04</td>\n",
       "      <td>16</td>\n",
       "      <td>70825</td>\n",
       "      <td>17</td>\n",
       "      <td>108571</td>\n",
       "      <td>84</td>\n",
       "      <td>425699</td>\n",
       "      <td>78</td>\n",
       "      <td>489567</td>\n",
       "      <td>717</td>\n",
       "      <td>770</td>\n",
       "      <td>2562</td>\n",
       "      <td>2573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-08-05</td>\n",
       "      <td>25</td>\n",
       "      <td>124218</td>\n",
       "      <td>23</td>\n",
       "      <td>92428</td>\n",
       "      <td>109</td>\n",
       "      <td>549917</td>\n",
       "      <td>101</td>\n",
       "      <td>581995</td>\n",
       "      <td>756</td>\n",
       "      <td>707</td>\n",
       "      <td>3318</td>\n",
       "      <td>3280</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  ordersPerDateA  revenuePerDateA  ordersPerDateB  \\\n",
       "0 2019-08-01              24           148579              21   \n",
       "1 2019-08-02              20            93822              24   \n",
       "2 2019-08-03              24           112473              16   \n",
       "3 2019-08-04              16            70825              17   \n",
       "4 2019-08-05              25           124218              23   \n",
       "\n",
       "   revenuePerDateB  ordersCummulativeA  revenueCummulativeA  \\\n",
       "0           101217                  24               148579   \n",
       "1           165531                  44               242401   \n",
       "2           114248                  68               354874   \n",
       "3           108571                  84               425699   \n",
       "4            92428                 109               549917   \n",
       "\n",
       "   ordersCummulativeB  revenueCummulativeB  visitorsPerDateA  \\\n",
       "0                  21               101217               719   \n",
       "1                  45               266748               619   \n",
       "2                  61               380996               507   \n",
       "3                  78               489567               717   \n",
       "4                 101               581995               756   \n",
       "\n",
       "   visitorsPerDateB  visitorsCummulativeA  visitorsCummulativeB  \n",
       "0               713                   719                   713  \n",
       "1               581                  1338                  1294  \n",
       "2               509                  1845                  1803  \n",
       "3               770                  2562                  2573  \n",
       "4               707                  3318                  3280  "
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# финальный датасет\n",
    "\n",
    "final_data = (\n",
    "    ordersADaily.merge(\n",
    "        ordersBDaily, left_on='date', right_on='date', how='left'\n",
    "    )\n",
    "    .merge(ordersACummulative, left_on='date', right_on='date', how='left')\n",
    "    .merge(ordersBCummulative, left_on='date', right_on='date', how='left')\n",
    "    .merge(visitorsADaily, left_on='date', right_on='date', how='left')\n",
    "    .merge(visitorsBDaily, left_on='date', right_on='date', how='left')\n",
    "    .merge(visitorsACummulative, left_on='date', right_on='date', how='left')\n",
    "    .merge(visitorsBCummulative, left_on='date', right_on='date', how='left')\n",
    ")\n",
    "\n",
    "final_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68f6e623",
   "metadata": {},
   "source": [
    "Проверим по получившимся \"сырым\" данным первую гипотезу.\n",
    "\n",
    "- **H_0**: Cреднее количество заказов на посетителя в группе A статистически значимо не отличается от среднего количества заказов на посетителя в группе B\n",
    "- **H_a**: Cреднее количество заказов на посетителя в группе A статистически значимо отличается от среднего количества заказов на посетителя в группе B\n",
    "- $\\alpha$ = 0.05\n",
    "\n",
    "Так как в данных есть большие (по сравнению с нормальным распределением) выбросы, алгебраические метрики работают плохо. Для проверки гипотезы мы вызовем непараметрический U-критерий Манна-Уитни методом `st.mannwhitneyu()`.\n",
    "\n",
    "Создадим переменные `ordersByUsersA` и `ordersByUsersB` со столбцами `['userId', 'orders']`. В них для пользователей, которые заказывали хотя бы 1 раз, укажем число совершённых заказов. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "8651f725",
   "metadata": {},
   "outputs": [],
   "source": [
    "# переменные ordersByUsersA и ordersByUsersB\n",
    "\n",
    "ordersByUsersA = (\n",
    "    orders[orders['group'] == 'A']\n",
    "    .groupby('visitorId', as_index=False)\n",
    "    .agg({'transactionId': pd.Series.nunique})\n",
    ")\n",
    "ordersByUsersA.columns = ['userId', 'orders']\n",
    "\n",
    "ordersByUsersB = (\n",
    "    orders[orders['group'] == 'B']\n",
    "    .groupby('visitorId', as_index=False)\n",
    "    .agg({'transactionId': pd.Series.nunique})\n",
    ")\n",
    "ordersByUsersB.columns = ['userId', 'orders']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4be52383",
   "metadata": {},
   "source": [
    "Объявим переменные `sampleA` и `sampleB`, в которых пользователям из разных групп будет соответствовать количество заказов. Тем, кто ничего не заказал, будут соответствовать нули. Это нужно, чтобы подготовить выборки к проверке критерием Манна-Уитни."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "1ea2d3ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "# переменные sampleA и sampleB\n",
    "\n",
    "sampleA = pd.concat(\n",
    "    [\n",
    "        ordersByUsersA['orders'],\n",
    "        pd.Series(\n",
    "            0,\n",
    "            index=np.arange(\n",
    "                final_data['visitorsPerDateA'].sum() - len(ordersByUsersA['orders'])\n",
    "            ),\n",
    "            name='orders',\n",
    "        ),\n",
    "    ],\n",
    "    axis=0,\n",
    ")\n",
    "\n",
    "sampleB = pd.concat(\n",
    "    [\n",
    "        ordersByUsersB['orders'],\n",
    "        pd.Series(\n",
    "            0,\n",
    "            index=np.arange(\n",
    "                final_data['visitorsPerDateB'].sum() - len(ordersByUsersB['orders'])\n",
    "            ),\n",
    "            name='orders',\n",
    "        ),\n",
    "    ],\n",
    "    axis=0,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c146200",
   "metadata": {},
   "source": [
    "Применим критерий и выведем относительный прирост среднего числа заказов группы B."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "7462f493",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mann–Whitney U test p-value = 0.017\n",
      "Относительный прирост группы B = 13.808%\n",
      "Отвергаем нулевую гипотезу для сравнения средних групп А и B\n"
     ]
    }
   ],
   "source": [
    "# результаты теста\n",
    "\n",
    "alpha = 0.05  # критический уровень статистической значимости\n",
    "\n",
    "results = stats.mannwhitneyu(sampleA, sampleB)\n",
    "print('Mann–Whitney U test p-value =', '{0:.3f}'.format(results[1]))\n",
    "print('Относительный прирост группы B =', '{0:.3%}'.format(sampleB.mean() / sampleA.mean() - 1))\n",
    "\n",
    "if results.pvalue < alpha:\n",
    "    print('Отвергаем нулевую гипотезу для сравнения средних групп А и B')\n",
    "else:\n",
    "    print(\n",
    "        'Не получилось отвергнуть нулевую гипотезу для сравнения средних групп А и B'\n",
    "    ) "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4eaca751",
   "metadata": {},
   "source": [
    "Вероятность получить наблюдемый результат при условии, что H_0 верна, составляет менее 5%. На данном уровне значимости мы отвергаем H_0 в пользу H_a. Cреднее количество заказов на посетителя в группе A **статистически значимо отличается** от среднего количества заказов на посетителя в группе B.\n",
    "\n",
    "Относительный выигрыш группы B составляет 13.8%."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3d380d4",
   "metadata": {},
   "source": [
    "#### Статистическая значимость различий в среднем чеке заказа между группами по \"сырым\" данным"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "104673ad",
   "metadata": {},
   "source": [
    "Проверим по \"сырым\" данным вторую гипотезу.\n",
    "\n",
    "- **H_0**: Cредний чек заказа в группе A статистически значимо не отличается от среднего чека заказа в группе B\n",
    "- **H_a**: Cредний чек заказа в группе A статистически значимо отличается от среднего чека заказа в группе B\n",
    "- $\\alpha$ = 0.05\n",
    "\n",
    "Аналогично предыдущей гипотезе, будем использовать критерий Манна-Уитни."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "c1f8ff9d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mann–Whitney U test p-value = 0.729\n",
      "Относительный прирост группы B = 25.871%\n",
      "Не получилось отвергнуть нулевую гипотезу для сравнения средних групп А и B\n"
     ]
    }
   ],
   "source": [
    "# результаты теста\n",
    "\n",
    "alpha = 0.05  # критический уровень статистической значимости\n",
    "\n",
    "sampleA = orders[orders['group']=='A']['revenue']\n",
    "sampleB = orders[orders['group']=='B']['revenue']                           \n",
    "\n",
    "results = stats.mannwhitneyu(sampleA, sampleB)\n",
    "print('Mann–Whitney U test p-value =', '{0:.3f}'.format(results[1]))\n",
    "print('Относительный прирост группы B =', '{0:.3%}'.format(sampleB.mean() / sampleA.mean() - 1))\n",
    "\n",
    "if results.pvalue < alpha:\n",
    "    print('Отвергаем нулевую гипотезу для сравнения средних групп А и B')\n",
    "else:\n",
    "    print(\n",
    "        'Не получилось отвергнуть нулевую гипотезу для сравнения средних групп А и B'\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc9830bc",
   "metadata": {},
   "source": [
    "Вероятность получить наблюдемый результат при условии, что H_0 верна, составляет более 7%. На данном уровне значимости нет оснований отвергать H_0. Cредний чек заказа в группе A **статистически значимо не отличается** от среднего чека заказа в группе B."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60dffb47",
   "metadata": {},
   "source": [
    "#### Статистическая значимость различий в среднем количестве заказов на посетителя между группами по очищенным данным"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f5e1283",
   "metadata": {},
   "source": [
    "На \"сырых\" данных мы получили такие результаты:\n",
    "- среднее количество заказов на посетителя в группе A **статистически значимо отличается** от среднего количества заказов на посетителя в группе B;\n",
    "- средний чек заказа в группе A **статистически значимо не отличается** от среднего чека заказа в группе B.\n",
    "\n",
    "Однако ранее мы обнаружили выбросы с аномально высоким количеством заказов или аномально высоким чеком, а также пересечение между группами A и B, что могло исказить результат.\n",
    "\n",
    "Проведём те же тесты на очищенных данных без выбросов и пользователей, ошибочно попавших в обе группы. \n",
    "\n",
    "99-й перцентиль количества заказов на пользователя составил `4`, 99-й перцентиль стоимости заказов составил `58233.2` у. е., а общие в группах пользователи сохранены в переменной `common_ids`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "b772d5e5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "67"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# пользователи с аномально высоким количеством заказов\n",
    "\n",
    "orders_lim = np.percentile(ordersByUsers['orders'], 99)\n",
    "\n",
    "usersWithManyOrders = pd.concat(\n",
    "    [\n",
    "        ordersByUsersA[ordersByUsersA['orders'] > orders_lim]['userId'],\n",
    "        ordersByUsersB[ordersByUsersB['orders'] > orders_lim]['userId'],\n",
    "    ],\n",
    "    axis=0,\n",
    ")\n",
    "\n",
    "# пользователи с аномально высоким количеством заказов\n",
    "\n",
    "revenue_lim = np.percentile(orders['revenue'], 99)\n",
    "\n",
    "usersWithExpensiveOrders = orders[orders['revenue'] > revenue_lim]['visitorId']\n",
    "\n",
    "# пользователи в обеих группах теста\n",
    "\n",
    "usersInBothGroups = orders[orders['visitorId'].isin(common_ids)]['visitorId']\n",
    "\n",
    "# все \"лишние\" пользователи\n",
    "\n",
    "abnormalUsers = (\n",
    "    pd.concat([usersWithManyOrders, usersWithExpensiveOrders, usersInBothGroups], axis=0)\n",
    "    .drop_duplicates()\n",
    "    .sort_values()\n",
    ")\n",
    "\n",
    "abnormalUsers.head()\n",
    "abnormalUsers.shape[0] "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "885c742e",
   "metadata": {},
   "source": [
    "Всего 67 аномальных пользователей.\n",
    "\n",
    "Подготовим очищенные выборки."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "fdd34d52",
   "metadata": {},
   "outputs": [],
   "source": [
    "# переменные sampleAFiltered и sampleBFiltered\n",
    "\n",
    "sampleAFiltered = pd.concat(\n",
    "    [\n",
    "        ordersByUsersA[\n",
    "            np.logical_not(ordersByUsersA['userId'].isin(abnormalUsers))\n",
    "        ]['orders'],\n",
    "        pd.Series(\n",
    "            0,\n",
    "            index=np.arange(\n",
    "                final_data['visitorsPerDateA'].sum() - len(ordersByUsersA['orders'])\n",
    "            ),\n",
    "            name='orders',\n",
    "        ),\n",
    "    ],\n",
    "    axis=0,\n",
    ")\n",
    "\n",
    "sampleBFiltered = pd.concat(\n",
    "    [\n",
    "        ordersByUsersB[\n",
    "            np.logical_not(ordersByUsersB['userId'].isin(abnormalUsers))\n",
    "        ]['orders'],\n",
    "        pd.Series(\n",
    "            0,\n",
    "            index=np.arange(\n",
    "                final_data['visitorsPerDateB'].sum() - len(ordersByUsersB['orders'])\n",
    "            ),\n",
    "            name='orders',\n",
    "        ),\n",
    "    ],\n",
    "    axis=0,\n",
    ") "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7d094f97",
   "metadata": {},
   "source": [
    "Проверим по получившимся очищенным данным первую гипотезу.\n",
    "\n",
    "- **H_0**: Cреднее количество заказов на посетителя в группе A статистически значимо не отличается от среднего количества заказов на посетителя в группе B\n",
    "- **H_a**: Cреднее количество заказов на посетителя в группе A статистически значимо отличается от среднего количества заказов на посетителя в группе B\n",
    "- $\\alpha$ = 0.05"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "a9a51fb6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mann–Whitney U test p-value = 0.010\n",
      "Относительный прирост группы B = 16.367%\n",
      "Отвергаем нулевую гипотезу для сравнения средних групп А и B\n"
     ]
    }
   ],
   "source": [
    "# результаты теста\n",
    "\n",
    "alpha = 0.05  # критический уровень статистической значимости\n",
    "\n",
    "results = stats.mannwhitneyu(sampleAFiltered, sampleBFiltered)\n",
    "print('Mann–Whitney U test p-value =', '{0:.3f}'.format(results[1]))\n",
    "print('Относительный прирост группы B =', '{0:.3%}'.format(sampleBFiltered.mean() / sampleAFiltered.mean() - 1))\n",
    "\n",
    "if results.pvalue < alpha:\n",
    "    print('Отвергаем нулевую гипотезу для сравнения средних групп А и B')\n",
    "else:\n",
    "    print(\n",
    "        'Не получилось отвергнуть нулевую гипотезу для сравнения средних групп А и B'\n",
    "    ) "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e52036a",
   "metadata": {},
   "source": [
    "Удаление аномальных пользователей **не повлияло** на результат. Cреднее количество заказов на посетителя в группе A **статистически значимо отличается** от среднего количества заказов на посетителя в группе B.\n",
    "\n",
    "Относительный выигрыш группы B составил 16.4% (без удаления аномальных пользователей прирост был 13.8%)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19f991af",
   "metadata": {},
   "source": [
    "#### Статистическая значимость различий в среднем чеке заказа между группами по очищенным данным"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8c68a253",
   "metadata": {},
   "source": [
    "Проверим по очищенным данным вторую гипотезу.\n",
    "\n",
    "- **H_0**: Cредний чек заказа в группе A статистически значимо не отличается от среднего чека заказа в группе B\n",
    "- **H_a**: Cредний чек заказа в группе A статистически значимо отличается от среднего чека заказа в группе B\n",
    "- $\\alpha$ = 0.05"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "91681c29",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mann–Whitney U test p-value = 0.901\n",
      "Относительный прирост группы B = -3.163%\n",
      "Не получилось отвергнуть нулевую гипотезу для сравнения средних групп А и B\n"
     ]
    }
   ],
   "source": [
    "# результаты теста\n",
    "\n",
    "alpha = 0.05  # критический уровень статистической значимости\n",
    "\n",
    "sampleAFiltered = orders[\n",
    "                np.logical_and(\n",
    "                    orders['group'] == 'A',\n",
    "                    np.logical_not(orders['visitorId'].isin(abnormalUsers)),\n",
    "                )\n",
    "            ]['revenue']\n",
    "sampleBFiltered = orders[\n",
    "                np.logical_and(\n",
    "                    orders['group'] == 'B',\n",
    "                    np.logical_not(orders['visitorId'].isin(abnormalUsers)),\n",
    "                )\n",
    "            ]['revenue']                           \n",
    "\n",
    "results = stats.mannwhitneyu(sampleAFiltered, sampleBFiltered)\n",
    "print('Mann–Whitney U test p-value =', '{0:.3f}'.format(results[1]))\n",
    "print('Относительный прирост группы B =', '{0:.3%}'.format(sampleBFiltered.mean() / sampleAFiltered.mean() - 1))\n",
    "\n",
    "if results.pvalue < alpha:\n",
    "    print('Отвергаем нулевую гипотезу для сравнения средних групп А и B')\n",
    "else:\n",
    "    print(\n",
    "        'Не получилось отвергнуть нулевую гипотезу для сравнения средних групп А и B'\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "83e5a1de",
   "metadata": {},
   "source": [
    "Удаление аномальных пользователей **не повлияло** на результат. Cредний чек заказа в группе A **статистически значимо не отличается** от среднего чека заказа в группе B.\n",
    "\n",
    "Относительный проигрыш группы B составил -3.2% (без удаления аномальных пользователей прирост был 25.9%).\n",
    "\n",
    "После удаления аномалий p-value увеличилось, а прирост группы B из положительного превратился в отрицательный. Несмотря на то, что вывод после проверки гипотезы не изменился, мы видим, как сильно аномалии могут влиять на результат."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7544510",
   "metadata": {},
   "source": [
    "### Вывод"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1507ab5a",
   "metadata": {},
   "source": [
    "Проанализировав результаты A/B-теста, мы принимаем решение остановить тест и зафиксировать победу группы B, так как:\n",
    "- среднее количество заказов на посетителя в группе A **статистически значимо отличается** от среднего количества заказов на посетителя в группе B;\n",
    "- график показывает, что кумулятивное среднее количество заказов на посетителя на текущий период выровнялось и показывает стабильно более высокий результат в группе B."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3da08ec9",
   "metadata": {},
   "source": [
    "## Общий вывод"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51c00947",
   "metadata": {},
   "source": [
    "<div style=\"border:solid green 2px; padding: 20px\">\n",
    "    \n",
    "**Часть 1. Приоритизация гипотез**\n",
    "    \n",
    "Перед проведением A/B-теста мы приоритизировали 9 гипотез по увеличению выручки интернет-магазина с помощью фреймворков ICE и RICE.\n",
    "    \n",
    "Наиболее приоритетными были гипотезы:\n",
    "    \n",
    "- По фреймворку ICE: *Запустить акцию, дающую скидку на товар в день рождения*.\n",
    "    \n",
    "- По фреймворку RICE: *Добавить форму подписки на все основные страницы, чтобы собрать базу клиентов для email-рассылок*.\n",
    "    \n",
    "**Часть 2. Анализ A/B-теста**\n",
    "    \n",
    "Мы провели предобработку данных, полученных в ходе A/B-теста. \n",
    "\n",
    "Нельзя сказать, что инструмент \"деления\" трафика работал безошибочно: 58 пользователей попали в обе группы A/B-теста.\n",
    "    \n",
    "В данных обнаружились выбросы, которые мы определили по границе 99-го перцентиля:\n",
    "- аномальное количество заказов на пользователя — > 4;\n",
    "- аномальная стоимость заказов — > 58233.2 у. е.\n",
    "    \n",
    "Так как в данных есть выбросы, мы использовали непараметрический <span style=\"background-color:yellow\">U-критерий Манна-Уитни</span> для статистического анализа и проверили 2 гипотезы о различиях средних в группах A и B.\n",
    "    \n",
    "1. <span style=\"background-color:yellow\">Cреднее количество заказов на посетителя в группе A статистически значимо отличается от среднего количества заказов на посетителя в группе B</span> как по \"сырым\", так и по очищенным от аномалий данным. Относительный выигрыш группы B составил 13.8% по \"сырым\" данным и 16.4% по очищенным данным.\n",
    "    \n",
    "2. <span style=\"background-color:yellow\">Cредний чек заказа в группе A статистически значимо не отличается от среднего чека заказа в группе B</span> как по \"сырым\", так и по очищенным от аномалий данным.\n",
    "    \n",
    "Таким образом, мы можем остановить тест и зафиксировать победу группы B."
   ]
  }
 ],
 "metadata": {
  "ExecuteTimeLog": [
   {
    "duration": 422,
    "start_time": "2023-03-29T10:52:30.457Z"
   },
   {
    "duration": 51,
    "start_time": "2023-03-29T10:52:30.882Z"
   },
   {
    "duration": 23,
    "start_time": "2023-03-29T10:52:30.935Z"
   },
   {
    "duration": 41,
    "start_time": "2023-03-29T10:52:30.961Z"
   },
   {
    "duration": 871,
    "start_time": "2023-03-29T10:52:31.003Z"
   },
   {
    "duration": 89,
    "start_time": "2023-03-29T10:52:31.875Z"
   },
   {
    "duration": 14,
    "start_time": "2023-03-29T10:52:31.966Z"
   },
   {
    "duration": 88,
    "start_time": "2023-03-29T10:52:31.982Z"
   },
   {
    "duration": 53,
    "start_time": "2023-03-29T10:52:32.071Z"
   },
   {
    "duration": 27,
    "start_time": "2023-03-29T10:52:32.127Z"
   },
   {
    "duration": 71,
    "start_time": "2023-03-29T10:52:32.155Z"
   },
   {
    "duration": 45,
    "start_time": "2023-03-29T10:52:32.228Z"
   },
   {
    "duration": 200,
    "start_time": "2023-03-29T10:52:32.275Z"
   },
   {
    "duration": 129,
    "start_time": "2023-03-29T10:52:32.477Z"
   },
   {
    "duration": 15,
    "start_time": "2023-03-29T10:52:32.608Z"
   },
   {
    "duration": 7,
    "start_time": "2023-03-29T10:52:32.625Z"
   },
   {
    "duration": 33,
    "start_time": "2023-03-29T10:52:32.633Z"
   },
   {
    "duration": 258,
    "start_time": "2023-03-29T10:52:32.668Z"
   },
   {
    "duration": 184,
    "start_time": "2023-03-29T10:52:32.928Z"
   },
   {
    "duration": 9,
    "start_time": "2023-03-29T10:52:33.114Z"
   },
   {
    "duration": 310,
    "start_time": "2023-03-29T10:52:33.124Z"
   },
   {
    "duration": 20,
    "start_time": "2023-03-29T10:52:33.436Z"
   },
   {
    "duration": 18,
    "start_time": "2023-03-29T10:52:33.457Z"
   },
   {
    "duration": 208,
    "start_time": "2023-03-29T10:52:33.476Z"
   },
   {
    "duration": 11,
    "start_time": "2023-03-29T10:52:33.685Z"
   },
   {
    "duration": 302,
    "start_time": "2023-03-29T10:52:33.697Z"
   },
   {
    "duration": 9,
    "start_time": "2023-03-29T10:52:34.000Z"
   },
   {
    "duration": 155,
    "start_time": "2023-03-29T10:52:34.012Z"
   },
   {
    "duration": 113,
    "start_time": "2023-03-29T10:52:34.169Z"
   },
   {
    "duration": 4,
    "start_time": "2023-03-29T10:52:34.284Z"
   },
   {
    "duration": 38,
    "start_time": "2023-03-29T10:52:34.290Z"
   },
   {
    "duration": 156,
    "start_time": "2023-03-29T10:52:34.329Z"
   },
   {
    "duration": 130,
    "start_time": "2023-03-29T10:52:34.486Z"
   },
   {
    "duration": 186,
    "start_time": "2023-03-29T10:52:34.618Z"
   },
   {
    "duration": 4,
    "start_time": "2023-03-29T10:52:34.805Z"
   },
   {
    "duration": 7,
    "start_time": "2023-03-29T10:52:34.811Z"
   },
   {
    "duration": 9,
    "start_time": "2023-03-29T10:52:34.819Z"
   },
   {
    "duration": 67,
    "start_time": "2023-03-29T10:52:34.829Z"
   },
   {
    "duration": 12,
    "start_time": "2023-03-29T10:52:34.897Z"
   },
   {
    "duration": 47,
    "start_time": "2023-03-29T10:52:34.911Z"
   },
   {
    "duration": 20,
    "start_time": "2023-03-29T10:52:34.960Z"
   },
   {
    "duration": 44,
    "start_time": "2023-03-29T10:52:34.981Z"
   },
   {
    "duration": 34,
    "start_time": "2023-03-29T10:52:35.027Z"
   },
   {
    "duration": 47,
    "start_time": "2023-03-29T10:52:35.063Z"
   },
   {
    "duration": 34,
    "start_time": "2023-03-29T10:52:35.112Z"
   },
   {
    "duration": 181,
    "start_time": "2023-03-29T10:52:35.148Z"
   },
   {
    "duration": 6,
    "start_time": "2023-03-29T10:52:35.331Z"
   },
   {
    "duration": 62,
    "start_time": "2023-03-29T10:52:35.338Z"
   },
   {
    "duration": 10,
    "start_time": "2023-03-29T10:52:35.402Z"
   },
   {
    "duration": 25,
    "start_time": "2023-03-29T10:52:35.414Z"
   },
   {
    "duration": 35,
    "start_time": "2023-03-29T10:52:35.440Z"
   },
   {
    "duration": 85,
    "start_time": "2023-03-29T10:52:35.476Z"
   },
   {
    "duration": 24,
    "start_time": "2023-03-29T10:52:35.563Z"
   },
   {
    "duration": 976,
    "start_time": "2023-03-31T07:44:33.264Z"
   },
   {
    "duration": 113,
    "start_time": "2023-03-31T07:44:34.242Z"
   },
   {
    "duration": 12,
    "start_time": "2023-03-31T07:44:34.357Z"
   },
   {
    "duration": 9,
    "start_time": "2023-03-31T07:44:34.371Z"
   },
   {
    "duration": 45,
    "start_time": "2023-03-31T07:44:34.382Z"
   },
   {
    "duration": 15,
    "start_time": "2023-03-31T07:44:34.434Z"
   },
   {
    "duration": 250,
    "start_time": "2023-03-31T07:44:34.450Z"
   },
   {
    "duration": 141,
    "start_time": "2023-03-31T07:44:34.701Z"
   },
   {
    "duration": 12,
    "start_time": "2023-03-31T07:44:34.843Z"
   },
   {
    "duration": 30,
    "start_time": "2023-03-31T07:44:34.856Z"
   },
   {
    "duration": 10,
    "start_time": "2023-03-31T07:44:34.887Z"
   },
   {
    "duration": 7,
    "start_time": "2023-03-31T07:44:34.899Z"
   },
   {
    "duration": 12,
    "start_time": "2023-03-31T07:44:34.907Z"
   },
   {
    "duration": 8,
    "start_time": "2023-03-31T07:44:34.922Z"
   },
   {
    "duration": 176,
    "start_time": "2023-03-31T07:44:34.931Z"
   },
   {
    "duration": 110,
    "start_time": "2023-03-31T07:44:35.108Z"
   },
   {
    "duration": 14,
    "start_time": "2023-03-31T07:44:35.220Z"
   },
   {
    "duration": 15,
    "start_time": "2023-03-31T07:44:35.248Z"
   },
   {
    "duration": 17,
    "start_time": "2023-03-31T07:44:35.265Z"
   },
   {
    "duration": 227,
    "start_time": "2023-03-31T07:44:35.283Z"
   },
   {
    "duration": 162,
    "start_time": "2023-03-31T07:44:35.512Z"
   },
   {
    "duration": 10,
    "start_time": "2023-03-31T07:44:35.676Z"
   },
   {
    "duration": 276,
    "start_time": "2023-03-31T07:44:35.687Z"
   },
   {
    "duration": 10,
    "start_time": "2023-03-31T07:44:35.964Z"
   },
   {
    "duration": 22,
    "start_time": "2023-03-31T07:44:35.976Z"
   },
   {
    "duration": 193,
    "start_time": "2023-03-31T07:44:35.999Z"
   },
   {
    "duration": 11,
    "start_time": "2023-03-31T07:44:36.194Z"
   },
   {
    "duration": 294,
    "start_time": "2023-03-31T07:44:36.207Z"
   },
   {
    "duration": 10,
    "start_time": "2023-03-31T07:44:36.503Z"
   },
   {
    "duration": 117,
    "start_time": "2023-03-31T07:44:36.515Z"
   },
   {
    "duration": 117,
    "start_time": "2023-03-31T07:44:36.633Z"
   },
   {
    "duration": 5,
    "start_time": "2023-03-31T07:44:36.751Z"
   },
   {
    "duration": 24,
    "start_time": "2023-03-31T07:44:36.757Z"
   },
   {
    "duration": 184,
    "start_time": "2023-03-31T07:44:36.782Z"
   },
   {
    "duration": 126,
    "start_time": "2023-03-31T07:44:36.967Z"
   },
   {
    "duration": 113,
    "start_time": "2023-03-31T07:44:37.095Z"
   },
   {
    "duration": 4,
    "start_time": "2023-03-31T07:44:37.210Z"
   },
   {
    "duration": 11,
    "start_time": "2023-03-31T07:44:37.215Z"
   },
   {
    "duration": 22,
    "start_time": "2023-03-31T07:44:37.227Z"
   },
   {
    "duration": 48,
    "start_time": "2023-03-31T07:44:37.251Z"
   },
   {
    "duration": 8,
    "start_time": "2023-03-31T07:44:37.302Z"
   },
   {
    "duration": 54,
    "start_time": "2023-03-31T07:44:37.311Z"
   },
   {
    "duration": 15,
    "start_time": "2023-03-31T07:44:37.366Z"
   },
   {
    "duration": 45,
    "start_time": "2023-03-31T07:44:37.382Z"
   },
   {
    "duration": 28,
    "start_time": "2023-03-31T07:44:37.428Z"
   },
   {
    "duration": 47,
    "start_time": "2023-03-31T07:44:37.458Z"
   },
   {
    "duration": 22,
    "start_time": "2023-03-31T07:44:37.506Z"
   },
   {
    "duration": 132,
    "start_time": "2023-03-31T07:44:37.530Z"
   },
   {
    "duration": 5,
    "start_time": "2023-03-31T07:44:37.664Z"
   },
   {
    "duration": 21,
    "start_time": "2023-03-31T07:44:37.670Z"
   },
   {
    "duration": 8,
    "start_time": "2023-03-31T07:44:37.693Z"
   },
   {
    "duration": 21,
    "start_time": "2023-03-31T07:44:37.703Z"
   },
   {
    "duration": 24,
    "start_time": "2023-03-31T07:44:37.725Z"
   },
   {
    "duration": 19,
    "start_time": "2023-03-31T07:44:37.751Z"
   },
   {
    "duration": 9,
    "start_time": "2023-03-31T07:44:37.772Z"
   }
  ],
  "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": "3.9.13"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}