github_final_AB_project_ya_ru/github_final_AB_project_ya_ru.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-0cpnmjXCKTH"
      },
      "source": [
        "# Проект по A/B-тестированию\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QN4lZ1S8WiI9"
      },
      "source": [
        "## Описание"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CWQyhZJUWlKM"
      },
      "source": [
        "Задача — провести оценку результатов A/B-теста. В нашем распоряжении есть датасет с действиями пользователей, техническое задание и несколько вспомогательных датасетов.\n",
        "\n",
        "- Оцените корректность проведения теста\n",
        "- Проанализируйте результаты теста"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cYhfvc_-Wvw7"
      },
      "source": [
        "## Техническое задание"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SrlXU8rnWydo"
      },
      "source": [
        "- Название теста: `recommender_system_test`;\n",
        "- группы: А — контрольная, B — новая платёжная воронка;\n",
        "- дата запуска: 2020-12-07;\n",
        "- дата остановки набора новых пользователей: 2020-12-21;\n",
        "- дата остановки: 2021-01-04;\n",
        "- аудитория: 15% новых пользователей из региона EU;\n",
        "- назначение теста: тестирование изменений, связанных с внедрением улучшенной рекомендательной системы;\n",
        "- ожидаемое количество участников теста: 6000.\n",
        "- ожидаемый эффект: за 14 дней с момента регистрации пользователи покажут улучшение каждой метрики не менее, чем на 10%:\n",
        "    - конверсии в просмотр карточек товаров — событие `product_page`,\n",
        "    - просмотры корзины — `product_cart`,\n",
        "    - покупки — `purchase`."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kCok7QiiW1dd"
      },
      "source": [
        "## Описание данных"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QuSDJTiDW8gz"
      },
      "source": [
        "`ab_project_marketing_events.csv` — календарь маркетинговых событий на 2020 год.\n",
        "\n",
        "Структура файла:\n",
        "\n",
        "- `name` — название маркетингового события;\n",
        "- `regions` — регионы, в которых будет проводиться рекламная кампания;\n",
        "- `start_dt` — дата начала кампании;\n",
        "- `finish_dt` — дата завершения кампании.\n",
        "\n",
        "`final_ab_new_users.csv` — пользователи, зарегистрировавшиеся с 7 по 21 декабря 2020 года.\n",
        "\n",
        "Структура файла:\n",
        "\n",
        "- `user_id` — идентификатор пользователя;\n",
        "- `first_date` — дата регистрации;\n",
        "- `region` — регион пользователя;\n",
        "- `device` — устройство, с которого происходила регистрация.\n",
        "\n",
        "`final_ab_events.csv` — действия новых пользователей в период с 7 декабря 2020 по 4 января 2021 года.\n",
        "\n",
        "Структура файла:\n",
        "\n",
        "- `user_id` — идентификатор пользователя;\n",
        "- `event_dt` — дата и время покупки;\n",
        "- `event_name` — тип события;\n",
        "- `details` — дополнительные данные о событии. Например, для покупок, `purchase,` в этом поле хранится стоимость покупки в долларах.\n",
        "\n",
        "`final_ab_participants.csv` — таблица участников тестов.\n",
        "\n",
        "Структура файла:\n",
        "\n",
        "- `user_id` — идентификатор пользователя;\n",
        "- `ab_test` — название теста;\n",
        "- `group` — группа пользователя."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rKwkDSCjW-Sy"
      },
      "source": [
        "## Откроем и изучим общую информацию о данных"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iSb1d5zHc6R6",
        "outputId": "3006ebe7-58f3-48ea-93c5-e3339b834f88",
        "scrolled": true,
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1nnwvut4fwpwwQb8uX92Qx_2GlxlDrZdK\n",
            "To: /content/final_ab_events.csv\n",
            "100% 22.3M/22.3M [00:00<00:00, 365MB/s]\n",
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1x6igMvlG7Ev_7K1XbGv31i1W8ceie1bw\n",
            "To: /content/ab_project_marketing_events.csv\n",
            "100% 879/879 [00:00<00:00, 1.77MB/s]\n",
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1zHpPW-7NcnazhY3FMBIjO4VqRLuZ2BVJ\n",
            "To: /content/final_ab_new_users.csv\n",
            "100% 2.37M/2.37M [00:00<00:00, 234MB/s]\n",
            "Downloading...\n",
            "From: https://drive.google.com/uc?id=1v5NrEAJLTrDpp43wiB9h90Y48hl38Tdj\n",
            "To: /content/final_ab_participants.csv\n",
            "100% 716k/716k [00:00<00:00, 113MB/s]\n"
          ]
        }
      ],
      "source": [
        "! gdown -- 1nnwvut4fwpwwQb8uX92Qx_2GlxlDrZdK\n",
        "! gdown -- 1x6igMvlG7Ev_7K1XbGv31i1W8ceie1bw\n",
        "! gdown -- 1zHpPW-7NcnazhY3FMBIjO4VqRLuZ2BVJ\n",
        "! gdown -- 1v5NrEAJLTrDpp43wiB9h90Y48hl38Tdj"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "XuWhzLD8bmNM",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "import plotly.express as px \n",
        "import plotly.graph_objects as go\n",
        "import numpy as np\n",
        "import pylab \n",
        "import math as mth\n",
        "from datetime import datetime, timedelta\n",
        "\n",
        "from scipy import stats as st\n",
        "import statsmodels.sandbox.stats.multicomp as multi_correction\n",
        "\n",
        "pd.set_option('display.float_format', '{:,.2f}'.format)\n",
        "pd.set_option('display.max_colwidth', None)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def open(df_google_csv, df_ya_csv):\n",
        "    try:\n",
        "        df = pd.read_csv(df_google_csv)\n",
        "    except:\n",
        "        df = pd.read_csv(df_ya_csv)\n",
        "    display(df.info(), df.head(3))\n",
        "    return df  \n",
        "    "
      ],
      "metadata": {
        "id": "TQUKlDiXrvwV"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "events = open('/content/final_ab_events.csv', 'https://code.s3.yandex.net/datasets/final_ab_events.csv')"
      ],
      "metadata": {
        "id": "tmi6j8cWF-3S",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "9884db46-6ce5-4c82-b7c7-9fad1300e09d"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 440317 entries, 0 to 440316\n",
            "Data columns (total 4 columns):\n",
            " #   Column      Non-Null Count   Dtype  \n",
            "---  ------      --------------   -----  \n",
            " 0   user_id     440317 non-null  object \n",
            " 1   event_dt    440317 non-null  object \n",
            " 2   event_name  440317 non-null  object \n",
            " 3   details     62740 non-null   float64\n",
            "dtypes: float64(1), object(3)\n",
            "memory usage: 13.4+ MB\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "None"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "            user_id             event_dt event_name  details\n",
              "0  E1BDDCE0DAFA2679  2020-12-07 20:22:03   purchase    99.99\n",
              "1  7B6452F081F49504  2020-12-07 09:22:53   purchase     9.99\n",
              "2  9CD9F34546DF254C  2020-12-07 12:59:29   purchase     4.99"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9aaaa187-bef4-4e80-8a58-945e414aa279\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>event_dt</th>\n",
              "      <th>event_name</th>\n",
              "      <th>details</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>E1BDDCE0DAFA2679</td>\n",
              "      <td>2020-12-07 20:22:03</td>\n",
              "      <td>purchase</td>\n",
              "      <td>99.99</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>7B6452F081F49504</td>\n",
              "      <td>2020-12-07 09:22:53</td>\n",
              "      <td>purchase</td>\n",
              "      <td>9.99</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>9CD9F34546DF254C</td>\n",
              "      <td>2020-12-07 12:59:29</td>\n",
              "      <td>purchase</td>\n",
              "      <td>4.99</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9aaaa187-bef4-4e80-8a58-945e414aa279')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-9aaaa187-bef4-4e80-8a58-945e414aa279 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-9aaaa187-bef4-4e80-8a58-945e414aa279');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "users = open('/content/final_ab_new_users.csv', 'https://code.s3.yandex.net/datasets/final_ab_new_users.csv') "
      ],
      "metadata": {
        "id": "qo3IgcmtA7dC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "1886c87f-6702-428a-e365-5f83a7a510b6"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 61733 entries, 0 to 61732\n",
            "Data columns (total 4 columns):\n",
            " #   Column      Non-Null Count  Dtype \n",
            "---  ------      --------------  ----- \n",
            " 0   user_id     61733 non-null  object\n",
            " 1   first_date  61733 non-null  object\n",
            " 2   region      61733 non-null  object\n",
            " 3   device      61733 non-null  object\n",
            "dtypes: object(4)\n",
            "memory usage: 1.9+ MB\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "None"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "            user_id  first_date     region   device\n",
              "0  D72A72121175D8BE  2020-12-07         EU       PC\n",
              "1  F1C668619DFE6E65  2020-12-07  N.America  Android\n",
              "2  2E1BF1D4C37EA01F  2020-12-07         EU       PC"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-482e6b8b-5936-403f-95ec-dde0f2952b4f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>first_date</th>\n",
              "      <th>region</th>\n",
              "      <th>device</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>D72A72121175D8BE</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>EU</td>\n",
              "      <td>PC</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>F1C668619DFE6E65</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>N.America</td>\n",
              "      <td>Android</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2E1BF1D4C37EA01F</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>EU</td>\n",
              "      <td>PC</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-482e6b8b-5936-403f-95ec-dde0f2952b4f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-482e6b8b-5936-403f-95ec-dde0f2952b4f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-482e6b8b-5936-403f-95ec-dde0f2952b4f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "participants = open('/content/final_ab_participants.csv', 'https://code.s3.yandex.net/datasets/final_ab_participants.csv')"
      ],
      "metadata": {
        "id": "GX-ovonfA7uv",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 346
        },
        "outputId": "01d01fb4-9272-44cb-997d-46bc7ad9b62a"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 18268 entries, 0 to 18267\n",
            "Data columns (total 3 columns):\n",
            " #   Column   Non-Null Count  Dtype \n",
            "---  ------   --------------  ----- \n",
            " 0   user_id  18268 non-null  object\n",
            " 1   group    18268 non-null  object\n",
            " 2   ab_test  18268 non-null  object\n",
            "dtypes: object(3)\n",
            "memory usage: 428.3+ KB\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "None"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "            user_id group                  ab_test\n",
              "0  D1ABA3E2887B6A73     A  recommender_system_test\n",
              "1  A7A3664BD6242119     A  recommender_system_test\n",
              "2  DABC14FDDFADD29E     A  recommender_system_test"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5258f24a-5aec-4cae-9ab9-4825c34e0edf\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>group</th>\n",
              "      <th>ab_test</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>D1ABA3E2887B6A73</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>A7A3664BD6242119</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>DABC14FDDFADD29E</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5258f24a-5aec-4cae-9ab9-4825c34e0edf')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5258f24a-5aec-4cae-9ab9-4825c34e0edf button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5258f24a-5aec-4cae-9ab9-4825c34e0edf');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "marketing = open('/content/ab_project_marketing_events.csv', 'https://code.s3.yandex.net/datasets/ab_project_marketing_events.csv')"
      ],
      "metadata": {
        "id": "gjEuXZZNA7_g",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "6a85a5ee-1411-4000-e0be-da2a3b1dbe9c"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 14 entries, 0 to 13\n",
            "Data columns (total 4 columns):\n",
            " #   Column     Non-Null Count  Dtype \n",
            "---  ------     --------------  ----- \n",
            " 0   name       14 non-null     object\n",
            " 1   regions    14 non-null     object\n",
            " 2   start_dt   14 non-null     object\n",
            " 3   finish_dt  14 non-null     object\n",
            "dtypes: object(4)\n",
            "memory usage: 576.0+ bytes\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "None"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "                           name                   regions    start_dt  \\\n",
              "0      Christmas&New Year Promo             EU, N.America  2020-12-25   \n",
              "1  St. Valentine's Day Giveaway  EU, CIS, APAC, N.America  2020-02-14   \n",
              "2        St. Patric's Day Promo             EU, N.America  2020-03-17   \n",
              "\n",
              "    finish_dt  \n",
              "0  2021-01-03  \n",
              "1  2020-02-16  \n",
              "2  2020-03-19  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-008deaab-a96b-4ac7-be39-ca633b9641d9\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>name</th>\n",
              "      <th>regions</th>\n",
              "      <th>start_dt</th>\n",
              "      <th>finish_dt</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Christmas&amp;New Year Promo</td>\n",
              "      <td>EU, N.America</td>\n",
              "      <td>2020-12-25</td>\n",
              "      <td>2021-01-03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>St. Valentine's Day Giveaway</td>\n",
              "      <td>EU, CIS, APAC, N.America</td>\n",
              "      <td>2020-02-14</td>\n",
              "      <td>2020-02-16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>St. Patric's Day Promo</td>\n",
              "      <td>EU, N.America</td>\n",
              "      <td>2020-03-17</td>\n",
              "      <td>2020-03-19</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-008deaab-a96b-4ac7-be39-ca633b9641d9')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-008deaab-a96b-4ac7-be39-ca633b9641d9 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-008deaab-a96b-4ac7-be39-ca633b9641d9');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VNoOliKMtSrZ"
      },
      "source": [
        "Данные открылись без замечаний. В ряде столбцов нужно поменять типы."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CovXt6y6XSnn"
      },
      "source": [
        "## Выполним предобработку данных"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ml5pC1Juh8iV",
        "outputId": "3744d2e0-9c13-419d-f938-ba55c6dfb622",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user_id      0.00\n",
              "first_date   0.00\n",
              "region       0.00\n",
              "device       0.00\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "users.isna().mean()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fmUTI0PnjNLM",
        "outputId": "5910bb3b-f208-466f-d21a-f600369ba0a9",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ],
      "source": [
        "users.duplicated().sum()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "82eUCA5xh_61",
        "outputId": "75d56ccd-678f-49c5-c4e7-7947439c890e",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user_id      0.00\n",
              "event_dt     0.00\n",
              "event_name   0.00\n",
              "details      0.86\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ],
      "source": [
        "events.isna().mean()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "V_eie5COi-4V",
        "outputId": "1b9aa33f-a0c0-4898-df09-d67ef91c8b1b",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ],
      "source": [
        "events.duplicated().sum()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-7d67JRbiC6a",
        "outputId": "75ebdb51-d8d7-4717-df3a-84437be6c70b",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user_id   0.00\n",
              "group     0.00\n",
              "ab_test   0.00\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "participants.isna().mean()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-4IGz-eyjRmW",
        "outputId": "a89213bf-7b62-4dc2-ed48-4ad0e71688aa",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ],
      "source": [
        "participants.duplicated().sum()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qc_zg5xDiHpZ",
        "outputId": "4f9a6937-a452-4374-cdd9-c742378c2e47",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "name        0.00\n",
              "regions     0.00\n",
              "start_dt    0.00\n",
              "finish_dt   0.00\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ],
      "source": [
        "marketing.isna().mean()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "17WgUHMSjXsu",
        "outputId": "35b29b44-1212-4b5d-9659-4ba6398bc209",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "marketing.duplicated().sum()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "AfPrGrG2mFzC",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "users['first_date'] = pd.to_datetime(users['first_date'], format=('%Y-%m-%d'))\n",
        "\n",
        "events['event_dt_date'] = pd.to_datetime(events['event_dt']).dt.date\n",
        "events['event_dt_date'] = pd.to_datetime(events['event_dt_date'], format=('%Y-%m-%d'))\n",
        "\n",
        "marketing['start_dt'] =  pd.to_datetime(marketing['start_dt'], format=('%Y-%m-%d'))\n",
        "marketing['finish_dt'] =  pd.to_datetime(marketing['finish_dt'], format=('%Y-%m-%d'))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R5BUGa4gtsLT"
      },
      "source": [
        "Проверили датафреймы на пропуски, явные дубликаты, поменяли тип в датах. Много пропусков в деталях событий. Они не повлияют на дальнейшее исследование."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n-IgoqQCXX2X"
      },
      "source": [
        "## Оценим корректность проведения теста"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LlWrjt8hXiW2"
      },
      "source": [
        "### Проверим корректность всех пунктов технического задания"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6NuSbByqxoEV",
        "outputId": "f918abdc-4af5-4812-a3d5-fe2965adc019",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Всего в датасете участников 18268 строк\n"
          ]
        }
      ],
      "source": [
        "print('Всего в датасете участников', len(participants), 'строк')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JHmn9U5mwlW8",
        "outputId": "4f5c920e-4e05-45c1-a6cb-3a66d2a166b0",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "В этих строках 16666 уникальных участников\n"
          ]
        }
      ],
      "source": [
        "print('В этих строках', participants['user_id'].nunique(), 'уникальных участников')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "lxBO_VIxxy9W",
        "outputId": "8b685c09-4c9e-4ea4-9417-deb3be73fa56",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Одинаковых пользователей для тестов recommender_system_test и interface_eu_test: 1602'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 19
        }
      ],
      "source": [
        "all_group = set(participants.query('ab_test == \"recommender_system_test\"')['user_id']) & set(participants.query('ab_test == \"interface_eu_test\"')['user_id']) \n",
        "\n",
        "f'Одинаковых пользователей для тестов recommender_system_test и interface_eu_test: {len(all_group)}'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "dG6xb4A55P_t",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "participants = participants.drop_duplicates('user_id')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_6xhmRe-wgzb"
      },
      "source": [
        "Удалили пересекающихся пользователей."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "F_lc6UNq5hkZ",
        "outputId": "ca7e3347-3b0f-4e1c-cbcb-4b4ec60593c7",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Одинаковых пользователей для тестов recommender_system_test и interface_eu_test: 0'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 21
        }
      ],
      "source": [
        "all_group = set(participants.query('ab_test == \"recommender_system_test\"')['user_id']) & set(participants.query('ab_test == \"interface_eu_test\"')['user_id']) \n",
        "\n",
        "f'Одинаковых пользователей для тестов recommender_system_test и interface_eu_test: {len(all_group)}'"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style('dark')\n",
        "\n",
        "plt.figure(figsize=(18, 6))\n",
        "\n",
        "sns.barplot(x='group', y='user_id', data=participants.groupby(['ab_test', 'group'])['user_id'].nunique().reset_index(), hue='ab_test')\n",
        "\n",
        "plt.title('Распределение уникальных участников по группам тестов')\n",
        "plt.xlabel('Тест и группа')\n",
        "plt.ylabel('Количество участников')\n",
        "plt.legend(loc='upper right', fontsize=10)\n",
        "plt.grid()\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "id": "82rfEkKmk7NM",
        "outputId": "dcddbe55-c6be-4f37-f186-69190400d6d1"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1296x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Группы смежного теста распределены поровну. В контрольной группе нашего теста `recommender_system_test` пользователей немного больше, чем в тестовой группе."
      ],
      "metadata": {
        "id": "jSDiLPX0lo1O"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Сохраним пользователей смежного теста, чтобы исключить их смешивание с пользователями нашего теста в остальных датафреймах."
      ],
      "metadata": {
        "id": "njs73WelmEso"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "users_the_othert_test = participants[participants['ab_test'] != 'recommender_system_test']['user_id'].tolist()"
      ],
      "metadata": {
        "id": "Qzh3nDUjmSMh"
      },
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EOGPI8RBwlfg",
        "outputId": "ba298c33-2453-431f-9c12-7839e58ef689",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ab_test                  group\n",
              "interface_eu_test        A        5012\n",
              "                         B        4953\n",
              "recommender_system_test  A        3824\n",
              "                         B        2877\n",
              "Name: user_id, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ],
      "source": [
        "participants.groupby(['ab_test', 'group'])['user_id'].count()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CgFMJAHewqlk"
      },
      "source": [
        "В датафрейме участники двух тестов. Удалим второй, чтобы гарантированно избавиться от пересечений по тестам."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "3Jyjb3kn2U03",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "participants= participants[participants['ab_test'] == 'recommender_system_test'].copy()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uPkiJovN2t9o",
        "outputId": "58aeb487-dcee-48eb-ffd9-b37550594594",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ab_test                  group\n",
              "recommender_system_test  A        3824\n",
              "                         B        2877\n",
              "Name: user_id, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ],
      "source": [
        "participants.groupby(['ab_test', 'group'])['user_id'].count()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cDpSKLdZw3kd"
      },
      "source": [
        "Выборки распределены неравномерно. В контрольной группе на 1 тысячу участников больше, чем в тестовой.\n",
        "\n",
        "Еще раз убедимся, что нет пересечений пользователей."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "8ySc41XNJmL7",
        "outputId": "8fde8015-488e-49a8-9ab9-ff73545eac7a",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Одинаковых пользователей для групп A и B: 0'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 27
        }
      ],
      "source": [
        "all_group = set(participants.query('group == \"A\"')['user_id']) & set(participants.query('group == \"B\"')['user_id']) \n",
        "\n",
        "f'Одинаковых пользователей для групп A и B: {len(all_group)}'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UR91zXH0290F",
        "outputId": "c74f8663-f6e6-4fc6-807e-fe9109075dbf",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "В тесте \"recommender_system_test\" 6701 уникальных пользователей\n"
          ]
        }
      ],
      "source": [
        "print('В тесте \"recommender_system_test\"', participants['user_id'].nunique(), 'уникальных пользователей')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "98g39VKniDnk"
      },
      "source": [
        "Рассмотрим датафрейм с пользователями со всего мира."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cWi-PezTx0z9"
      },
      "source": [
        "Исследуем регионы."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QYHUX2gR6eAp",
        "outputId": "a3b31c08-8302-4c84-bd8a-336a0e373a18",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "EU           46270\n",
              "N.America     9155\n",
              "CIS           3155\n",
              "APAC          3153\n",
              "Name: region, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ],
      "source": [
        "total_users = users.copy()\n",
        "\n",
        "total_users_by_regions = total_users['region'].value_counts()\n",
        "total_users_by_regions"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ya9cOojp2_A4"
      },
      "source": [
        "Оставим только европейских пользователей, так как регион проведения теста - Европа."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "E4wcgEBC-6NM",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "users = users[users['region'] == 'EU'].copy()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B3p0CSmzybGo"
      },
      "source": [
        "Исследуем даты регистрации европейских пользователей. Для начала избавимся от пользователей из смежного теста, если такие есть."
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "before = users['user_id'].nunique()"
      ],
      "metadata": {
        "id": "2yIaK-43ol4S"
      },
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "users = users.query('user_id not in @users_the_othert_test').copy()"
      ],
      "metadata": {
        "id": "6_jORiUOoJem"
      },
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "after = users['user_id'].nunique()"
      ],
      "metadata": {
        "id": "Kn1nEfseotsr"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "if before == after:\n",
        "  print('Количество уникальных пользователей не изменилось -', after)\n",
        "else:\n",
        "  print('Количество уникальных пользователей уменьшилось. Было -', before, 'Стало -', after,\\\n",
        "        'Изменения -', round(((after / before * 100) - 100), 2), '%')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-UnmYm9Po-hR",
        "outputId": "2e4a7b5d-32ca-4bf2-8c87-a86e9ccfec81"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Количество уникальных пользователей уменьшилось. Было - 46270 Стало - 36305 Изменения - -21.54 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 582
        },
        "id": "sydWdR1f7LvS",
        "outputId": "af4e12a6-f6ca-454b-b4dc-28afd6df572d",
        "scrolled": true,
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        index  first_date\n",
              "2  2020-12-07        3337\n",
              "8  2020-12-08        1900\n",
              "14 2020-12-09        1258\n",
              "9  2020-12-10        1888\n",
              "13 2020-12-11        1415\n",
              "5  2020-12-12        2290\n",
              "3  2020-12-13        2773\n",
              "1  2020-12-14        3356\n",
              "10 2020-12-15        1797\n",
              "15 2020-12-16        1258\n",
              "11 2020-12-17        1782\n",
              "7  2020-12-18        2016\n",
              "6  2020-12-19        2119\n",
              "4  2020-12-20        2539\n",
              "0  2020-12-21        3649\n",
              "12 2020-12-22        1705\n",
              "16 2020-12-23        1223"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5d919756-c4ae-484c-95f0-4e01a3b6d9ea\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>index</th>\n",
              "      <th>first_date</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>3337</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>2020-12-08</td>\n",
              "      <td>1900</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>2020-12-09</td>\n",
              "      <td>1258</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>2020-12-10</td>\n",
              "      <td>1888</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>2020-12-11</td>\n",
              "      <td>1415</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2020-12-12</td>\n",
              "      <td>2290</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2020-12-13</td>\n",
              "      <td>2773</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2020-12-14</td>\n",
              "      <td>3356</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>2020-12-15</td>\n",
              "      <td>1797</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>2020-12-16</td>\n",
              "      <td>1258</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>2020-12-17</td>\n",
              "      <td>1782</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>2020-12-18</td>\n",
              "      <td>2016</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>2020-12-19</td>\n",
              "      <td>2119</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2020-12-20</td>\n",
              "      <td>2539</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2020-12-21</td>\n",
              "      <td>3649</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>2020-12-22</td>\n",
              "      <td>1705</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>2020-12-23</td>\n",
              "      <td>1223</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5d919756-c4ae-484c-95f0-4e01a3b6d9ea')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5d919756-c4ae-484c-95f0-4e01a3b6d9ea button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5d919756-c4ae-484c-95f0-4e01a3b6d9ea');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ],
      "source": [
        "users['first_date'].value_counts().reset_index().sort_values(by = 'index')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k_4Avg1Bym8C"
      },
      "source": [
        "Согласно ТЗ, в тесте участвуют пользователи, зарегистрировавшиеся не позднее 2020-12-21. Отсечем лишнее."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {
        "id": "ntzJNgo9_OY9",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "users = users[users['first_date'] <= '2020-12-21']"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 519
        },
        "id": "0SD19H3Gib2V",
        "outputId": "1dd6cd80-e708-44e3-9a24-4796688ab77d",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        index  first_date\n",
              "2  2020-12-07        3337\n",
              "8  2020-12-08        1900\n",
              "13 2020-12-09        1258\n",
              "9  2020-12-10        1888\n",
              "12 2020-12-11        1415\n",
              "5  2020-12-12        2290\n",
              "3  2020-12-13        2773\n",
              "1  2020-12-14        3356\n",
              "10 2020-12-15        1797\n",
              "14 2020-12-16        1258\n",
              "11 2020-12-17        1782\n",
              "7  2020-12-18        2016\n",
              "6  2020-12-19        2119\n",
              "4  2020-12-20        2539\n",
              "0  2020-12-21        3649"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-c3aac13f-4fc5-4e3d-ac62-842fc3efa096\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>index</th>\n",
              "      <th>first_date</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>3337</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>2020-12-08</td>\n",
              "      <td>1900</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>2020-12-09</td>\n",
              "      <td>1258</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>2020-12-10</td>\n",
              "      <td>1888</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>2020-12-11</td>\n",
              "      <td>1415</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2020-12-12</td>\n",
              "      <td>2290</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2020-12-13</td>\n",
              "      <td>2773</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2020-12-14</td>\n",
              "      <td>3356</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>2020-12-15</td>\n",
              "      <td>1797</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>2020-12-16</td>\n",
              "      <td>1258</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>2020-12-17</td>\n",
              "      <td>1782</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>2020-12-18</td>\n",
              "      <td>2016</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>2020-12-19</td>\n",
              "      <td>2119</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2020-12-20</td>\n",
              "      <td>2539</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2020-12-21</td>\n",
              "      <td>3649</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c3aac13f-4fc5-4e3d-ac62-842fc3efa096')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-c3aac13f-4fc5-4e3d-ac62-842fc3efa096 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-c3aac13f-4fc5-4e3d-ac62-842fc3efa096');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ],
      "source": [
        "users['first_date'].value_counts().reset_index().sort_values(by = 'index')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 38,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "id": "VPH1JQprzTEn",
        "outputId": "5525c3d1-d4f7-4c35-a4cb-711cf751d1ad",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1296x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "users['first_date'].hist(bins = 15, figsize=(18, 6))\n",
        "\n",
        "plt.title('Распределение регистраций пользователей по периоду наблюдения')\n",
        "plt.xlabel('Даты регистраций')\n",
        "plt.ylabel('Количество регистраций');"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gRwrQ9v9z2J6"
      },
      "source": [
        "Новые пользователи приходили каждый день. Пиковые значения наблюдаются в начале каждой недели проведения теста. Отметим, что эти даты - 7, 14, 21 декабря - понедельники."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 39,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cNrPpN6EN0Jp",
        "outputId": "dc5214b1-838f-4eeb-f720-ef5bc933fc41",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "На данном этапе в датафрейме с пользователями 33377 уникальных пользователей из Европы\n"
          ]
        }
      ],
      "source": [
        "print('На данном этапе в датафрейме с пользователями', users['user_id'].nunique(), 'уникальных пользователей из Европы')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0axhJaKy__PH"
      },
      "source": [
        "Вспомним, сколько у нас уникальных участников теста."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YD7s2bD9P4Tu",
        "outputId": "45bdb944-12c2-49a5-b487-fcd1430d47f4",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "6701"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ],
      "source": [
        "participants['user_id'].nunique()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lQMtnW9P1I39"
      },
      "source": [
        "Отфильтруем пользователей, оставим тех, кто является участниками теста."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 41,
      "metadata": {
        "id": "bqEMZnZNOOrE",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "test_participants = participants['user_id'].tolist()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 42,
      "metadata": {
        "id": "pP9ZZ5HDhcuL",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "users = users.query('user_id in @test_participants')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 43,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pJy7WCSHQDRE",
        "outputId": "5514cff9-fa1e-4e68-9d45-be24ff44d4c6",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Изначально уникальных участников теста -  6701\n",
            "Уникальных участников после фильтрации по дате регистрации и региону -  6351\n",
            "Количество уникальных участников уменьшилось на  5.22 %\n"
          ]
        }
      ],
      "source": [
        "print('Изначально уникальных участников теста - ', participants['user_id'].nunique())\n",
        "print('Уникальных участников после фильтрации по дате регистрации и региону - ', users['user_id'].nunique())\n",
        "print('Количество уникальных участников уменьшилось на ', round(100 - (users['user_id'].nunique() / participants['user_id'].nunique() * 100), 2), '%')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DyVFcpU9_mVw"
      },
      "source": [
        "Обновим переменную `test_participants` с отфильтрованными по региону, дате регистрации уникальными участниками теста."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 44,
      "metadata": {
        "id": "tGHIbdkVTbYF",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "test_participants = users['user_id'].tolist()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KNDM7MCHAVBs"
      },
      "source": [
        "Рассмотрим датафрейм с событиями."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 45,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-Cdwc_tp_PLP",
        "outputId": "d77a551a-2828-41d8-8bbd-d66f576e2fa4",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "login           189552\n",
              "product_page    125563\n",
              "purchase         62740\n",
              "product_cart     62462\n",
              "Name: event_name, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 45
        }
      ],
      "source": [
        "events['event_name'].value_counts()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KBgZC7NtAjlm"
      },
      "source": [
        "Перед нами простая продуктовая воронка. Последовательность действий клиентов для осуществления цели действия нестрогая. Число оплативших больше, чем число положивших в корзину. Возможно, на сайте есть опция быстрой покупки или покупки со сторонних площадок.\n",
        "\n",
        "Логично предположить, что последовательность действия нарушена. \n",
        "Верная последовательность:\n",
        "\n",
        "    login           \n",
        "    product_page  \n",
        "    product_cart\n",
        "    purchase        \n",
        "\n",
        "Учтем это в дальнейшем.\n",
        "     "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 46,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 802
        },
        "id": "teVg4_qX_2ZW",
        "outputId": "3c67bd7d-9a25-4d44-ce05-a8606dd32749",
        "scrolled": true,
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        index  event_dt_date\n",
              "20 2020-12-07          11385\n",
              "17 2020-12-08          12547\n",
              "19 2020-12-09          12122\n",
              "14 2020-12-10          14077\n",
              "16 2020-12-11          13864\n",
              "12 2020-12-12          17634\n",
              "9  2020-12-13          20985\n",
              "3  2020-12-14          26184\n",
              "6  2020-12-15          23469\n",
              "10 2020-12-16          20909\n",
              "8  2020-12-17          21751\n",
              "7  2020-12-18          22871\n",
              "5  2020-12-19          24273\n",
              "2  2020-12-20          26425\n",
              "0  2020-12-21          32559\n",
              "1  2020-12-22          29472\n",
              "4  2020-12-23          26108\n",
              "11 2020-12-24          19399\n",
              "13 2020-12-25          16556\n",
              "15 2020-12-26          14058\n",
              "18 2020-12-27          12420\n",
              "21 2020-12-28          11014\n",
              "22 2020-12-29          10146\n",
              "23 2020-12-30             89"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-c29babe6-861c-4ed2-a565-0aa7542b2394\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>index</th>\n",
              "      <th>event_dt_date</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>11385</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>2020-12-08</td>\n",
              "      <td>12547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>2020-12-09</td>\n",
              "      <td>12122</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>2020-12-10</td>\n",
              "      <td>14077</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>2020-12-11</td>\n",
              "      <td>13864</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>2020-12-12</td>\n",
              "      <td>17634</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>2020-12-13</td>\n",
              "      <td>20985</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>2020-12-14</td>\n",
              "      <td>26184</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>2020-12-15</td>\n",
              "      <td>23469</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>2020-12-16</td>\n",
              "      <td>20909</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>2020-12-17</td>\n",
              "      <td>21751</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>2020-12-18</td>\n",
              "      <td>22871</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2020-12-19</td>\n",
              "      <td>24273</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2020-12-20</td>\n",
              "      <td>26425</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2020-12-21</td>\n",
              "      <td>32559</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2020-12-22</td>\n",
              "      <td>29472</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>2020-12-23</td>\n",
              "      <td>26108</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>2020-12-24</td>\n",
              "      <td>19399</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>2020-12-25</td>\n",
              "      <td>16556</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>2020-12-26</td>\n",
              "      <td>14058</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>2020-12-27</td>\n",
              "      <td>12420</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>2020-12-28</td>\n",
              "      <td>11014</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>22</th>\n",
              "      <td>2020-12-29</td>\n",
              "      <td>10146</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23</th>\n",
              "      <td>2020-12-30</td>\n",
              "      <td>89</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c29babe6-861c-4ed2-a565-0aa7542b2394')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-c29babe6-861c-4ed2-a565-0aa7542b2394 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-c29babe6-861c-4ed2-a565-0aa7542b2394');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 46
        }
      ],
      "source": [
        "events['event_dt_date'].value_counts().reset_index().sort_values(by = 'index')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HpdKwqgCCMbf"
      },
      "source": [
        "По ТЗ тест длится с 2020-12-07 по 2021-01-04. На практике события ограничиваются 2020-12-30 включительно. Причем за последний день данные неполные."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 47,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "id": "lJveSznTQvG4",
        "outputId": "4f2f79a0-330c-4ae5-da50-fe2cd607ef92",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1296x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "events['event_dt_date'].hist(bins = 24, figsize=(18, 6))\n",
        "\n",
        "plt.title('Распределение событий по периоду наблюдения')\n",
        "plt.xlabel('Дата события')\n",
        "plt.ylabel('Количество событий');"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4FQiCa85Cd_2"
      },
      "source": [
        "В датафрейме нет пропусков по дням событий. В начале 20-х чисел, перед католическим Рождеством, - пиковое значение."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jKTARYbSDKja"
      },
      "source": [
        "Посмотрим, были ли маркетинговые активности в период проведения теста."
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "marketing[(marketing['start_dt'] <= '2021-01-04') | (marketing['finish_dt'] >= '2020-12-07')]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "F76BsJ_0RUcc",
        "outputId": "29596528-3977-4c0d-fd48-93b583c0eaa2"
      },
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                name                   regions   start_dt  \\\n",
              "0           Christmas&New Year Promo             EU, N.America 2020-12-25   \n",
              "1       St. Valentine's Day Giveaway  EU, CIS, APAC, N.America 2020-02-14   \n",
              "2             St. Patric's Day Promo             EU, N.America 2020-03-17   \n",
              "3                       Easter Promo  EU, CIS, APAC, N.America 2020-04-12   \n",
              "4                  4th of July Promo                 N.America 2020-07-04   \n",
              "5          Black Friday Ads Campaign  EU, CIS, APAC, N.America 2020-11-26   \n",
              "6             Chinese New Year Promo                      APAC 2020-01-25   \n",
              "7   Labor day (May 1st) Ads Campaign             EU, CIS, APAC 2020-05-01   \n",
              "8    International Women's Day Promo             EU, CIS, APAC 2020-03-08   \n",
              "9    Victory Day CIS (May 9th) Event                       CIS 2020-05-09   \n",
              "10         CIS New Year Gift Lottery                       CIS 2020-12-30   \n",
              "11     Dragon Boat Festival Giveaway                      APAC 2020-06-25   \n",
              "12           Single's Day Gift Promo                      APAC 2020-11-11   \n",
              "13             Chinese Moon Festival                      APAC 2020-10-01   \n",
              "\n",
              "    finish_dt  \n",
              "0  2021-01-03  \n",
              "1  2020-02-16  \n",
              "2  2020-03-19  \n",
              "3  2020-04-19  \n",
              "4  2020-07-11  \n",
              "5  2020-12-01  \n",
              "6  2020-02-07  \n",
              "7  2020-05-03  \n",
              "8  2020-03-10  \n",
              "9  2020-05-11  \n",
              "10 2021-01-07  \n",
              "11 2020-07-01  \n",
              "12 2020-11-12  \n",
              "13 2020-10-07  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-bfcc82cd-7c82-4903-9ac7-69e6b735b883\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>name</th>\n",
              "      <th>regions</th>\n",
              "      <th>start_dt</th>\n",
              "      <th>finish_dt</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Christmas&amp;New Year Promo</td>\n",
              "      <td>EU, N.America</td>\n",
              "      <td>2020-12-25</td>\n",
              "      <td>2021-01-03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>St. Valentine's Day Giveaway</td>\n",
              "      <td>EU, CIS, APAC, N.America</td>\n",
              "      <td>2020-02-14</td>\n",
              "      <td>2020-02-16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>St. Patric's Day Promo</td>\n",
              "      <td>EU, N.America</td>\n",
              "      <td>2020-03-17</td>\n",
              "      <td>2020-03-19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Easter Promo</td>\n",
              "      <td>EU, CIS, APAC, N.America</td>\n",
              "      <td>2020-04-12</td>\n",
              "      <td>2020-04-19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4th of July Promo</td>\n",
              "      <td>N.America</td>\n",
              "      <td>2020-07-04</td>\n",
              "      <td>2020-07-11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Black Friday Ads Campaign</td>\n",
              "      <td>EU, CIS, APAC, N.America</td>\n",
              "      <td>2020-11-26</td>\n",
              "      <td>2020-12-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Chinese New Year Promo</td>\n",
              "      <td>APAC</td>\n",
              "      <td>2020-01-25</td>\n",
              "      <td>2020-02-07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Labor day (May 1st) Ads Campaign</td>\n",
              "      <td>EU, CIS, APAC</td>\n",
              "      <td>2020-05-01</td>\n",
              "      <td>2020-05-03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>International Women's Day Promo</td>\n",
              "      <td>EU, CIS, APAC</td>\n",
              "      <td>2020-03-08</td>\n",
              "      <td>2020-03-10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Victory Day CIS (May 9th) Event</td>\n",
              "      <td>CIS</td>\n",
              "      <td>2020-05-09</td>\n",
              "      <td>2020-05-11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>CIS New Year Gift Lottery</td>\n",
              "      <td>CIS</td>\n",
              "      <td>2020-12-30</td>\n",
              "      <td>2021-01-07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>Dragon Boat Festival Giveaway</td>\n",
              "      <td>APAC</td>\n",
              "      <td>2020-06-25</td>\n",
              "      <td>2020-07-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>Single's Day Gift Promo</td>\n",
              "      <td>APAC</td>\n",
              "      <td>2020-11-11</td>\n",
              "      <td>2020-11-12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>Chinese Moon Festival</td>\n",
              "      <td>APAC</td>\n",
              "      <td>2020-10-01</td>\n",
              "      <td>2020-10-07</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-bfcc82cd-7c82-4903-9ac7-69e6b735b883')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-bfcc82cd-7c82-4903-9ac7-69e6b735b883 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-bfcc82cd-7c82-4903-9ac7-69e6b735b883');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y0M9k48jDbzI"
      },
      "source": [
        "Да, рождественско-новогодняя рекламная кампания в том числе и в Европе. Ее, как и праздничные распродажи, можно считать внешним фактором, который может влиять на результаты теста. Позже посмотрим, параллельно ли делались замеры по группам для снижения влияния внешних факторов. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PJpTz-PyFS6q"
      },
      "source": [
        "Объединим таблицы с участниками теста и событиями, отфильтровав по участникам теста."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 49,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "-sRWUiKVKmaa",
        "outputId": "6476c169-1e63-4d88-fff2-012c79cb66a4",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            user_id             event_dt    event_name  details event_dt_date  \\\n",
              "0  831887FE7F2D6CBA  2020-12-07 06:50:29      purchase     4.99    2020-12-07   \n",
              "1  831887FE7F2D6CBA  2020-12-09 02:19:17      purchase    99.99    2020-12-09   \n",
              "2  831887FE7F2D6CBA  2020-12-07 06:50:30  product_cart      NaN    2020-12-07   \n",
              "\n",
              "  group                  ab_test  \n",
              "0     A  recommender_system_test  \n",
              "1     A  recommender_system_test  \n",
              "2     A  recommender_system_test  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-83f5d07e-77b6-4db2-b6b5-90c06a63507f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>event_dt</th>\n",
              "      <th>event_name</th>\n",
              "      <th>details</th>\n",
              "      <th>event_dt_date</th>\n",
              "      <th>group</th>\n",
              "      <th>ab_test</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>831887FE7F2D6CBA</td>\n",
              "      <td>2020-12-07 06:50:29</td>\n",
              "      <td>purchase</td>\n",
              "      <td>4.99</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>831887FE7F2D6CBA</td>\n",
              "      <td>2020-12-09 02:19:17</td>\n",
              "      <td>purchase</td>\n",
              "      <td>99.99</td>\n",
              "      <td>2020-12-09</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>831887FE7F2D6CBA</td>\n",
              "      <td>2020-12-07 06:50:30</td>\n",
              "      <td>product_cart</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-83f5d07e-77b6-4db2-b6b5-90c06a63507f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-83f5d07e-77b6-4db2-b6b5-90c06a63507f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-83f5d07e-77b6-4db2-b6b5-90c06a63507f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 49
        }
      ],
      "source": [
        "events_new = pd.merge(events, participants, how='inner', on='user_id') #параметр inner, чтобы в итоговую выборку попали пользователи из participants,\n",
        "                                                                       #о действиях которых есть информация \n",
        "\n",
        "events_new.head(3)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XzqnlED2D_e0"
      },
      "source": [
        "Дополнительно отфильтруем события по уникальным пользователям, которые соответствуют требованям ТЗ."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 50,
      "metadata": {
        "id": "SzaoShZ4OUOK",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "events_new = events_new.query('user_id in @test_participants')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 51,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mr9tYMNVEnJR",
        "outputId": "5ab71997-7340-4c8e-f6f2-eb772f54d5df",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Кто-то потерялся. Количество уникальных участников после фильтрации не совпадает. Было - 6351 Стало - 3481 Изменения - -45.19 %\n"
          ]
        }
      ],
      "source": [
        "if users['user_id'].nunique() == events_new['user_id'].nunique():\n",
        "  print('Уникальных участников теста по-прежнему', events_new['user_id'].nunique(),'- никто не потерялся')\n",
        "else:\n",
        "  print('Кто-то потерялся. Количество уникальных участников после фильтрации не совпадает. Было -', users['user_id'].nunique(), 'Стало -',events_new['user_id'].nunique(),\\\n",
        "        'Изменения -', round(((events_new['user_id'].nunique() / users['user_id'].nunique() * 100) - 100), 2), '%')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IPJ_uQKBEStg"
      },
      "source": [
        "Почти у половины отфильтрованных участников теста нет сведений по событиям. Рассмотри их подробнее."
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "users_without_events = pd.merge(events, participants, how='right', on='user_id') \n",
        "\n",
        "users_without_events = users_without_events[users_without_events['event_dt_date'].isna()]\n",
        "\n",
        "users_without_events = pd.merge(users, users_without_events, how='inner', on='user_id')\n",
        "\n",
        "users_without_events = users_without_events.groupby(['first_date', 'group'])['user_id'].nunique().reset_index()\n",
        "\n",
        "users_without_events['first_date'] = users_without_events['first_date'].dt.strftime('%Y-%m-%d')"
      ],
      "metadata": {
        "id": "9ZHSXXjprGg6"
      },
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style('dark')\n",
        "\n",
        "plt.figure(figsize=(18, 6))\n",
        "\n",
        "sns.barplot(x='first_date', y='user_id', data=users_without_events, hue='group')\n",
        "\n",
        "plt.title('Распределение пользователей, у которых нет событий, по дням регистраций')\n",
        "plt.xlabel('Дата')\n",
        "plt.ylabel('Количество зарегистрировавшихся пользователей')\n",
        "plt.legend(loc='upper right', fontsize=10)\n",
        "plt.grid()\n",
        "plt.xticks(rotation=80)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 453
        },
        "id": "W3-t2LdOrG_M",
        "outputId": "360cd826-175c-4c3f-9227-423ef1664fd2"
      },
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1296x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Обнаружили странность. Начиная с понедельника, 14 декабря, уникальные пользователи без событий попадали только в тестовую группу B, в тот время как в A с этого периода не оказалось ни одного. Возможно, кто-то внес изменения в настройки сбора сведений для теста, начиная с 14 декабря. Либо с этой даты возникла ошибка при загрузке сведений об уникальных пользователях без событий из контрольной группы."
      ],
      "metadata": {
        "id": "NSE-3X5b2LhH"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YYGgoCqFTcHW"
      },
      "source": [
        "Добавим даты регистрации в датафрейм с отфильтрованными данными."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 54,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "7I1ftvU0Tgv9",
        "outputId": "1c3004f1-a7bb-4610-8a00-49b012b38695",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            user_id first_date region   device             event_dt  \\\n",
              "0  D72A72121175D8BE 2020-12-07     EU       PC  2020-12-07 21:52:10   \n",
              "1  D72A72121175D8BE 2020-12-07     EU       PC  2020-12-07 21:52:07   \n",
              "2  DD4352CDCF8C3D57 2020-12-07     EU  Android  2020-12-07 15:32:54   \n",
              "\n",
              "     event_name  details event_dt_date group                  ab_test  \n",
              "0  product_page      NaN    2020-12-07     A  recommender_system_test  \n",
              "1         login      NaN    2020-12-07     A  recommender_system_test  \n",
              "2  product_page      NaN    2020-12-07     B  recommender_system_test  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d0bbb6a7-51d3-4bb5-96d5-066dae67db4b\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user_id</th>\n",
              "      <th>first_date</th>\n",
              "      <th>region</th>\n",
              "      <th>device</th>\n",
              "      <th>event_dt</th>\n",
              "      <th>event_name</th>\n",
              "      <th>details</th>\n",
              "      <th>event_dt_date</th>\n",
              "      <th>group</th>\n",
              "      <th>ab_test</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>D72A72121175D8BE</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>EU</td>\n",
              "      <td>PC</td>\n",
              "      <td>2020-12-07 21:52:10</td>\n",
              "      <td>product_page</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>D72A72121175D8BE</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>EU</td>\n",
              "      <td>PC</td>\n",
              "      <td>2020-12-07 21:52:07</td>\n",
              "      <td>login</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>A</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>DD4352CDCF8C3D57</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>EU</td>\n",
              "      <td>Android</td>\n",
              "      <td>2020-12-07 15:32:54</td>\n",
              "      <td>product_page</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2020-12-07</td>\n",
              "      <td>B</td>\n",
              "      <td>recommender_system_test</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d0bbb6a7-51d3-4bb5-96d5-066dae67db4b')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-d0bbb6a7-51d3-4bb5-96d5-066dae67db4b button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-d0bbb6a7-51d3-4bb5-96d5-066dae67db4b');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 54
        }
      ],
      "source": [
        "events_new = pd.merge(users, events_new, how='inner', on='user_id') #параметр inner, чтобы в итоговую выборку попали пользователи из participants,\n",
        "                                                                       #о действиях которых есть информация \n",
        "\n",
        "events_new.head(3)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sURPCcXJ8-1b"
      },
      "source": [
        "По условиям ТЗ, анализируется 14-дневный срок с момента регистрации пользователей. Учтем это требование."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 55,
      "metadata": {
        "id": "Tok5VOKvUrJh",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "events_new['14_days_last'] = events_new['first_date']  + timedelta(days=14)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 56,
      "metadata": {
        "id": "6-Cjjsyw98SO",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "events_new = events_new[events_new['event_dt_date'] <= events_new['14_days_last']].copy()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WG2l8AlHIOEk"
      },
      "source": [
        "Еще раз проверим пересекающихся пользователей в группах."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 57,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "8wIK3vBkIUYe",
        "outputId": "a23e5bf5-7dd4-4246-f938-4c8cfb1e7bd4",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Одинаковых пользователей для групп A и B: 0'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 57
        }
      ],
      "source": [
        "all_group = set(events_new.query('group == \"A\"')['user_id']) & set(events_new.query('group == \"B\"')['user_id']) \n",
        "\n",
        "f'Одинаковых пользователей для групп A и B: {len(all_group)}'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 58,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BalsDftglHTP",
        "outputId": "a43773af-ab0a-4783-f80a-0cc5492906da",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "После фильтрации в нашем распоряжении 22828 событий, а именно:\n"
          ]
        }
      ],
      "source": [
        "print('После фильтрации в нашем распоряжении', events_new['event_dt_date'].count(), 'событий, а именно:')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 59,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kz0uIxggEoev",
        "outputId": "43c48c65-6c8d-4796-af28-4451998ceccc",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "A    17977\n",
              "B     4851\n",
              "Name: group, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 59
        }
      ],
      "source": [
        "events_new['group'].value_counts()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 60,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uv-LflnCE7Fb",
        "outputId": "5e78d047-cff2-4c42-f187-111c095217c0",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "A   0.79\n",
              "B   0.21\n",
              "Name: group, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 60
        }
      ],
      "source": [
        "events_new['group'].value_counts(normalize=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-tQr6KjGFDUY"
      },
      "source": [
        "Выборки сильно отличаются."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 61,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "vQ5FvQkQl6BD",
        "outputId": "f0e83374-88db-4db9-a350-8697d2477a66",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  group    event_name  event_dt_date\n",
              "0     A         login           7968\n",
              "1     A  product_cart           2385\n",
              "2     A  product_page           5125\n",
              "3     A      purchase           2499\n",
              "4     B         login           2345\n",
              "5     B  product_cart            625\n",
              "6     B  product_page           1257\n",
              "7     B      purchase            624"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5d8ee6c1-8548-4aad-a180-79a5838808d4\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>group</th>\n",
              "      <th>event_name</th>\n",
              "      <th>event_dt_date</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>A</td>\n",
              "      <td>login</td>\n",
              "      <td>7968</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>A</td>\n",
              "      <td>product_cart</td>\n",
              "      <td>2385</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>A</td>\n",
              "      <td>product_page</td>\n",
              "      <td>5125</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>A</td>\n",
              "      <td>purchase</td>\n",
              "      <td>2499</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>B</td>\n",
              "      <td>login</td>\n",
              "      <td>2345</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>B</td>\n",
              "      <td>product_cart</td>\n",
              "      <td>625</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>B</td>\n",
              "      <td>product_page</td>\n",
              "      <td>1257</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>B</td>\n",
              "      <td>purchase</td>\n",
              "      <td>624</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5d8ee6c1-8548-4aad-a180-79a5838808d4')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5d8ee6c1-8548-4aad-a180-79a5838808d4 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5d8ee6c1-8548-4aad-a180-79a5838808d4');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 61
        }
      ],
      "source": [
        "events_new.groupby(['group', 'event_name'])['event_dt_date'].count().reset_index()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 62,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "id": "yyUOG9PFngDb",
        "outputId": "171f15c3-1577-4c84-ffbe-402d4ac98500",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1296x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "sns.set_style('dark')\n",
        "\n",
        "plt.figure(figsize=(18, 6))\n",
        "\n",
        "sns.barplot(x='group', y='event_dt_date', data=events_new.groupby(['group', 'event_name'])['event_dt_date'].count().reset_index(), hue='event_name')\n",
        "\n",
        "plt.title('Распределение событий по группам')\n",
        "plt.xlabel('Название группы')\n",
        "plt.ylabel('Количество событий')\n",
        "plt.legend(loc='upper right', fontsize=10)\n",
        "plt.grid()\n",
        "\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "be0pQHMIGGQo"
      },
      "source": [
        "Количество сообытий по группам распределено пропорционально, хотя в группе В втрое меньше событий. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UFO-LuhrHV8V"
      },
      "source": [
        "Посмотрим, составили ли отфильтрованные участники теста 15% от всех клиентов в Европе, как было установлено в ТЗ."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 63,
      "metadata": {
        "id": "pZGy_0cnIxqq",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "test_participants = events_new['user_id'].tolist()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 64,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oskCjvWdJFkk",
        "outputId": "78586323-a506-4dc9-91bc-bcc6ffc522cb",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Доля отфильтрованных участников теста от общего числа клиентов из Европы - 7.52 %\n",
            "\n",
            "Согласно ТЗ, требуется 15%, не хватает 7.48 п.п.\n"
          ]
        }
      ],
      "source": [
        "print('Доля отфильтрованных участников теста от общего числа клиентов из Европы -',\\\n",
        "      round(total_users.query('user_id in @test_participants')['region'].value_counts()[0] / total_users_by_regions[0] * 100, 2), '%')\n",
        "print()\n",
        "print('Согласно ТЗ, требуется 15%, не хватает',\\\n",
        "      round(15 - round(total_users.query('user_id in @test_participants')['region'].value_counts()[0] / total_users_by_regions[0] * 100, 2), 2), 'п.п.')\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3Oqzx5iRL4BX"
      },
      "source": [
        "Таким образом, выводы на соответствие данным теста ТЗ:\n",
        "\n",
        "ДА:\n",
        "- Название теста: `recommender_system_test`;\n",
        "- группы: А — контрольная, B — новая платёжная воронка;\n",
        "- дата запуска: 2020-12-07;\n",
        "- дата остановки набора новых пользователей: 2020-12-21;\n",
        "- дата остановки: 2021-01-04;\n",
        "- отфильтровали 14 дней с момента регистрации.\n",
        "\n",
        "НЕТ:\n",
        "- Данные заканчиваются 2020-12-30, вместо 2021-01-04;\n",
        "- аудитория новых пользователей из региона EU не равна заявленным 15%, по факту - вдвое меньше \n",
        "- ожидаемое количество участников теста: 6000.\n",
        "- участников гораздо меньше заявленных 6000, но тех, у кого есть события - 3481.\n",
        "\n",
        "У нас совсем мало дней наблюдений, чтобы набрать когорты посетителей и оценить изменения событий в этих когортах. С 7 по 24 декабря - 18 дней. С 25-го начинается рождественская промо-акция. Она пересекается с тестом формально 11 дней, а фактически, поскольку данные событий только до 30 декабря - 6.  \n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oBSBDiCpX8rg"
      },
      "source": [
        "## Проведем исследовательский анализ данных"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7qm3uEm1YEZ6"
      },
      "source": [
        "### Количество событий на пользователя одинаково распределены в выборках?\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 65,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "nolDFh2u7bs-",
        "outputId": "3fcb8dcd-1a32-4404-ee0c-37146537f971",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       event_name\n",
              "count    3,481.00\n",
              "mean         6.56\n",
              "std          3.77\n",
              "min          1.00\n",
              "25%          4.00\n",
              "50%          6.00\n",
              "75%          9.00\n",
              "max         24.00"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-78bb0c69-bae2-40f8-8a0f-1b76aa3f01ce\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>event_name</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>3,481.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>6.56</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>3.77</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>4.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>6.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>9.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>24.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-78bb0c69-bae2-40f8-8a0f-1b76aa3f01ce')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-78bb0c69-bae2-40f8-8a0f-1b76aa3f01ce button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-78bb0c69-bae2-40f8-8a0f-1b76aa3f01ce');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 65
        }
      ],
      "source": [
        "events_new.groupby('user_id').agg({'event_name': 'count'}).describe()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3D8J0rijIdKU"
      },
      "source": [
        "На одного пользователя в среднем 7 событий. Медиана немного ниже - 6. Значит, выбросов не должно быть много."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 66,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "id": "YWBqwdZoIlKC",
        "outputId": "44b00bbb-9972-4abb-df53-54e675cecd31",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1296x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "events_new.groupby('user_id').agg({'event_name': 'count'}).boxplot(figsize=(18, 6))\n",
        "\n",
        "plt.title('Распределение средних значений событий на одного пользоватея')\n",
        "plt.ylabel('Количество событий');"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m_pyl1uUI_2u"
      },
      "source": [
        "Выбросы есть, но их не мало.Вполне возможно, некоторые пользователи рьяно посещают сайт, изучают содержимое, но так ничего не покупают."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a7WWh7JLYJxZ"
      },
      "source": [
        "Посмотрим на события в выборках по дням."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 67,
      "metadata": {
        "id": "43utPJyQkpZr",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "distribution_by_days = events_new.groupby(['event_dt_date', 'group'])['event_name'].count().reset_index()\n",
        "distribution_by_days['event_dt_date'] = distribution_by_days['event_dt_date'].dt.strftime('%Y-%m-%d')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 68,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 453
        },
        "id": "w02LLhZiJoBM",
        "outputId": "79a6ba1c-28b5-41b6-d0f8-ecb162345e37",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "sns.set_style('dark')\n",
        "\n",
        "plt.figure(figsize=(15, 6))\n",
        "\n",
        "sns.barplot(x='event_dt_date', y='event_name', data=distribution_by_days, hue='group')\n",
        "\n",
        "plt.title('Распределение событий по группам по дням')\n",
        "plt.xlabel('Дата')\n",
        "plt.ylabel('Количество событий')\n",
        "plt.legend(loc='upper right', fontsize=10)\n",
        "plt.grid()\n",
        "plt.xticks(rotation=80)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZoqK0xptuJKV"
      },
      "source": [
        "В первый день выборка группы B даже немного выше. Со второй недели (с 7 декабря, в понедельник) у выборки А наблюдается четырехкратное количественное опережение выборки В. Показатели группы А растут всю вторую неделю. На начало третьей (21 декабря, снова понедельник) приходится пик обеих выборок. После 21 декабря у группы A явное снижение количества событий, у группы B - небольшое снижение. Отметим, что снижение приходится на начало рекламной кампании и период рождественско-новогодних праздников."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sP7Z7FugYN0l"
      },
      "source": [
        "### Воронки событий"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gEed7LTO2JlG"
      },
      "source": [
        "Определим, сколько пользователей совершали каждое из событий, а также долю пользователей, которые хоть раз совершали событие. Не забудем о последовательности действий в воронке, которую мы установили выше."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 69,
      "metadata": {
        "id": "9iFrJaUc3JeS",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "events_new_A = events_new.query('group == \"A\"')\n",
        "events_new_B = events_new.query('group == \"B\"')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 70,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "-LW9phLI2SmN",
        "outputId": "244aada1-1307-4c93-c676-44f975e86141",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     event_name  number_of_users  user_share\n",
              "0         login             2604      100.00\n",
              "2  product_page             1685       64.71\n",
              "1  product_cart              782       30.03\n",
              "3      purchase              833       31.99"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5135a7a6-ded8-4c94-b0af-29e91782e22a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>event_name</th>\n",
              "      <th>number_of_users</th>\n",
              "      <th>user_share</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>login</td>\n",
              "      <td>2604</td>\n",
              "      <td>100.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>product_page</td>\n",
              "      <td>1685</td>\n",
              "      <td>64.71</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>product_cart</td>\n",
              "      <td>782</td>\n",
              "      <td>30.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>purchase</td>\n",
              "      <td>833</td>\n",
              "      <td>31.99</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5135a7a6-ded8-4c94-b0af-29e91782e22a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-5135a7a6-ded8-4c94-b0af-29e91782e22a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-5135a7a6-ded8-4c94-b0af-29e91782e22a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ],
      "source": [
        "parts_users_per_event_part_A= (events_new_A\n",
        "                               .groupby('event_name').agg(number_of_users = ('user_id', 'nunique'))\n",
        "                               .reset_index()\n",
        "                               .sort_values('number_of_users', ascending=False)\n",
        "                               .assign(user_share = lambda x: x['number_of_users'] * 100 / events_new_A['user_id']\n",
        "                                       .nunique())\n",
        "                                       )\n",
        "\n",
        "parts_users_per_event_part_A.reindex([0,2,1,3])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 71,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "xjfiCdA5wSv8",
        "outputId": "f7f02867-6c4b-476f-d584-8c7a59d4c8e4",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     event_name  number_of_users  user_share\n",
              "0         login              876       99.89\n",
              "2  product_page              493       56.21\n",
              "1  product_cart              244       27.82\n",
              "3      purchase              249       28.39"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-485dfdd3-eab5-4268-9e9f-6108c9400cdf\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>event_name</th>\n",
              "      <th>number_of_users</th>\n",
              "      <th>user_share</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>login</td>\n",
              "      <td>876</td>\n",
              "      <td>99.89</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>product_page</td>\n",
              "      <td>493</td>\n",
              "      <td>56.21</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>product_cart</td>\n",
              "      <td>244</td>\n",
              "      <td>27.82</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>purchase</td>\n",
              "      <td>249</td>\n",
              "      <td>28.39</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-485dfdd3-eab5-4268-9e9f-6108c9400cdf')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-485dfdd3-eab5-4268-9e9f-6108c9400cdf button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-485dfdd3-eab5-4268-9e9f-6108c9400cdf');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 71
        }
      ],
      "source": [
        "parts_users_per_event_part_B= (events_new_B\n",
        "                               .groupby('event_name').agg(number_of_users = ('user_id', 'nunique'))\n",
        "                               .reset_index()\n",
        "                               .sort_values('number_of_users', ascending=False)\n",
        "                               .assign(user_share = lambda x: x['number_of_users'] * 100 / events_new_B['user_id']\n",
        "                                       .nunique())\n",
        "                                       )\n",
        "\n",
        "parts_users_per_event_part_B.reindex([0,2,1,3])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bpLvo9iONs6U"
      },
      "source": [
        "Судя по тому, что не все залогинились, покупать можно и через сторонние сайты."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RBnEPWuAMJk1"
      },
      "source": [
        "Перейдем к построению воронки."
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fig = go.Figure()\n",
        "\n",
        "fig.add_trace(go.Funnel(\n",
        "    name = 'Группа А',\n",
        "    y = [\"login\", \"product_page\",\\\n",
        "         \"product_cart\", \"purchase\t\"],\n",
        "    x = [2604, 1685, 782, 833],\n",
        "    textinfo = \"value+percent initial\"))\n",
        "\n",
        "fig.add_trace(go.Funnel(\n",
        "    name = 'Группа B',\n",
        "    orientation = \"h\",\n",
        "    y = [\"login\", \"product_page\",\\\n",
        "         \"product_cart\", \"purchase\t\"],\n",
        "    x = [876, 493, 244, 249],\n",
        "    textposition = \"inside\",\n",
        "    textinfo = \"value+percent initial\"))\n",
        "\n",
        "fig.update_layout(title=\"Воронка событий\", width=1200, height=500)\n",
        "fig.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 537
        },
        "id": "bM1eY3hKJ2ua",
        "outputId": "54212553-27c8-40a6-defc-01ea7c252806"
      },
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-2.8.3.min.js\"></script>                <div id=\"77d781b2-0744-44a3-97dc-d1c97edd70bd\" class=\"plotly-graph-div\" style=\"height:500px; width:1200px;\"></div>            <script type=\"text/javascript\">                                    window.PLOTLYENV=window.PLOTLYENV || {};                                    if (document.getElementById(\"77d781b2-0744-44a3-97dc-d1c97edd70bd\")) {                    Plotly.newPlot(                        \"77d781b2-0744-44a3-97dc-d1c97edd70bd\",                        [{\"name\":\"\\u0413\\u0440\\u0443\\u043f\\u043f\\u0430 \\u0410\",\"textinfo\":\"value+percent initial\",\"x\":[2604,1685,782,833],\"y\":[\"login\",\"product_page\",\"product_cart\",\"purchase\\t\"],\"type\":\"funnel\"},{\"name\":\"\\u0413\\u0440\\u0443\\u043f\\u043f\\u0430 B\",\"orientation\":\"h\",\"textinfo\":\"value+percent initial\",\"textposition\":\"inside\",\"x\":[876,493,244,249],\"y\":[\"login\",\"product_page\",\"product_cart\",\"purchase\\t\"],\"type\":\"funnel\"}],                        {\"template\":{\"data\":{\"bar\":[{\"error_x\":{\"color\":\"#2a3f5f\"},\"error_y\":{\"color\":\"#2a3f5f\"},\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"bar\"}],\"barpolar\":[{\"marker\":{\"line\":{\"color\":\"#E5ECF6\",\"width\":0.5},\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"barpolar\"}],\"carpet\":[{\"aaxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"baxis\":{\"endlinecolor\":\"#2a3f5f\",\"gridcolor\":\"white\",\"linecolor\":\"white\",\"minorgridcolor\":\"white\",\"startlinecolor\":\"#2a3f5f\"},\"type\":\"carpet\"}],\"choropleth\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"choropleth\"}],\"contour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"contour\"}],\"contourcarpet\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"contourcarpet\"}],\"heatmap\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmap\"}],\"heatmapgl\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"heatmapgl\"}],\"histogram\":[{\"marker\":{\"pattern\":{\"fillmode\":\"overlay\",\"size\":10,\"solidity\":0.2}},\"type\":\"histogram\"}],\"histogram2d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2d\"}],\"histogram2dcontour\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"histogram2dcontour\"}],\"mesh3d\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"type\":\"mesh3d\"}],\"parcoords\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"parcoords\"}],\"pie\":[{\"automargin\":true,\"type\":\"pie\"}],\"scatter\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter\"}],\"scatter3d\":[{\"line\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatter3d\"}],\"scattercarpet\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattercarpet\"}],\"scattergeo\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergeo\"}],\"scattergl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattergl\"}],\"scattermapbox\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scattermapbox\"}],\"scatterpolar\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolar\"}],\"scatterpolargl\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterpolargl\"}],\"scatterternary\":[{\"marker\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"type\":\"scatterternary\"}],\"surface\":[{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"},\"colorscale\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"type\":\"surface\"}],\"table\":[{\"cells\":{\"fill\":{\"color\":\"#EBF0F8\"},\"line\":{\"color\":\"white\"}},\"header\":{\"fill\":{\"color\":\"#C8D4E3\"},\"line\":{\"color\":\"white\"}},\"type\":\"table\"}]},\"layout\":{\"annotationdefaults\":{\"arrowcolor\":\"#2a3f5f\",\"arrowhead\":0,\"arrowwidth\":1},\"autotypenumbers\":\"strict\",\"coloraxis\":{\"colorbar\":{\"outlinewidth\":0,\"ticks\":\"\"}},\"colorscale\":{\"diverging\":[[0,\"#8e0152\"],[0.1,\"#c51b7d\"],[0.2,\"#de77ae\"],[0.3,\"#f1b6da\"],[0.4,\"#fde0ef\"],[0.5,\"#f7f7f7\"],[0.6,\"#e6f5d0\"],[0.7,\"#b8e186\"],[0.8,\"#7fbc41\"],[0.9,\"#4d9221\"],[1,\"#276419\"]],\"sequential\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]],\"sequentialminus\":[[0.0,\"#0d0887\"],[0.1111111111111111,\"#46039f\"],[0.2222222222222222,\"#7201a8\"],[0.3333333333333333,\"#9c179e\"],[0.4444444444444444,\"#bd3786\"],[0.5555555555555556,\"#d8576b\"],[0.6666666666666666,\"#ed7953\"],[0.7777777777777778,\"#fb9f3a\"],[0.8888888888888888,\"#fdca26\"],[1.0,\"#f0f921\"]]},\"colorway\":[\"#636efa\",\"#EF553B\",\"#00cc96\",\"#ab63fa\",\"#FFA15A\",\"#19d3f3\",\"#FF6692\",\"#B6E880\",\"#FF97FF\",\"#FECB52\"],\"font\":{\"color\":\"#2a3f5f\"},\"geo\":{\"bgcolor\":\"white\",\"lakecolor\":\"white\",\"landcolor\":\"#E5ECF6\",\"showlakes\":true,\"showland\":true,\"subunitcolor\":\"white\"},\"hoverlabel\":{\"align\":\"left\"},\"hovermode\":\"closest\",\"mapbox\":{\"style\":\"light\"},\"paper_bgcolor\":\"white\",\"plot_bgcolor\":\"#E5ECF6\",\"polar\":{\"angularaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"radialaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"scene\":{\"xaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"yaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"},\"zaxis\":{\"backgroundcolor\":\"#E5ECF6\",\"gridcolor\":\"white\",\"gridwidth\":2,\"linecolor\":\"white\",\"showbackground\":true,\"ticks\":\"\",\"zerolinecolor\":\"white\"}},\"shapedefaults\":{\"line\":{\"color\":\"#2a3f5f\"}},\"ternary\":{\"aaxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"baxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"},\"bgcolor\":\"#E5ECF6\",\"caxis\":{\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\"}},\"title\":{\"x\":0.05},\"xaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2},\"yaxis\":{\"automargin\":true,\"gridcolor\":\"white\",\"linecolor\":\"white\",\"ticks\":\"\",\"title\":{\"standoff\":15},\"zerolinecolor\":\"white\",\"zerolinewidth\":2}}},\"title\":{\"text\":\"\\u0412\\u043e\\u0440\\u043e\\u043d\\u043a\\u0430 \\u0441\\u043e\\u0431\\u044b\\u0442\\u0438\\u0439\"},\"width\":1200,\"height\":500},                        {\"responsive\": true}                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('77d781b2-0744-44a3-97dc-d1c97edd70bd');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })                };                            </script>        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HENqc6AEOVR4"
      },
      "source": [
        "Судя по воронкам, показатели группы A выглядят лучше. Клиентов в долях от первоначального количества теряется меньше. До этапа `product_page` доходит 65% от первоначального кличества, в группе B только 56%."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5g_xuRDbYSx3"
      },
      "source": [
        "### Какие особенности данных нужно учесть, прежде чем приступать к A/B-тестированию?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pDidS3IPPHHj"
      },
      "source": [
        "Сильная диспропорция в группах. Данные соответствуют ТЗ частично. Число клиентов вдвое меньше ожидаемого количества. Не о всех днях теста есть информация по действиям пользователей. Крайне неудачное время проведения. На тест могут оказывать влиянеи внешние факторы в виде рождественско-новогодней рекламной кампании, предпраздничных распродаж."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HK85i98pYWLg"
      },
      "source": [
        "## Изучим результаты A/B-тестирования"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N8GglGVuRYAT"
      },
      "source": [
        "Определим, сколько пользователей в каждой группе, добавим колонку с общим количеством пользователей в каждой из групп."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 73,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "hzAR7Hi8R_ZM",
        "outputId": "793d6aed-e3be-4cde-bdf2-40ccfa4bd2ae",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "event_name  login  product_page  product_cart  purchase  users_total\n",
              "group                                                               \n",
              "A            2604          1685           782       833         2604\n",
              "B             876           493           244       249          877"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-2edad396-1ddf-4f71-9edd-c3ffe8645c8a\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>event_name</th>\n",
              "      <th>login</th>\n",
              "      <th>product_page</th>\n",
              "      <th>product_cart</th>\n",
              "      <th>purchase</th>\n",
              "      <th>users_total</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>group</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>A</th>\n",
              "      <td>2604</td>\n",
              "      <td>1685</td>\n",
              "      <td>782</td>\n",
              "      <td>833</td>\n",
              "      <td>2604</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>B</th>\n",
              "      <td>876</td>\n",
              "      <td>493</td>\n",
              "      <td>244</td>\n",
              "      <td>249</td>\n",
              "      <td>877</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2edad396-1ddf-4f71-9edd-c3ffe8645c8a')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-2edad396-1ddf-4f71-9edd-c3ffe8645c8a button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-2edad396-1ddf-4f71-9edd-c3ffe8645c8a');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 73
        }
      ],
      "source": [
        "users_by_events = events_new\\\n",
        ".pivot_table(index=['group'], columns = ['event_name'], values = 'user_id', aggfunc = 'nunique').reset_index()\n",
        "\n",
        "users_by_events = users_by_events[['group',\t'login',\t'product_page', 'product_cart',\t'purchase']]\n",
        "\n",
        "users_total = events_new.groupby('group')['user_id'].nunique()\n",
        "\n",
        "users_by_events['users_total'] = users_by_events['group'].apply(lambda x: users_total.loc[x])\n",
        "users_by_events = users_by_events.set_index('group')\n",
        "\n",
        "users_by_events"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yao-OA2DV9rS"
      },
      "source": [
        "Используем Z-тест. Применяем именно его, так как мы сравниваем доли на каждом этапе от общего числа уникальных пользователей. Важно проверить количество уникальных пользователей, посетивших тот или иной этап. Это необходимо, чтобы понимать, что пользователь вообще заходил на ту или иную страницу.\n",
        "\n",
        "Напомним формулу:\n",
        "\n",
        "$$Z = \\frac{(P_1-P_2)(π_1 - π_2)}{\\sqrt{P(1-P)(1/n_1+1/n_2)}}\\sim N(0, 1)$$\n",
        "\n",
        "Величина Z — стандартная для критерия со стандартным нормальным распределением: со средним, равным нулю, и стандартным отклонением, равным единице. Это указано в правой части формулы после ～ знака, который говорит, что выражение распределено как N(0,1).\n",
        "\n",
        "Для удобства напишем функцию."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 74,
      "metadata": {
        "id": "Uutjr4h7SrdV",
        "trusted": true
      },
      "outputs": [],
      "source": [
        "def st_test(part0, part1, total0, total1, alpha):  \n",
        "    p1 = part0 / total0 \n",
        "    p2 = part1 / total1 \n",
        "    p_combined = (part0 + part1) / (total0 + total1) \n",
        "    difference = p1 - p2 \n",
        "    z_value = difference / mth.sqrt(p_combined * (1 - p_combined) * (1/total0 + 1/total1))\n",
        "    distr = st.norm(0, 1) \n",
        "    p_value = (1 - distr.cdf(abs(z_value))) * 2\n",
        "    return p_value\n",
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z3XAs_T7WdGe"
      },
      "source": [
        "    H_0: Доля события выборки группы A = доле события группы В\n",
        "    Н_а: Доля события выборки группы А ≠ доле события группы В\n",
        "   \n",
        "    alpha = 0.05"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 75,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 175
        },
        "id": "k7SKkiJFW-5u",
        "outputId": "56be1137-71d6-4c08-93bc-4d9f5226a129",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Группа_1 Группа_2       Событие  p_alpha  p_value  \\\n",
              "0        A        B         login     0.05     0.08   \n",
              "1        A        B  product_page     0.05     0.00   \n",
              "2        A        B  product_cart     0.05     0.21   \n",
              "3        A        B      purchase     0.05     0.05   \n",
              "\n",
              "                                                                       Результат  \n",
              "0  Не получилось отвергнуть нулевую гипотезу, нет оснований считать доли разными  \n",
              "1            Отвергаем нулевую гипотезу - между долями есть статзначимая разница  \n",
              "2  Не получилось отвергнуть нулевую гипотезу, нет оснований считать доли разными  \n",
              "3            Отвергаем нулевую гипотезу - между долями есть статзначимая разница  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ec174119-3bc4-4373-bafc-b81b942dc74f\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Группа_1</th>\n",
              "      <th>Группа_2</th>\n",
              "      <th>Событие</th>\n",
              "      <th>p_alpha</th>\n",
              "      <th>p_value</th>\n",
              "      <th>Результат</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>login</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.08</td>\n",
              "      <td>Не получилось отвергнуть нулевую гипотезу, нет оснований считать доли разными</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>product_page</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.00</td>\n",
              "      <td>Отвергаем нулевую гипотезу - между долями есть статзначимая разница</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>product_cart</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.21</td>\n",
              "      <td>Не получилось отвергнуть нулевую гипотезу, нет оснований считать доли разными</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>purchase</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.05</td>\n",
              "      <td>Отвергаем нулевую гипотезу - между долями есть статзначимая разница</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ec174119-3bc4-4373-bafc-b81b942dc74f')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ec174119-3bc4-4373-bafc-b81b942dc74f button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ec174119-3bc4-4373-bafc-b81b942dc74f');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 75
        }
      ],
      "source": [
        "pd.options.display.max_colwidth = 150\n",
        "AB_pivot = pd.DataFrame(columns=['Группа_1', 'Группа_2', 'Событие', 'p_alpha', 'p_value', 'Результат'])\n",
        " \n",
        "result_options = [\"Отвергаем нулевую гипотезу - между долями есть статзначимая разница\",\n",
        "                  \"Не получилось отвергнуть нулевую гипотезу, нет оснований считать доли разными\"]\n",
        "\n",
        "group_1 = 'A'\n",
        "current_row = users_by_events.loc[group_1]\n",
        "group_2 = 'B'\n",
        "for event_number in range(4):\n",
        "    curr_event = users_by_events.columns[event_number]\n",
        "    p_value_result = st_test(current_row[curr_event], \n",
        "                                   users_by_events[curr_event][group_2],\n",
        "                                   current_row['users_total'],\n",
        "                                   users_by_events['users_total'][group_2],\n",
        "                                   alpha = 0.05)    \n",
        "    new_row = {'Группа_1': group_1,\n",
        "               'Группа_2': group_2,\n",
        "               'Событие': users_by_events.columns[event_number],\n",
        "               'p_alpha': 0.05,\n",
        "               'p_value': p_value_result,\n",
        "               'Результат': result_options[int(p_value_result >= 0.05)]}\n",
        "    AB_pivot = AB_pivot.append([new_row])\n",
        " \n",
        "AB_pivot.index = range(len(AB_pivot))\n",
        "AB_pivot              "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lVE8BwHcY4_n"
      },
      "source": [
        "Тест показал различие в конверсии между группами на этапе `product_page` и `purchase`. Откорректируем наши конечные p_value, использую поправку Холма-Шидака. Коррекция нужна, так как перед нами множественный непараметрический тест."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 76,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "id": "9zG-IRF0XGw7",
        "outputId": "343308ca-a62f-46ef-f7fc-59a61d7be910",
        "trusted": true
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Группа_1 Группа_2       Событие  p_alpha  p_value  p_value_corr  \\\n",
              "0        A        B         login     0.05     0.08          0.16   \n",
              "1        A        B  product_page     0.05     0.00          0.00   \n",
              "2        A        B  product_cart     0.05     0.21          0.21   \n",
              "3        A        B      purchase     0.05     0.05          0.13   \n",
              "\n",
              "                                                                          Итоговый результат  \n",
              "0  Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными  \n",
              "1                        Отвергаем нулевую гипотезу - между долями есть статзначимая разница  \n",
              "2  Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными  \n",
              "3  Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-01b3e819-8c06-4f0d-a256-02530894c4b6\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Группа_1</th>\n",
              "      <th>Группа_2</th>\n",
              "      <th>Событие</th>\n",
              "      <th>p_alpha</th>\n",
              "      <th>p_value</th>\n",
              "      <th>p_value_corr</th>\n",
              "      <th>Итоговый результат</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>login</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.08</td>\n",
              "      <td>0.16</td>\n",
              "      <td>Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>product_page</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>Отвергаем нулевую гипотезу - между долями есть статзначимая разница</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>product_cart</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.21</td>\n",
              "      <td>0.21</td>\n",
              "      <td>Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>A</td>\n",
              "      <td>B</td>\n",
              "      <td>purchase</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.13</td>\n",
              "      <td>Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-01b3e819-8c06-4f0d-a256-02530894c4b6')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-01b3e819-8c06-4f0d-a256-02530894c4b6 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-01b3e819-8c06-4f0d-a256-02530894c4b6');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 76
        }
      ],
      "source": [
        "result_options = [\"Отвергаем нулевую гипотезу - между долями есть статзначимая разница\",\n",
        "                  \"Не получилось отвергнуть нулевую гипотезу, нет достаточных оснований считать доли разными\"]\n",
        " \n",
        "pvalues_corr_list = []  # откорректированные p-value\n",
        "final_results = []  # итоговые результаты (отклоняем или нет)\n",
        "\n",
        "for i in range(0, len(AB_pivot), 4):\n",
        "    pvalues_list = AB_pivot.loc[i:i+3, 'p_value'].tolist()\n",
        "    results, p_value_corr, _, _ = multi_correction.multipletests(pvalues_list, method='holm-sidak')\n",
        "    pvalues_corr_list.extend([round(x, 5) for x in p_value_corr])\n",
        "    final_results.extend([result_options[0] if x else result_options[1] for x in results])\n",
        " \n",
        "AB_pivot = AB_pivot.drop('Результат', axis=1)\n",
        "AB_pivot['p_value_corr'] = pvalues_corr_list\n",
        "AB_pivot['Итоговый результат'] = final_results\n",
        " \n",
        "AB_pivot"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Cp2S5LJaexjs"
      },
      "source": [
        "Лишь на этапе `product_page` обнаружена статистически значимая разница в конверсиях."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vt3VCx2mYoOv"
      },
      "source": [
        "## Выводы"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uASqZXs0YsPk"
      },
      "source": [
        "Для проведения теста выбрано крайне неудачное время. Мощнейшие внешние факторы в виде новогодне-рождественских праздников, распродаж, промо-кампании, 11 дней которой пересекаются с датами проведения теста. Хотя по факту данных о событиях меньше - они заканчиваются не 4 января, когда останавливается тест, а 30 декабря. И то в этот день данные явно неполные.\n",
        "\n",
        "Собранные данные соответствуют ТЗ лишь частично. Количество участников теста меньше заявленных 6000, аудитория сильно меньше планировавшихся 15% от новых пользователей из Европы. В выборках сильная диспропорция. Уникальных пользователей в тестовой группе - лишь 20%  от общего количества. \n",
        "\n",
        "Анализ воронки события показал, что конверсия в событиях контрольной группы A выглядит лучше. Клиентов в долях от первоначального количества теряется меньше. До этапа product_page доходит 65% от первоначального кличества, в группе B только 56%. Ни о каком улучшении метрик не менее, чем на 10%, и речи нет. \n",
        "\n",
        "Конверсия события `product_page` в тестовой группе B более чем на 13% ниже.\n",
        "\n",
        "Конверсия события `product_cart` в тестовой группе B более чем на 7% ниже.\n",
        "\n",
        "Конверсия события `purchase` в тестовой группе B более чем на 12% ниже.\n",
        "\n",
        "Статтест не увидел статистически значимую разницу в конверсиях трех из четырех событий контрольной и тестовой групп. Лишь на этапе `product_page` обнаружена статистически значимая разница в конверсиях.\n",
        "\n",
        "Таким образом, исследование показало, что внедрение улучшенной рекомендательной системы не достигло заявленных целей. Однако нельзя забывать про сильные внешние факторы, а также данные, которые лишь частично соответствовали ТЗ, плюс сомнения в адекватности данных вызывает тот факт, что, начиная с 14 декабря, в контрольную группу перестали попадать уникальные пользователи без событий.\n",
        "\n",
        "Для более корректного исследования рекомендуем повторить тест и учесть недоcтатки, выявленные в ходе данного анализа.\n",
        "\n"
      ]
    }
  ],
  "metadata": {
    "ExecuteTimeLog": [
      {
        "duration": 2381,
        "start_time": "2022-12-03T13:37:35.859Z"
      },
      {
        "duration": 1289,
        "start_time": "2022-12-03T13:37:38.244Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.539Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.541Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.542Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.543Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.544Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.545Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.546Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.547Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.548Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.550Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.551Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.552Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.553Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.555Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.556Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.557Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.558Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.560Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.561Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.562Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.563Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.565Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.566Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.567Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.568Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.569Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.616Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.618Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.619Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.621Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.622Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.624Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.626Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.627Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.628Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.629Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.631Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.632Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.634Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.635Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.636Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.638Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.639Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.641Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.643Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.645Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.646Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.647Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.649Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.651Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.652Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.715Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.717Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.718Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.719Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.721Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.722Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.723Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.725Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.728Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.729Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.731Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.732Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.733Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.734Z"
      },
      {
        "duration": 1,
        "start_time": "2022-12-03T13:37:39.735Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.737Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.738Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.740Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.741Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.742Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.745Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.746Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.748Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T13:37:39.749Z"
      },
      {
        "duration": 2367,
        "start_time": "2022-12-03T16:22:55.626Z"
      },
      {
        "duration": 1551,
        "start_time": "2022-12-03T16:22:57.997Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.551Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.552Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.553Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.554Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.555Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.557Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.558Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.559Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.560Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.561Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.562Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.563Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.578Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.579Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.586Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.587Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.588Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.589Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.590Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.591Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.592Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.593Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.594Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.595Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.597Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.598Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.599Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.600Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.601Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.602Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.604Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.605Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.606Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.607Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.608Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.609Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.611Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.612Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.613Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.614Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.616Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.617Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.618Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.620Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.621Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.622Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.624Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.625Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.626Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.628Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.678Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.679Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.680Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.684Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.684Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.687Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.688Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.688Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.689Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.696Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.697Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.697Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.698Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.699Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.700Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.700Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.701Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.702Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.703Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.704Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.706Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.707Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.708Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.709Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:22:59.710Z"
      },
      {
        "duration": 2379,
        "start_time": "2022-12-03T16:24:20.108Z"
      },
      {
        "duration": 1283,
        "start_time": "2022-12-03T16:24:22.489Z"
      },
      {
        "duration": 2157,
        "start_time": "2022-12-03T16:24:23.774Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.934Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.935Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.936Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.938Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.939Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.940Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.941Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.942Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.944Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.945Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.946Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.947Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.948Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.949Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.950Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.951Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.952Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.953Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.954Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.955Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.957Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.958Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.959Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.960Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.961Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.963Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.964Z"
      },
      {
        "duration": 1,
        "start_time": "2022-12-03T16:24:25.964Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.966Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.967Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.967Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.979Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.980Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.981Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.981Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.982Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.983Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.983Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.984Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.985Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.985Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.986Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.992Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:25.999Z"
      },
      {
        "duration": 1,
        "start_time": "2022-12-03T16:24:25.999Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.002Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.003Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.008Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.010Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.011Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.013Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.016Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.017Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.017Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.020Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.021Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.023Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.024Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.026Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.028Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.029Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.031Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.078Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.079Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.083Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.084Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.085Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.088Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.095Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.096Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.097Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.098Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.099Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:24:26.100Z"
      },
      {
        "duration": 2097,
        "start_time": "2022-12-03T16:25:40.254Z"
      },
      {
        "duration": 2425,
        "start_time": "2022-12-03T16:25:53.174Z"
      },
      {
        "duration": 1228,
        "start_time": "2022-12-03T16:25:55.601Z"
      },
      {
        "duration": 2097,
        "start_time": "2022-12-03T16:25:56.830Z"
      },
      {
        "duration": 1,
        "start_time": "2022-12-03T16:25:58.931Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.933Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.934Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.935Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.936Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.937Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.938Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.939Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.940Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.941Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.943Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.944Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.945Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.946Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.947Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.949Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.950Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.951Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.952Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.953Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.954Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.955Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.956Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.957Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.958Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.960Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.961Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.962Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.963Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.981Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.984Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.990Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.991Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.991Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.992Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.993Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.994Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.997Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.998Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:58.999Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.000Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.001Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.002Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.003Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.005Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.006Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.007Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.008Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.009Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.010Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.020Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.022Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.023Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.027Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.028Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.029Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.031Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.033Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.034Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.081Z"
      },
      {
        "duration": 6,
        "start_time": "2022-12-03T16:25:59.082Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.089Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.090Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.091Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.092Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.093Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.094Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.096Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.097Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.098Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.099Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.100Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.101Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:25:59.103Z"
      },
      {
        "duration": 2487,
        "start_time": "2022-12-03T16:26:19.885Z"
      },
      {
        "duration": 1263,
        "start_time": "2022-12-03T16:26:22.374Z"
      },
      {
        "duration": 2312,
        "start_time": "2022-12-03T16:26:23.638Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.954Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.955Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.957Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.958Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.960Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.960Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.962Z"
      },
      {
        "duration": 1,
        "start_time": "2022-12-03T16:26:25.962Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.964Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.965Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.966Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.967Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.968Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.969Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.970Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.971Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.973Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.974Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.975Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.976Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.979Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.980Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.980Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.981Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.982Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.986Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.987Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.987Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.988Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.989Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.990Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.990Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.991Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.994Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.995Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:25.995Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.000Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.001Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.002Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.002Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.003Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.004Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.005Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.010Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.011Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.015Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.016Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.018Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.019Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.020Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.021Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.023Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.024Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.025Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.026Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.027Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.028Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.029Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.030Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.031Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.032Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.033Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.034Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.077Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.078Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.079Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.079Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.082Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.083Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.087Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.088Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.089Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.092Z"
      },
      {
        "duration": 0,
        "start_time": "2022-12-03T16:26:26.093Z"
      },
      {
        "duration": 2354,
        "start_time": "2022-12-03T16:27:26.157Z"
      },
      {
        "duration": 29,
        "start_time": "2022-12-03T16:39:16.870Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T16:39:28.150Z"
      },
      {
        "duration": 2265,
        "start_time": "2022-12-03T16:39:33.076Z"
      },
      {
        "duration": 94,
        "start_time": "2022-12-03T16:45:01.551Z"
      },
      {
        "duration": 1306,
        "start_time": "2022-12-03T16:45:19.640Z"
      },
      {
        "duration": 82,
        "start_time": "2022-12-03T16:46:15.425Z"
      },
      {
        "duration": 67,
        "start_time": "2022-12-03T16:46:24.944Z"
      },
      {
        "duration": 2402,
        "start_time": "2022-12-03T16:49:49.209Z"
      },
      {
        "duration": 1230,
        "start_time": "2022-12-03T16:49:51.613Z"
      },
      {
        "duration": 1211,
        "start_time": "2022-12-03T16:49:52.844Z"
      },
      {
        "duration": 106,
        "start_time": "2022-12-03T16:49:54.057Z"
      },
      {
        "duration": 21,
        "start_time": "2022-12-03T16:49:54.165Z"
      },
      {
        "duration": 276,
        "start_time": "2022-12-03T16:49:54.189Z"
      },
      {
        "duration": 18,
        "start_time": "2022-12-03T16:49:54.466Z"
      },
      {
        "duration": 56,
        "start_time": "2022-12-03T16:49:54.486Z"
      },
      {
        "duration": 151,
        "start_time": "2022-12-03T16:49:54.544Z"
      },
      {
        "duration": 14,
        "start_time": "2022-12-03T16:49:54.698Z"
      },
      {
        "duration": 18,
        "start_time": "2022-12-03T16:49:54.713Z"
      },
      {
        "duration": 42,
        "start_time": "2022-12-03T16:49:54.733Z"
      },
      {
        "duration": 20,
        "start_time": "2022-12-03T16:49:54.777Z"
      },
      {
        "duration": 27,
        "start_time": "2022-12-03T16:49:54.798Z"
      },
      {
        "duration": 20,
        "start_time": "2022-12-03T16:49:54.827Z"
      },
      {
        "duration": 50,
        "start_time": "2022-12-03T16:49:54.849Z"
      },
      {
        "duration": 58,
        "start_time": "2022-12-03T16:49:54.901Z"
      },
      {
        "duration": 294,
        "start_time": "2022-12-03T16:49:54.961Z"
      },
      {
        "duration": 7,
        "start_time": "2022-12-03T16:49:55.256Z"
      },
      {
        "duration": 29,
        "start_time": "2022-12-03T16:49:55.265Z"
      },
      {
        "duration": 14,
        "start_time": "2022-12-03T16:49:55.295Z"
      },
      {
        "duration": 7,
        "start_time": "2022-12-03T16:49:55.310Z"
      },
      {
        "duration": 288,
        "start_time": "2022-12-03T16:49:55.319Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T16:49:55.609Z"
      },
      {
        "duration": 13,
        "start_time": "2022-12-03T16:49:55.614Z"
      },
      {
        "duration": 17,
        "start_time": "2022-12-03T16:49:55.629Z"
      },
      {
        "duration": 9,
        "start_time": "2022-12-03T16:49:55.647Z"
      },
      {
        "duration": 60,
        "start_time": "2022-12-03T16:49:55.660Z"
      },
      {
        "duration": 13,
        "start_time": "2022-12-03T16:49:55.722Z"
      },
      {
        "duration": 5,
        "start_time": "2022-12-03T16:49:55.737Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T16:49:55.744Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T16:49:55.756Z"
      },
      {
        "duration": 7,
        "start_time": "2022-12-03T16:49:55.768Z"
      },
      {
        "duration": 20,
        "start_time": "2022-12-03T16:49:55.777Z"
      },
      {
        "duration": 26,
        "start_time": "2022-12-03T16:49:55.803Z"
      },
      {
        "duration": 18,
        "start_time": "2022-12-03T16:49:55.831Z"
      },
      {
        "duration": 5,
        "start_time": "2022-12-03T16:49:55.851Z"
      },
      {
        "duration": 15,
        "start_time": "2022-12-03T16:49:55.857Z"
      },
      {
        "duration": 292,
        "start_time": "2022-12-03T16:49:55.874Z"
      },
      {
        "duration": 30,
        "start_time": "2022-12-03T16:49:56.168Z"
      },
      {
        "duration": 16,
        "start_time": "2022-12-03T16:49:56.202Z"
      },
      {
        "duration": 8,
        "start_time": "2022-12-03T16:49:56.219Z"
      },
      {
        "duration": 14,
        "start_time": "2022-12-03T16:49:56.228Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T16:49:56.244Z"
      },
      {
        "duration": 2,
        "start_time": "2022-12-03T16:49:56.257Z"
      },
      {
        "duration": 54,
        "start_time": "2022-12-03T16:49:56.261Z"
      },
      {
        "duration": 16,
        "start_time": "2022-12-03T16:49:56.317Z"
      },
      {
        "duration": 238,
        "start_time": "2022-12-03T16:49:56.334Z"
      },
      {
        "duration": 32,
        "start_time": "2022-12-03T16:49:56.574Z"
      },
      {
        "duration": 122,
        "start_time": "2022-12-03T16:49:56.609Z"
      },
      {
        "duration": 13,
        "start_time": "2022-12-03T16:49:56.733Z"
      },
      {
        "duration": 29,
        "start_time": "2022-12-03T16:49:56.747Z"
      },
      {
        "duration": 35,
        "start_time": "2022-12-03T16:49:56.783Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T16:49:56.820Z"
      },
      {
        "duration": 13,
        "start_time": "2022-12-03T16:49:56.826Z"
      },
      {
        "duration": 15,
        "start_time": "2022-12-03T16:49:56.841Z"
      },
      {
        "duration": 3,
        "start_time": "2022-12-03T16:49:56.857Z"
      },
      {
        "duration": 17,
        "start_time": "2022-12-03T16:49:56.862Z"
      },
      {
        "duration": 29,
        "start_time": "2022-12-03T16:49:56.883Z"
      },
      {
        "duration": 13,
        "start_time": "2022-12-03T16:49:56.914Z"
      },
      {
        "duration": 217,
        "start_time": "2022-12-03T16:49:56.929Z"
      },
      {
        "duration": 3,
        "start_time": "2022-12-03T16:49:57.147Z"
      },
      {
        "duration": 38,
        "start_time": "2022-12-03T16:49:57.152Z"
      },
      {
        "duration": 20,
        "start_time": "2022-12-03T16:49:57.202Z"
      },
      {
        "duration": 139,
        "start_time": "2022-12-03T16:49:57.224Z"
      },
      {
        "duration": 236,
        "start_time": "2022-12-03T16:49:57.365Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T16:49:57.602Z"
      },
      {
        "duration": 434,
        "start_time": "2022-12-03T16:49:57.614Z"
      },
      {
        "duration": 9,
        "start_time": "2022-12-03T16:49:58.050Z"
      },
      {
        "duration": 40,
        "start_time": "2022-12-03T16:49:58.061Z"
      },
      {
        "duration": 17,
        "start_time": "2022-12-03T16:49:58.105Z"
      },
      {
        "duration": 233,
        "start_time": "2022-12-03T16:49:58.124Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T16:49:58.359Z"
      },
      {
        "duration": 44,
        "start_time": "2022-12-03T16:49:58.371Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T16:49:58.416Z"
      },
      {
        "duration": 82,
        "start_time": "2022-12-03T16:49:58.421Z"
      },
      {
        "duration": 19,
        "start_time": "2022-12-03T16:49:58.504Z"
      },
      {
        "duration": 2310,
        "start_time": "2022-12-03T21:15:46.303Z"
      },
      {
        "duration": 1330,
        "start_time": "2022-12-03T21:15:48.615Z"
      },
      {
        "duration": 1488,
        "start_time": "2022-12-03T21:15:49.946Z"
      },
      {
        "duration": 111,
        "start_time": "2022-12-03T21:15:51.436Z"
      },
      {
        "duration": 12,
        "start_time": "2022-12-03T21:15:51.548Z"
      },
      {
        "duration": 202,
        "start_time": "2022-12-03T21:15:51.561Z"
      },
      {
        "duration": 17,
        "start_time": "2022-12-03T21:15:51.764Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T21:15:51.783Z"
      },
      {
        "duration": 120,
        "start_time": "2022-12-03T21:15:51.794Z"
      },
      {
        "duration": 17,
        "start_time": "2022-12-03T21:15:51.918Z"
      },
      {
        "duration": 15,
        "start_time": "2022-12-03T21:15:51.937Z"
      },
      {
        "duration": 143,
        "start_time": "2022-12-03T21:15:51.953Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T21:15:52.098Z"
      },
      {
        "duration": 14,
        "start_time": "2022-12-03T21:15:52.109Z"
      },
      {
        "duration": 31,
        "start_time": "2022-12-03T21:15:52.125Z"
      },
      {
        "duration": 35,
        "start_time": "2022-12-03T21:15:52.158Z"
      },
      {
        "duration": 69,
        "start_time": "2022-12-03T21:15:52.194Z"
      },
      {
        "duration": 271,
        "start_time": "2022-12-03T21:15:52.265Z"
      },
      {
        "duration": 8,
        "start_time": "2022-12-03T21:15:52.538Z"
      },
      {
        "duration": 16,
        "start_time": "2022-12-03T21:15:52.547Z"
      },
      {
        "duration": 5,
        "start_time": "2022-12-03T21:15:52.565Z"
      },
      {
        "duration": 23,
        "start_time": "2022-12-03T21:15:52.571Z"
      },
      {
        "duration": 304,
        "start_time": "2022-12-03T21:15:52.596Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T21:15:52.902Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:52.917Z"
      },
      {
        "duration": 25,
        "start_time": "2022-12-03T21:15:52.930Z"
      },
      {
        "duration": 6,
        "start_time": "2022-12-03T21:15:52.957Z"
      },
      {
        "duration": 12,
        "start_time": "2022-12-03T21:15:52.967Z"
      },
      {
        "duration": 9,
        "start_time": "2022-12-03T21:15:52.980Z"
      },
      {
        "duration": 28,
        "start_time": "2022-12-03T21:15:52.991Z"
      },
      {
        "duration": 24,
        "start_time": "2022-12-03T21:15:53.021Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T21:15:53.047Z"
      },
      {
        "duration": 6,
        "start_time": "2022-12-03T21:15:53.059Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:53.067Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:53.079Z"
      },
      {
        "duration": 31,
        "start_time": "2022-12-03T21:15:53.092Z"
      },
      {
        "duration": 24,
        "start_time": "2022-12-03T21:15:53.127Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:53.153Z"
      },
      {
        "duration": 294,
        "start_time": "2022-12-03T21:15:53.166Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:53.462Z"
      },
      {
        "duration": 6,
        "start_time": "2022-12-03T21:15:53.475Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T21:15:53.482Z"
      },
      {
        "duration": 40,
        "start_time": "2022-12-03T21:15:53.488Z"
      },
      {
        "duration": 13,
        "start_time": "2022-12-03T21:15:53.530Z"
      },
      {
        "duration": 3,
        "start_time": "2022-12-03T21:15:53.545Z"
      },
      {
        "duration": 39,
        "start_time": "2022-12-03T21:15:53.549Z"
      },
      {
        "duration": 40,
        "start_time": "2022-12-03T21:15:53.590Z"
      },
      {
        "duration": 238,
        "start_time": "2022-12-03T21:15:53.632Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T21:15:53.872Z"
      },
      {
        "duration": 123,
        "start_time": "2022-12-03T21:15:53.884Z"
      },
      {
        "duration": 18,
        "start_time": "2022-12-03T21:15:54.009Z"
      },
      {
        "duration": 31,
        "start_time": "2022-12-03T21:15:54.030Z"
      },
      {
        "duration": 27,
        "start_time": "2022-12-03T21:15:54.063Z"
      },
      {
        "duration": 3,
        "start_time": "2022-12-03T21:15:54.092Z"
      },
      {
        "duration": 16,
        "start_time": "2022-12-03T21:15:54.115Z"
      },
      {
        "duration": 16,
        "start_time": "2022-12-03T21:15:54.133Z"
      },
      {
        "duration": 5,
        "start_time": "2022-12-03T21:15:54.151Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:54.157Z"
      },
      {
        "duration": 18,
        "start_time": "2022-12-03T21:15:54.170Z"
      },
      {
        "duration": 35,
        "start_time": "2022-12-03T21:15:54.190Z"
      },
      {
        "duration": 229,
        "start_time": "2022-12-03T21:15:54.227Z"
      },
      {
        "duration": 3,
        "start_time": "2022-12-03T21:15:54.457Z"
      },
      {
        "duration": 28,
        "start_time": "2022-12-03T21:15:54.462Z"
      },
      {
        "duration": 26,
        "start_time": "2022-12-03T21:15:54.492Z"
      },
      {
        "duration": 163,
        "start_time": "2022-12-03T21:15:54.520Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T21:15:54.684Z"
      },
      {
        "duration": 477,
        "start_time": "2022-12-03T21:15:54.696Z"
      },
      {
        "duration": 10,
        "start_time": "2022-12-03T21:15:55.175Z"
      },
      {
        "duration": 38,
        "start_time": "2022-12-03T21:15:55.187Z"
      },
      {
        "duration": 46,
        "start_time": "2022-12-03T21:15:55.227Z"
      },
      {
        "duration": 395,
        "start_time": "2022-12-03T21:15:55.274Z"
      },
      {
        "duration": 11,
        "start_time": "2022-12-03T21:15:55.670Z"
      },
      {
        "duration": 67,
        "start_time": "2022-12-03T21:15:55.683Z"
      },
      {
        "duration": 4,
        "start_time": "2022-12-03T21:15:55.752Z"
      },
      {
        "duration": 91,
        "start_time": "2022-12-03T21:15:55.758Z"
      },
      {
        "duration": 55,
        "start_time": "2022-12-03T21:15:55.851Z"
      }
    ],
    "colab": {
      "provenance": [],
      "toc_visible": true
    },
    "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": "3.9.5"
    },
    "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": false,
      "toc_position": {},
      "toc_section_display": true,
      "toc_window_display": true
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}