sem09_pandas/sem09_pandas_advanced_2012-209.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "16d9b3ae",
   "metadata": {
    "toc": true
   },
   "source": [
    "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n",
    "<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#Pandas:-exploratory-data-analysis\" data-toc-modified-id=\"Pandas:-exploratory-data-analysis-1\"><span class=\"toc-item-num\">1&nbsp;&nbsp;</span>Pandas: exploratory data analysis</a></span><ul class=\"toc-item\"><li><ul class=\"toc-item\"><li><span><a href=\"#План:\" data-toc-modified-id=\"План:-1.0.1\"><span class=\"toc-item-num\">1.0.1&nbsp;&nbsp;</span>План:</a></span></li></ul></li><li><span><a href=\"#Соединения-таблиц\" data-toc-modified-id=\"Соединения-таблиц-1.1\"><span class=\"toc-item-num\">1.1&nbsp;&nbsp;</span>Соединения таблиц</a></span><ul class=\"toc-item\"><li><span><a href=\"#1.-Join\" data-toc-modified-id=\"1.-Join-1.1.1\"><span class=\"toc-item-num\">1.1.1&nbsp;&nbsp;</span>1. Join</a></span></li><li><span><a href=\"#2.-Конкатенация-строк\" data-toc-modified-id=\"2.-Конкатенация-строк-1.1.2\"><span class=\"toc-item-num\">1.1.2&nbsp;&nbsp;</span>2. Конкатенация строк</a></span></li><li><span><a href=\"#Пример-на-клипах-по-ютубу\" data-toc-modified-id=\"Пример-на-клипах-по-ютубу-1.1.3\"><span class=\"toc-item-num\">1.1.3&nbsp;&nbsp;</span>Пример на клипах по ютубу</a></span></li><li><span><a href=\"#Задача-№1:\" data-toc-modified-id=\"Задача-№1:-1.1.4\"><span class=\"toc-item-num\">1.1.4&nbsp;&nbsp;</span>Задача №1:</a></span></li><li><span><a href=\"#Задача-№2:\" data-toc-modified-id=\"Задача-№2:-1.1.5\"><span class=\"toc-item-num\">1.1.5&nbsp;&nbsp;</span>Задача №2:</a></span></li></ul></li><li><span><a href=\"#Типы-переменных\" data-toc-modified-id=\"Типы-переменных-1.2\"><span class=\"toc-item-num\">1.2&nbsp;&nbsp;</span>Типы переменных</a></span></li><li><span><a href=\"#Работа-с-категориальными-переменными\" data-toc-modified-id=\"Работа-с-категориальными-переменными-1.3\"><span class=\"toc-item-num\">1.3&nbsp;&nbsp;</span>Работа с категориальными переменными</a></span><ul class=\"toc-item\"><li><span><a href=\"#Когда-категории-две:\" data-toc-modified-id=\"Когда-категории-две:-1.3.1\"><span class=\"toc-item-num\">1.3.1&nbsp;&nbsp;</span>Когда категории две:</a></span></li><li><span><a href=\"#Задача-№3:\" data-toc-modified-id=\"Задача-№3:-1.3.2\"><span class=\"toc-item-num\">1.3.2&nbsp;&nbsp;</span>Задача №3:</a></span></li><li><span><a href=\"#Когда-категорий-больше-двух\" data-toc-modified-id=\"Когда-категорий-больше-двух-1.3.3\"><span class=\"toc-item-num\">1.3.3&nbsp;&nbsp;</span>Когда категорий больше двух</a></span></li></ul></li><li><span><a href=\"#Работа-с-датой-и-временем\" data-toc-modified-id=\"Работа-с-датой-и-временем-1.4\"><span class=\"toc-item-num\">1.4&nbsp;&nbsp;</span>Работа с датой и временем</a></span><ul class=\"toc-item\"><li><span><a href=\"#Задача-№4:\" data-toc-modified-id=\"Задача-№4:-1.4.1\"><span class=\"toc-item-num\">1.4.1&nbsp;&nbsp;</span>Задача №4:</a></span></li><li><span><a href=\"#Задача-№5:\" data-toc-modified-id=\"Задача-№5:-1.4.2\"><span class=\"toc-item-num\">1.4.2&nbsp;&nbsp;</span>Задача №5:</a></span></li><li><span><a href=\"#Задача-№6:\" data-toc-modified-id=\"Задача-№6:-1.4.3\"><span class=\"toc-item-num\">1.4.3&nbsp;&nbsp;</span>Задача №6:</a></span></li></ul></li><li><span><a href=\"#Пропущенные-значения\" data-toc-modified-id=\"Пропущенные-значения-1.5\"><span class=\"toc-item-num\">1.5&nbsp;&nbsp;</span>Пропущенные значения</a></span><ul class=\"toc-item\"><li><span><a href=\"#Выявление\" data-toc-modified-id=\"Выявление-1.5.1\"><span class=\"toc-item-num\">1.5.1&nbsp;&nbsp;</span>Выявление</a></span></li><li><span><a href=\"#Удаление\" data-toc-modified-id=\"Удаление-1.5.2\"><span class=\"toc-item-num\">1.5.2&nbsp;&nbsp;</span>Удаление</a></span></li><li><span><a href=\"#Заполнение\" data-toc-modified-id=\"Заполнение-1.5.3\"><span class=\"toc-item-num\">1.5.3&nbsp;&nbsp;</span>Заполнение</a></span></li><li><span><a href=\"#Задача-№7:\" data-toc-modified-id=\"Задача-№7:-1.5.4\"><span class=\"toc-item-num\">1.5.4&nbsp;&nbsp;</span>Задача №7:</a></span></li></ul></li><li><span><a href=\"#Обнаружение-аномалий\" data-toc-modified-id=\"Обнаружение-аномалий-1.6\"><span class=\"toc-item-num\">1.6&nbsp;&nbsp;</span>Обнаружение аномалий</a></span><ul class=\"toc-item\"><li><span><a href=\"#Оцениваем-средний-рост-людей-в-школьном-классе:\" data-toc-modified-id=\"Оцениваем-средний-рост-людей-в-школьном-классе:-1.6.1\"><span class=\"toc-item-num\">1.6.1&nbsp;&nbsp;</span>Оцениваем средний рост людей в школьном классе:</a></span></li><li><span><a href=\"#Выявление-и-удаление-выбросов-на-примере-клипов:\" data-toc-modified-id=\"Выявление-и-удаление-выбросов-на-примере-клипов:-1.6.2\"><span class=\"toc-item-num\">1.6.2&nbsp;&nbsp;</span>Выявление и удаление выбросов на примере клипов:</a></span></li><li><span><a href=\"#Задача-№8:\" data-toc-modified-id=\"Задача-№8:-1.6.3\"><span class=\"toc-item-num\">1.6.3&nbsp;&nbsp;</span>Задача №8:</a></span></li></ul></li></ul></li></ul></div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "active-finding",
   "metadata": {},
   "source": [
    "# Pandas: exploratory data analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "rubber-people",
   "metadata": {},
   "source": [
    "### План:\n",
    "* Соединения таблиц\n",
    "* Типы переменных\n",
    "* Работа с категориальными переменными\n",
    "* Работа с датой и временем\n",
    "* Пропущенные значения\n",
    "* Обнаружение аномалий"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "sticky-thickness",
   "metadata": {},
   "source": [
    "Также по ходу знакомимся с визуализацией, но подробнее об этом на следующей паре"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "regional-embassy",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np         # библиотека для матриц и математики\n",
    "import pandas as pd        # библиотека для работы с табличками\n",
    "from scipy import stats    # модуль для работы со статистикой\n",
    "\n",
    "# библиотеки для визуализации\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "plt.style.use('ggplot')    # стиль графиков\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "reported-shade",
   "metadata": {},
   "source": [
    "## Соединения таблиц"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bizarre-premises",
   "metadata": {},
   "source": [
    "Часто хранить информацию в одной таблице бывает довольно неудобно. Когда у вас очень крупная IT-компания, данных настолько много, что если их добавить в одну таблицу, работать с ней будет очень неэффективно по времени и памяти. Поэтому информацию кладут в кучу разных таблиц, а над ними строят специальные *отношения* - так называемые колонки, по которым можно эти таблицы объединять. \n",
    "\n",
    "Пример устройства такой структуры хранения данных:\n",
    "\n",
    "<center>\n",
    "<img src=\"https://progi.pro/media/main/82/d3/34/82d334f5b4b49d424a2460e796a30cd5.png\" height=\"200\" width=\"600\">\n",
    "</center>   "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "worse-pathology",
   "metadata": {},
   "source": [
    "### 1. Join"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "excited-problem",
   "metadata": {},
   "source": [
    "<center>\n",
    "<img src=\"https://community.qlik.com/legacyfs/online/87693_all-joins.png\" height=\"400\" width=\"400\">\n",
    "</center>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "exciting-transport",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Student</th>\n",
       "      <th>group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tom</td>\n",
       "      <td>01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ujin</td>\n",
       "      <td>02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ann</td>\n",
       "      <td>02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Polina</td>\n",
       "      <td>01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Sam</td>\n",
       "      <td>02</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Student group\n",
       "0     Tom    01\n",
       "1    Ujin    02\n",
       "2     Ann    02\n",
       "3  Polina    01\n",
       "4     Sam    02"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Name</th>\n",
       "      <th>GPA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tom</td>\n",
       "      <td>7.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ujin</td>\n",
       "      <td>6.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ann</td>\n",
       "      <td>8.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Polina</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Kit</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Name  GPA\n",
       "0     Tom  7.8\n",
       "1    Ujin  6.4\n",
       "2     Ann  8.3\n",
       "3  Polina    9\n",
       "4     Kit   10"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df1 = pd.DataFrame({'Student': ['Tom', 'Ujin', 'Ann', 'Polina','Sam'],\n",
    "                    'group': ['01', '02', '02', '01','02']})\n",
    "df2 = pd.DataFrame({'Name': ['Tom', 'Ujin', 'Ann', 'Polina', 'Kit'],\n",
    "                    'GPA': ['7.8', '6.4', '8.3', '9', '10']})\n",
    "display(df1, df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "delayed-judge",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Student</th>\n",
       "      <th>group</th>\n",
       "      <th>Name</th>\n",
       "      <th>GPA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tom</td>\n",
       "      <td>01</td>\n",
       "      <td>Tom</td>\n",
       "      <td>7.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ujin</td>\n",
       "      <td>02</td>\n",
       "      <td>Ujin</td>\n",
       "      <td>6.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ann</td>\n",
       "      <td>02</td>\n",
       "      <td>Ann</td>\n",
       "      <td>8.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Polina</td>\n",
       "      <td>01</td>\n",
       "      <td>Polina</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Student group    Name  GPA\n",
       "0     Tom    01     Tom  7.8\n",
       "1    Ujin    02    Ujin  6.4\n",
       "2     Ann    02     Ann  8.3\n",
       "3  Polina    01  Polina    9"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# inner join по умолчанию \n",
    "pd.merge(df1, df2, left_on='Student', right_on='Name')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "continent-advantage",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Student</th>\n",
       "      <th>group</th>\n",
       "      <th>Name</th>\n",
       "      <th>GPA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tom</td>\n",
       "      <td>01</td>\n",
       "      <td>Tom</td>\n",
       "      <td>7.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ujin</td>\n",
       "      <td>02</td>\n",
       "      <td>Ujin</td>\n",
       "      <td>6.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ann</td>\n",
       "      <td>02</td>\n",
       "      <td>Ann</td>\n",
       "      <td>8.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Polina</td>\n",
       "      <td>01</td>\n",
       "      <td>Polina</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Sam</td>\n",
       "      <td>02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Student group    Name  GPA\n",
       "0     Tom    01     Tom  7.8\n",
       "1    Ujin    02    Ujin  6.4\n",
       "2     Ann    02     Ann  8.3\n",
       "3  Polina    01  Polina    9\n",
       "4     Sam    02     NaN  NaN"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# left join - оставляем все, что в левой таблице\n",
    "pd.merge(df1, df2, left_on='Student', right_on='Name', how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "consecutive-wrestling",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Student</th>\n",
       "      <th>group</th>\n",
       "      <th>Name</th>\n",
       "      <th>GPA</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Tom</td>\n",
       "      <td>01</td>\n",
       "      <td>Tom</td>\n",
       "      <td>7.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ujin</td>\n",
       "      <td>02</td>\n",
       "      <td>Ujin</td>\n",
       "      <td>6.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Ann</td>\n",
       "      <td>02</td>\n",
       "      <td>Ann</td>\n",
       "      <td>8.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Polina</td>\n",
       "      <td>01</td>\n",
       "      <td>Polina</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Sam</td>\n",
       "      <td>02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Kit</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Student group    Name  GPA\n",
       "0     Tom    01     Tom  7.8\n",
       "1    Ujin    02    Ujin  6.4\n",
       "2     Ann    02     Ann  8.3\n",
       "3  Polina    01  Polina    9\n",
       "4     Sam    02     NaN  NaN\n",
       "5     NaN   NaN     Kit   10"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# outer join\n",
    "pd.merge(df1, df2, left_on='Student', right_on='Name', how='outer')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "patient-television",
   "metadata": {},
   "source": [
    "### 2. Конкатенация строк"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "improving-inclusion",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>students</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Max</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Phil</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Vlad</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Nastya</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     name  students\n",
       "0     Max        20\n",
       "1    Phil        21\n",
       "2    Vlad        30\n",
       "3  Nastya        42"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>students</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Ahmed</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Andrey</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     name  students\n",
       "0   Ahmed        50\n",
       "1  Andrey        40"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df1 = pd.DataFrame({'name': ['Max', 'Phil', 'Vlad', 'Nastya'], 'students': [20, 21, 30, 42]})\n",
    "df2 = pd.DataFrame({'name': ['Ahmed', 'Andrey'], 'students': [50, 40]})\n",
    "\n",
    "display(df1, df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "competitive-graph",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>students</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Max</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Phil</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Vlad</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Nastya</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Ahmed</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Andrey</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     name  students\n",
       "0     Max        20\n",
       "1    Phil        21\n",
       "2    Vlad        30\n",
       "3  Nastya        42\n",
       "4   Ahmed        50\n",
       "5  Andrey        40"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([df1, df2]).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "nutritional-murray",
   "metadata": {},
   "source": [
    "### Пример на клипах по ютубу"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "brief-patrick",
   "metadata": {},
   "source": [
    "Работать в дальнейшем будем с базой данных, содержащей информацию о клипах ютуба. В базе несколько таблиц, между которыми выстроены соединения. Структура выглядит таким образом:  \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "brilliant-vanilla",
   "metadata": {},
   "source": [
    "<center>\n",
    "<img src=\"https://github.com/hse-econ-data-science/dap_2021_spring/blob/main/sem09_pandas/images/struct.png?raw=true\" height=\"500\" width=\"500\">\n",
    "</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "future-shell",
   "metadata": {},
   "source": [
    "__Описание колонок:__\n",
    "\n",
    "* `id_title` $-$ уникальный id названия видео\n",
    "* `id_music_style` $-$ уникальный id жанра\n",
    "* `id_performer` $-$ уникальный id исполнителя\n",
    "* `title` $-$ название видео\n",
    "* `commentCount` $-$ число комментариев\n",
    "* `dislikeCount` $-$ число дизлайков\n",
    "* `likeCount` $-$ число лайков\n",
    "* `music_style` $-$ музыкальный стиль\n",
    "* `performer` $-$ исполнитель\n",
    "* `viewCount` $-$ число просмотров\n",
    "* `commentators_uniq` $-$ число уникальных комментаторов\n",
    "* `comments_obscene_cnt` $-$ число негативных комментариев\n",
    "* `video_age` $-$ возраст видео в днях (на момент '2019-03-15')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "modern-shoulder",
   "metadata": {},
   "outputs": [],
   "source": [
    "# подгрузим все таблицы\n",
    "df_perf = pd.read_csv('data/performer_table.csv', sep=';')\n",
    "df_title = pd.read_csv('data/title_table.csv', sep=';')\n",
    "df_genre = pd.read_csv('data/music_style_table.csv', sep=';')\n",
    "\n",
    "df_stat = pd.read_csv('data/title_stat_table.csv', sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "13cc4742",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_performer</th>\n",
       "      <th>performer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>99991</td>\n",
       "      <td>2маши</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>99992</td>\n",
       "      <td>лазарев</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id_performer performer\n",
       "0         99991     2маши\n",
       "1         99992   лазарев"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_title</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>98765431</td>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>98765432</td>\n",
       "      <td>\"Моя любимая кухня\". Дорада на скорую руку</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id_title                                       title\n",
       "0  98765431                   \"#2Маши \"Свободные линии\"\n",
       "1  98765432  \"Моя любимая кухня\". Дорада на скорую руку"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_music_style</th>\n",
       "      <th>music_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12345</td>\n",
       "      <td>popsa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12346</td>\n",
       "      <td>rock</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id_music_style music_style\n",
       "0           12345       popsa\n",
       "1           12346        rock"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>id_music_style</th>\n",
       "      <th>id_performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>98765431</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>12345</td>\n",
       "      <td>99991</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>98765432</td>\n",
       "      <td>36.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>12345</td>\n",
       "      <td>99991</td>\n",
       "      <td>74666.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2160</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id_title  commentCount  dislikeCount  likeCount  id_music_style  \\\n",
       "0  98765431          34.0          43.0     1402.0           12345   \n",
       "1  98765432          36.0          90.0      512.0           12345   \n",
       "\n",
       "   id_performer  viewCount  commentators_uniq  comments_obscene_cnt  video_age  \n",
       "0         99991   214301.0               24.0                   0.0       1141  \n",
       "1         99991    74666.0               27.0                   0.0       2160  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for df in [df_perf,df_title,df_genre,df_stat]:\n",
    "    display(df.head(2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "b1cf927b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19124"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stat['id_title'].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "40a3be6d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19124"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_stat['id_title'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "tutorial-priest",
   "metadata": {},
   "source": [
    "### Задача №1:\n",
    "\n",
    "Какой исполнитель получает больше всего хейта? (хейт определяется числом негативных комментариев) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "automated-living",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Ваш код\n",
    "df = pd.merge(df_stat, df_perf, left_on='id_performer', right_on='id_performer')\n",
    "agg = df.groupby('performer')['comments_obscene_cnt'].sum()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "856c1a3a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "performer\n",
       "gazgolder       24865.0\n",
       "ленинград       12013.0\n",
       "oxxxymiron       7190.0\n",
       "элджей           6789.0\n",
       "тимати           5958.0\n",
       "noizemc          5271.0\n",
       "lone             3576.0\n",
       "лсп              3504.0\n",
       "пошлая_молли     3345.0\n",
       "face             3107.0\n",
       "Name: comments_obscene_cnt, dtype: float64"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg.sort_values(ascending=False)[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "049e40d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'gazgolder'"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg.idxmax()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "missing-mailing",
   "metadata": {},
   "source": [
    "### Задача №2:\n",
    "\n",
    "Для каждого жанра выведите число просмотров самого популярного клипа? Кто его исполнитель? Как называется клип?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "5e78d39f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_music_style</th>\n",
       "      <th>music_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12345</td>\n",
       "      <td>popsa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12346</td>\n",
       "      <td>rock</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12347</td>\n",
       "      <td>classic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12348</td>\n",
       "      <td>indi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12349</td>\n",
       "      <td>estrada</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>12350</td>\n",
       "      <td>shanson</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>12351</td>\n",
       "      <td>rap</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id_music_style music_style\n",
       "0           12345       popsa\n",
       "1           12346        rock\n",
       "2           12347     classic\n",
       "3           12348        indi\n",
       "4           12349     estrada\n",
       "5           12350     shanson\n",
       "6           12351         rap"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_genre"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "north-respondent",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "music_style\n",
       "classic     9256\n",
       "estrada    13660\n",
       "indi       10301\n",
       "popsa       2409\n",
       "rap        18293\n",
       "rock        7229\n",
       "shanson    15602\n",
       "Name: viewCount, dtype: int64"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Ваш код\n",
    "\n",
    "df = pd.merge(df_stat, df_perf, on='id_performer')\n",
    "df = pd.merge(df, df_title, on='id_title')\n",
    "df = pd.merge(df, df_genre, on='id_music_style')\n",
    "\n",
    "\n",
    "\n",
    "agg = df.groupby('music_style')['viewCount'].idxmax()\n",
    "agg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "beb422f8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 9256, 13660, 10301,  2409, 18293,  7229, 15602])"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agg.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "b7e3f8e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>performer</th>\n",
       "      <th>title</th>\n",
       "      <th>music_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9256</th>\n",
       "      <td>classical_channel</td>\n",
       "      <td>Four Seasons ~ Vivaldi</td>\n",
       "      <td>classic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13660</th>\n",
       "      <td>михайлов</td>\n",
       "      <td>Стас Михайлов и Елена Север - Не зови, не слыш...</td>\n",
       "      <td>estrada</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10301</th>\n",
       "      <td>мальбэк</td>\n",
       "      <td>Мальбэк — Равнодушие ft. Сюзанна</td>\n",
       "      <td>indi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2409</th>\n",
       "      <td>loboda</td>\n",
       "      <td>LOBODA — Твои Глаза</td>\n",
       "      <td>popsa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18293</th>\n",
       "      <td>miyagi&amp;эндшпиль</td>\n",
       "      <td>Miyagi &amp; Эндшпиль feat. Рем Дигга - I Got Love...</td>\n",
       "      <td>rap</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7229</th>\n",
       "      <td>ленинград</td>\n",
       "      <td>Ленинград — Экспонат</td>\n",
       "      <td>rock</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15602</th>\n",
       "      <td>ирина_круг</td>\n",
       "      <td>Ирина Круг и Виктор Королев - Букет из белых роз</td>\n",
       "      <td>shanson</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               performer                                              title  \\\n",
       "9256   classical_channel                             Four Seasons ~ Vivaldi   \n",
       "13660           михайлов  Стас Михайлов и Елена Север - Не зови, не слыш...   \n",
       "10301            мальбэк                   Мальбэк — Равнодушие ft. Сюзанна   \n",
       "2409              loboda                                LOBODA — Твои Глаза   \n",
       "18293    miyagi&эндшпиль  Miyagi & Эндшпиль feat. Рем Дигга - I Got Love...   \n",
       "7229           ленинград                               Ленинград — Экспонат   \n",
       "15602         ирина_круг   Ирина Круг и Виктор Королев - Букет из белых роз   \n",
       "\n",
       "      music_style  \n",
       "9256      classic  \n",
       "13660     estrada  \n",
       "10301        indi  \n",
       "2409        popsa  \n",
       "18293         rap  \n",
       "7229         rock  \n",
       "15602     shanson  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.iloc[agg.values][['performer','title','music_style',]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "0bcf5ef2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "music_style\n",
       "classic    196812152.0\n",
       "estrada     43969730.0\n",
       "indi        48972150.0\n",
       "popsa      132015566.0\n",
       "rap        272047706.0\n",
       "rock       141509635.0\n",
       "shanson     57860686.0\n",
       "Name: viewCount, dtype: float64"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# способ 2\n",
    "\n",
    "agg = df.groupby('music_style')['viewCount'].max()\n",
    "agg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "1a26f61c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>performer</th>\n",
       "      <th>title</th>\n",
       "      <th>music_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>classical_channel</td>\n",
       "      <td>Four Seasons ~ Vivaldi</td>\n",
       "      <td>classic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>михайлов</td>\n",
       "      <td>Стас Михайлов и Елена Север - Не зови, не слыш...</td>\n",
       "      <td>estrada</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>мальбэк</td>\n",
       "      <td>Мальбэк — Равнодушие ft. Сюзанна</td>\n",
       "      <td>indi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>loboda</td>\n",
       "      <td>LOBODA — Твои Глаза</td>\n",
       "      <td>popsa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>miyagi&amp;эндшпиль</td>\n",
       "      <td>Miyagi &amp; Эндшпиль feat. Рем Дигга - I Got Love...</td>\n",
       "      <td>rap</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>ленинград</td>\n",
       "      <td>Ленинград — Экспонат</td>\n",
       "      <td>rock</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ирина_круг</td>\n",
       "      <td>Ирина Круг и Виктор Королев - Букет из белых роз</td>\n",
       "      <td>shanson</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           performer                                              title  \\\n",
       "0  classical_channel                             Four Seasons ~ Vivaldi   \n",
       "1           михайлов  Стас Михайлов и Елена Север - Не зови, не слыш...   \n",
       "2            мальбэк                   Мальбэк — Равнодушие ft. Сюзанна   \n",
       "3             loboda                                LOBODA — Твои Глаза   \n",
       "4    miyagi&эндшпиль  Miyagi & Эндшпиль feat. Рем Дигга - I Got Love...   \n",
       "5          ленинград                               Ленинград — Экспонат   \n",
       "6         ирина_круг   Ирина Круг и Виктор Королев - Букет из белых роз   \n",
       "\n",
       "  music_style  \n",
       "0     classic  \n",
       "1     estrada  \n",
       "2        indi  \n",
       "3       popsa  \n",
       "4         rap  \n",
       "5        rock  \n",
       "6     shanson  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.merge(agg, df, on='viewCount')[['performer','title','music_style',]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "german-wrapping",
   "metadata": {},
   "source": [
    "## Типы переменных"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "93782ff4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id_title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>id_music_style</th>\n",
       "      <th>id_performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "      <th>performer</th>\n",
       "      <th>title</th>\n",
       "      <th>music_style</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>98765431</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>12345</td>\n",
       "      <td>99991</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "      <td>2маши</td>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>popsa</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>98765432</td>\n",
       "      <td>36.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>512.0</td>\n",
       "      <td>12345</td>\n",
       "      <td>99991</td>\n",
       "      <td>74666.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2160</td>\n",
       "      <td>2маши</td>\n",
       "      <td>\"Моя любимая кухня\". Дорада на скорую руку</td>\n",
       "      <td>popsa</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id_title  commentCount  dislikeCount  likeCount  id_music_style  \\\n",
       "0  98765431          34.0          43.0     1402.0           12345   \n",
       "1  98765432          36.0          90.0      512.0           12345   \n",
       "\n",
       "   id_performer  viewCount  commentators_uniq  comments_obscene_cnt  \\\n",
       "0         99991   214301.0               24.0                   0.0   \n",
       "1         99991    74666.0               27.0                   0.0   \n",
       "\n",
       "   video_age performer                                       title music_style  \n",
       "0       1141     2маши                   \"#2Маши \"Свободные линии\"       popsa  \n",
       "1       2160     2маши  \"Моя любимая кухня\". Дорада на скорую руку       popsa  "
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "678c964a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "id_title                  int64\n",
       "commentCount            float64\n",
       "dislikeCount            float64\n",
       "likeCount               float64\n",
       "id_music_style            int64\n",
       "id_performer              int64\n",
       "viewCount               float64\n",
       "commentators_uniq       float64\n",
       "comments_obscene_cnt    float64\n",
       "video_age                 int64\n",
       "performer                object\n",
       "title                    object\n",
       "music_style              object\n",
       "dtype: object"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "complex-delight",
   "metadata": {},
   "source": [
    "__Классификация переменных по типу данных:__  \n",
    "\n",
    "* `int` $-$ числовой, целые числа (1000, 0, -1)\n",
    "* `float` $-$ числовой, вещественные числа (1.0, 2.564523)\n",
    "* `object` $-$ строковый (бутырка, JPMorgan, f12aaz//)\n",
    "* `bool` $-$ булевский (True, False)\n",
    "* `datetime` $-$ дата (2020-01-12, 23.Jan.20)\n",
    "\n",
    "\n",
    "__Классификация переменных по смыслу:__  \n",
    "\n",
    "* `Непрерывные` $-$ содержат непрерывное множество значений на каком-то промежутке (з/п, рост человека, вес буханки хлеба и тд)\n",
    "* `Категориальные` $-$ содержат конечное небольшое множество значений (возраст детей, тип магазина)\n",
    "* `Бинарные` $-$ подтип категориальной переменной, имеющей только 2 категории (пол, столицы двух стран)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "leading-flexibility",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>2маши</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )</td>\n",
       "      <td>NaN</td>\n",
       "      <td>690.0</td>\n",
       "      <td>10532.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>2221419.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3217</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       title  commentCount  dislikeCount  \\\n",
       "0                  \"#2Маши \"Свободные линии\"          34.0          43.0   \n",
       "1  \"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )           NaN         690.0   \n",
       "\n",
       "   likeCount music_style performer  viewCount  commentators_uniq  \\\n",
       "0     1402.0       popsa     2маши   214301.0               24.0   \n",
       "1    10532.0       popsa   лазарев  2221419.0                NaN   \n",
       "\n",
       "   comments_obscene_cnt  video_age  \n",
       "0                   0.0       1141  \n",
       "1                   NaN       3217  "
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('data/youtube_data_short.csv', sep='\\t')\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "august-funds",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title                    object\n",
       "commentCount            float64\n",
       "dislikeCount            float64\n",
       "likeCount               float64\n",
       "music_style              object\n",
       "performer                object\n",
       "viewCount               float64\n",
       "commentators_uniq       float64\n",
       "comments_obscene_cnt    float64\n",
       "video_age                 int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# смотрим на типы переменных \n",
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "global-camcorder",
   "metadata": {},
   "source": [
    "## Работа с категориальными переменными"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "handmade-davis",
   "metadata": {},
   "source": [
    "### Когда категории две:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "serious-flood",
   "metadata": {},
   "source": [
    "Отберем только клипы жанра `rock` и `rap`. Представим, что только они нас интересуют. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "hispanic-equality",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>\"Hey Tovarish\" Video Shooting</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>мумий_тролль</td>\n",
       "      <td>3815.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>\"SOS Матросу!\" - тизер фильма</td>\n",
       "      <td>10.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>мумий_тролль</td>\n",
       "      <td>7632.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>\"ВОИН ДУМА\"</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>49.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>король_и_шут</td>\n",
       "      <td>1950.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>\"Виктор Цой. 50 лет\" Кукрыниксы - Печаль, Конч...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>65.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>кукрыниксы</td>\n",
       "      <td>9986.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>\"Всё, что было\". Начало.</td>\n",
       "      <td>23.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>171.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>ария</td>\n",
       "      <td>11168.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2351</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                title  commentCount  \\\n",
       "8                       \"Hey Tovarish\" Video Shooting          10.0   \n",
       "15                      \"SOS Матросу!\" - тизер фильма          10.0   \n",
       "36                                        \"ВОИН ДУМА\"           4.0   \n",
       "38  \"Виктор Цой. 50 лет\" Кукрыниксы - Печаль, Конч...           2.0   \n",
       "40                           \"Всё, что было\". Начало.          23.0   \n",
       "\n",
       "    dislikeCount  likeCount music_style     performer  viewCount  \\\n",
       "8            1.0       34.0        rock  мумий_тролль     3815.0   \n",
       "15           4.0       68.0        rock  мумий_тролль     7632.0   \n",
       "36           0.0       49.0        rock  король_и_шут     1950.0   \n",
       "38           4.0       65.0        rock    кукрыниксы     9986.0   \n",
       "40           5.0      171.0        rock          ария    11168.0   \n",
       "\n",
       "    commentators_uniq  comments_obscene_cnt  video_age  \n",
       "8                 9.0                   0.0       2539  \n",
       "15                5.0                   0.0       1739  \n",
       "36                NaN                   NaN       3819  \n",
       "38                2.0                   0.0       2351  \n",
       "40               21.0                   1.0       2351  "
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genres_to_keep = ['rock', 'rap']\n",
    "\n",
    "df_small = df[df['music_style'].isin(genres_to_keep)]\n",
    "df_small.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "turned-touch",
   "metadata": {},
   "source": [
    "В таком \"поддатафрейме\" наша колонка `music_style` **имеет тип object** (т.к. в ней находятся строки), а также **является бинарной**, так как содержит всего два значения (`rock` и `rap`)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "minor-saver",
   "metadata": {},
   "source": [
    "Однако проблема строковых переменных в том, что многие модели не умеют с ними работать. Компьютеру привычнее оперировать с числами. \n",
    "\n",
    "Именно поэтому над строковыми бинарными переменными производят нехитрые манипуляции. Либо кодируют их в числа 1 / 0, либо создают булевский тип True / False, что в какой-то степени тоже самое. \n",
    "\n",
    "Попробуем преобразовать нашу бинарную строковую переменную music_style:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "selected-disposition",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-101-21b19c4d8243>:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df_small['is_rock'] = df_small['music_style'].map(d)\n"
     ]
    }
   ],
   "source": [
    "d = {'rock': 1, 'rap': 0}\n",
    "\n",
    "df_small['is_rock'] = df_small['music_style'].map(d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "engaging-render",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "      <th>is_rock</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19104</th>\n",
       "      <td>репетиция с оркестром донецк рвать</td>\n",
       "      <td>12.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>130.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>чичерина</td>\n",
       "      <td>833.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>199</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19106</th>\n",
       "      <td>саундчек-репетиция</td>\n",
       "      <td>8.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>93.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>ария</td>\n",
       "      <td>13552.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3050</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19107</th>\n",
       "      <td>снег начнется (Live)</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>земфира</td>\n",
       "      <td>5580.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>811</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19121</th>\n",
       "      <td>モスクワライブショー ' 17</td>\n",
       "      <td>4872.0</td>\n",
       "      <td>15494.0</td>\n",
       "      <td>30603.0</td>\n",
       "      <td>rap</td>\n",
       "      <td>элджей</td>\n",
       "      <td>1187340.0</td>\n",
       "      <td>1788.0</td>\n",
       "      <td>352.0</td>\n",
       "      <td>339</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19122</th>\n",
       "      <td>舞梅鬼   -  新年好宝贝 / Mumiy Troll - Happy New Year ...</td>\n",
       "      <td>59.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>260.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>мумий_тролль</td>\n",
       "      <td>10147.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>997</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   title  commentCount  \\\n",
       "19104                 репетиция с оркестром донецк рвать          12.0   \n",
       "19106                                 саундчек-репетиция           8.0   \n",
       "19107                               снег начнется (Live)           0.0   \n",
       "19121                                    モスクワライブショー ' 17        4872.0   \n",
       "19122  舞梅鬼   -  新年好宝贝 / Mumiy Troll - Happy New Year ...          59.0   \n",
       "\n",
       "       dislikeCount  likeCount music_style     performer  viewCount  \\\n",
       "19104           7.0      130.0        rock      чичерина      833.0   \n",
       "19106           3.0       93.0        rock          ария    13552.0   \n",
       "19107           4.0       51.0        rock       земфира     5580.0   \n",
       "19121       15494.0    30603.0         rap        элджей  1187340.0   \n",
       "19122          10.0      260.0        rock  мумий_тролль    10147.0   \n",
       "\n",
       "       commentators_uniq  comments_obscene_cnt  video_age  is_rock  \n",
       "19104               11.0                   0.0        199        1  \n",
       "19106                5.0                   0.0       3050        1  \n",
       "19107                NaN                   NaN        811        1  \n",
       "19121             1788.0                 352.0        339        0  \n",
       "19122               19.0                   0.0        997        1  "
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_small.tail(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "strong-numbers",
   "metadata": {},
   "source": [
    "### Задача №3: \n",
    "Какая доля наблюдений обоих жанров?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "other-ranking",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.743810005053057"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Ваш код\n",
    "df_small['is_rock'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "506582f8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.25618999494694294"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "1-df_small['is_rock'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "84e5deae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPgAAADnCAYAAAAzUZtFAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Z1A+gAAAACXBIWXMAAAsTAAALEwEAmpwYAAAbxUlEQVR4nO3deZwcZZ3H8c8zPZmZBJJASEIgKa4QIEA41hLkCEdziBYIylKggsgiK8qiKLQoNywqy7Iix67LtShhBQpkFSmEcCgQRNIlEAIBCYkJRUIucpK5Z2r/qE6YTDIz3T1d9VRX/96vV7+m0zPdz6+T+aaqq556fioIAoQQ6VSnuwAhRHQk4EKkmARciBSTgAuRYhJwIVJMAi5EiknAhUgxCbgQKSYBFyLFJOBCpJgEXIgUk4ALkWIScCFSTAIuRIpJwIVIMQm4ECkmARcixSTgQqSYBFyIFJOAC5FiEnAhUkwCLkSKScCFSDEJuBApJgEXIsUk4EKkmARciBSr112AiJZvmQ3AeGBCr9uOwHBg6BZuAC1Ac4+vzcBaYDHgAx8Ubj7woeF63fG8I1EKJc0H08O3zN2AfwAOAPYt3HYl+j21DuAdYFbh9jowy3C95RGPKwYgAa9SvmUqYApwTOF2KLCt1qI2txh4CXgGeNZwvXma66k5EvAq4lvmBODzhIE+Ghijt6KSLQSeLdyeNFxvpeZ6Uk8CnnC+ZY4DTgNOJ9xKK70VVUwH8DTwEPBbw/XWaq4nlSTgCeRb5raATRjqI0n/2Y424A+EYf+d4XotmutJDQl4gviWuR9wIfAVYJjmcnRZBfwK+G/D9f6mu5hqJwHXzLfMeuCLhMGeqrmcJAkId+FvBp4yXE9+UcsgAdfEt8xG4HzgEsLz0qJvbwM/AX4t59tLIwGPmW+ZQ4BzgcuRYJdqDnA18BvZohdHAh4T3zIzwNnAlcAuequpeq8BVxmu97juQpJOAh4D3zKPBW4D9tJdS8r8GbjAcL3XdReSVBLwCPmWuSPhQSJbdy0p1gXcDlxpuN463cUkjQQ8AoUj4xcC1xJe0CGitxj4vuF6D+kuJElqMuBKqT8BlwRB4FX6tX3LPBD4JbBfpV9bFOVp4DzD9RbqLiQJUjFDSoW0vhffMpVvmTngL0i4dToOmOVb5hm6C0mCqg24UmoXpdTflFL3AW8C9yil3lRKzVZKnd7j5y4tPDZLKXVDr9eoU0r9Uil1/WBq8S1zPOGW40agYTCvJSpiJPCAb5m/8i1za93F6FTtCz5MIjz1NJ5w0sj+wGggr5R6gfC66JOBg4MgaFZKjerx3Hrgf4E3gyD4cbkF+Jb5ReBuYNRAPyti9zXgMN8yv2K43kzdxehQtVvwgoVBEPwFOBx4IAiCriAIlgLPA58GjgXuDYKgGSAIgp6XJ97BIMLtW2bGt8ybgUeRcCfZROAl3zK/q7sQHao94OsH8dw/A0crpZpKfaJvmSMBF7hoEOOL+NQDP/ct887CTMKaUe0B3+BF4HSlVEYpNQY4AphJ+Ln4HKXUMIBeu+j3AE8AjlKq6I8qvmVOIjyQ9tlKFS9icx4w3bfM7XQXEpe0BPz/gDcI1wN7DvhBEARLgiB4EngM8JRSrxNe2LFREAQ/I5z2OK2Yo/C+ZR4DvILMSKtmRwGv+JY5WXchcajJ8+Dl8C3za4Rb/Wo/MClCa4GTDNd7QXchUUrLFjxSvmV+h3DyioQ7PUYAf/At8zjdhURJAj4A3zIvB24hPWuhiU8MA37vW+aJuguJigS8H75lXgMMahKMSLxG4FHfMk/VXUgUJOB9KIT7at11iFgMAR7yLfPLugupNDnItgW+ZV4I3Kq7DhG7TsIDb0/qLqRSJOC9+JZ5GvAgsndTq9YDRxmuV/ErDXWQgPfgW+ZRwJOEn8tE7VoGHGa43nu6CxksCXiBb5lTCGfEjdRdi0iE+cAhhust013IYEjA2dgeyCO8Kk2IDTxgquF6rboLKVfNf84sLK/kIOEWmzMJ13urWjUfcOAGpKOI6Nu5vmX+k+4iylXTu+iFyQ2P6K5DJF4r4efx13UXUqqaDbhvmXsAecI5yUIMZB5gGq63WnchpajJXXTfMhsIt9wSblGsicC9uosoVU0GnHAK6hTdRYiqc4pvmV/XXUQpam4X3bfMTxGuyCKXfopyrAGmGK7n6y6kGDW1BS/smt+LhFuUbyRwl+4iilVTAQeuQHbNxeB91rfMs3QXUYya2UX3LXN/wplJsvUWlfARMNlwveW6C+lPLW3Bb0HCLSpnO+BfdRcxkJrYghe6jzyquw6ROl3A/obrvaW7kL6kfgteWOj+Rt11iFTKADfpLqI/qQ848C/A7lEPMu/jVk54cc7G297TX+Puvy/d+P075y9lpyf+ysr2zj5fY11HFwc99wZXvvU+AG1d3Zw1cy7HvvAW9y385KrFS2cvZPaa5ujejCjFCb5lHq+7iL6kOuC+ZY4CroxjrIlbN/Hk1L15cureuIdPZmhdHSeM2waAxS3tvLBiLeOb+m88etPcxRy87SfNMJ9fsZZPj9qa6VP35tFFYVu1OWub6Q4CpowcFtl7ESW7ybfMRGYpkUVV0KXAtnEP+tKKdey0VSMThoYLw1z7ts9le41H9bPw8htr1rOirYMjxnwye3aIUrR0ddPRHbDhSMlN7y7m4j12jLB6UYYphJ1MEye1AfctcxvgWzrGfuzDlZy8Q9gGbfrS1YxramDvEX1vcbuDgOvf/oAr9pqwyeNTR4/gg+Y2Tnn5Hc7ZeSzTl65m3xHDGDfAnoDQ4lLfMhO3dn6aTxtdAAyPe9D27m6eXrqaS/ccT0tXN7e/9yH3H7RHv8+5b+Fyjh4zkh2Gbhrc+jrFbQfuBkBHd8BZ+bnc/amJXDfHZ1FrO6eO347jt98mqrciSrMXcAphn7zESGXAfcscCmjpB/2n5WvZd+QwxjQO4Z21Lfgt7ZwwYw4AH7a28/kZc3jssMmMbfyki+2rq9czc+U6pr2/nPWdXXQEAcMydfyoxxb9voXLOHX8KF5dtZ7hQzL81+TdOOOVdyXgyXIpEvBYnAuM0THw7xZ/snu+14ihvHbs/hu/d+gfZ/P4YZMZ1bDpX/utB+y68f7DH6zgjTXNm4R7dUcnzy5bw/0HTeKZZWuoQ6GA1q7uaN+MKNXBvmUebbjeH3UXskHqPoMX1li7WMfYzZ1dvLhiLSeMG/i43qzV6/nBGwuKet1b5n7IhbvvQJ1SHDF6BDNXreO4F+fwpfE10+a6mvxQdwE9pW4mm2+Zp5Cw3SRRcyYbrveO7iIghVtwoGoXyBOpcY7uAjZI1Ra8sL65T3qPLYjqsAQwDNfre9piTIoOgm3bjcBVwJeB7RzHGWnb9vHAHo7jJGXt6LORcAv9xgGfA36vu5BSdtFvBvYFvgobJ1a9habJJH1IzK6RqHmJ+F0sJeBfBL7iOM7LQDeA4ziLSEhHEN8yDwP21F2HEAUn+pap5VRtT6UEvJ1eu7+2bY8hXNkiCWzdBQjRwxDgS7qLKCXgDwO/sm17VwDbtncg7Nv0YBSFleEk3QUI0Yv238lSAn4Z8HdgNrANMBdYDFxb+bJK41vmvsCuA/6gEPE6xrdMrdf1lnWarLBrvsJxnEScY/Mt80fAT3TXIcQWnGy43mO6Bu/3lJJt27v18+3hth1+7HUcZ34liyrDFzSPL0RfTgKSGXDgPcJTYv1d5xoQrk2lhW+ZY4GDdI0vxABO9C1TGa6nZW+334A7jlMNU1mPIZ1TbkU6jCOcPzJbx+BFB8O27ZNt29a2pe7H4boLEGIAh+gauJQt33XAEtu2b7dt++CoCiqDBFwknbaAl3QU3bbt/YEzCeejrwemAfc7jrMgkuoG4FvmcGA1sosuku1dw/W0zLIs9zSZIvzs+x+Eny9eAu4AHnAcJ7ZlRnzLzALPxjWeEIMw2nC92Gd9lrzls217IuFVZb8Amgr37yJsMPBIRasbmBw9F9XiMzoGLeVy0QuAs4BJwEPAWY7j/KXH938DLOvj6VE5IObxhCiXCbhxD1rKtdOfI9wlf8xxnLbe33Qcp9m27bgn18vVY6Ja9L92dkRK2UX/k+M4D/cOt23b399w33Gc6RWrrDiTYh5PiHIlPuBX9fH4FZUopFS+ZY4HttIxthBl0LIxGnAX3bbt7IaftW37aDadtrobsC6Kwoqg5X9EIco00rfM7Q3XWzrwj1ZOMZ/B7yl8bQT+p8fjAbAUuLDSRRVJAi6qzSTCzMRmwIA7jrNhgYf7HMdJUgfF/q50EyKJdgdmxDlg0Z/Be4fbtu2jbds+ovIlFW2sxrGFKEfsa7SVcrHJ87ZtH1a4fynhUk0P2LZ9WVTFDUD69ohqE/vvbClH0fcFNkxsOQ84mnB2zvmVLqpIozSNK0S5Yg94KRNd6oCgMFVVOY4zB8C27YE77UVDtuCi2sS+USol4DMIV1HdgUJzv0LYV0RQVzEk4KLaJHoX/euEl2a+AVxdeGwv4JbKllQ02UUX1Sb2gFe0+aBt267jOFbFXrAPhR7gHVGPI0SFzTNcb/c4B6z0QglTK/x6QqRJ7EueyUooQsQn9oBXa6vd/pZxFoOwsmH48vMOuSKJi2tWPUX32ldiHrNaAy4i0pJpagmU2kl3HWkUkFkT95iV3kWXLWuVa6lv3GwxD1ExnXEPWOmAx9UfrJNCj3JRWS2ZxnbdNaRYcgNu2/ajtm1P7fXYVNu2Ny606DjOTytZXF8KbWBWxTFWrWmub5LTj9FpjXvAUrbgRwJ/7vXYy4Rz0nVYqWncVGvONErAoxPrteBQWsBb2XyJpK3RN+FEAh6Blvqm2Hcja8jiuAcsJeBPAXfYtj0CoPD1duDJKAorggQ8As2ZJjm2EZ1EB/xiYASw0rbtZYQBGwlcFEFdxZCAR6C5vlECHp3YA170eXDHcVYBlm3bOwATAN9xnCWRVTawuJss1IQW2YJHKVkBt21bOY4TFO5v2NovLdw2PhZnP7Ie5mkYM/Wa65t0l5BmyQo4sIZwtxzCc3i9Lz1Thcd0TG2cq2HM1GvJNOouIc0+jHvAgQK+T4/7u0ZZSBne1V1AGrVmGmQ2YjS6gdg/0vYbcMdx/B73F/b8nm3bQ4HuLfUpi8n7QBvheu2iQlozjRLwaLyfz2UTPZPtJtu2DyrctwiPYq+ybfukqIrrj+F63cB8HWOnWVtmiFxJFg1Px6ClnCb7KvBm4f5VwJnAF4hv/vmWyG56hbXVScAjkviADyu0CN4O2M1xnN84jvMMsHNEtRXjrxrHTqWOuiFyCXE08joGLeUf813btr9K2H7laQDbtkcDLVEUVqSXNY6dSh119UN015BCAZo2RqUE/NuEK6i2A+cWHvssEHdP8J5mEh6dlKWnKqRTZSTglfdePpeNfbEHqPCqqjr4ljmbsOuKqIDTj/jxvI66IRN115EyD+Rz2a/oGLjoLXiPPuGbcRznucqUU5a/IAGvmG7qZCpb5Wn5/A2l7aLf0+vPY4AG4AP0tvJ9GfiGxvFTpVspmVdQeckP+IY+4RvYtp0BrgDWVbqoEj2jefxUCVBDddeQMqv5pGln7Mo+OOU4ThfwY+AHlSundIbrvQ/M1llDykjAK8vVMYNtg8EefT6OZCx++HvdBaRBe119G0rJGYnK+q3OwUs5yOaz6dVkWxHOA7+g0kWV4XHgMt1FVLvWuoYWZG5/JbWhb8UjoLSDbGf2+vN64F3HcdZWsJ5yvQIsJzzwJ8rUmmmIfdXPlHs2n8t+rLOAUnbHXgeOAL4LXE44B/0R27Z1TnQBNl548oTuOqpda6ZRAl5Zv9NdQClb8IcJF3b4P/ROT+2LA5ytu4hq1lIvTQ8qqJsqC/hngNGO4yT1l2A64VJS2+supFpJV5OKeiWfy8a+DnpvpeyizwD2iqqQwTJcrxN4QHcd1Uy6mlTUr3UXAKVtwb8OPGHb9iv06tDgOM51lSxqEP4Hfcs4V73mjAS8QpqBabqLgNK24D8GDMJd4Ek9brtHUFdZDNebTXhEXZShub6pS3cNKfGgrqvHeitlC34GsIfjOLGvDFmiu4CDdRdRjSTgFfPfugvYoJSAz0dfH7JS/Br4KXJOvGQtmei6mrSvXsbfH7yBjo9XgVKMOdhi+8NPZdH0X7Fipkv9VtsAMP6Ec9lm8ub/P3e2fMzCR26iZckCUIpdTruErXfehw+euJM178xk2I67s+sZPwTgo1efpnP9WrafempUb6c/+Xwuq+3ikt5KCfg04DHbtm9j88/gOi8X3YThei2+Zd4GJOW4QNVoyTRGtzhAXYYJJ57PVhP2oKu1mTm3ns+ISZ8CYPup/8i4I+1+n+4/djsj9vg0E8+6hu7ODro72uhs+Zj1i+ayz/fvZsHDN9H84XyaRo9nhfcUk869IbK3MoCf6Rp4S0oJ+IYpqb0XWQzQe7nolvwn4UUwW+supJpE2dWkYcR2NIzYDoBM0zCGjt2Z9jUrinpuZ8vHrJs/m13sSwGoqx9CXf0QulqbCbq6CIKA7o42VKaeJc87jD30FOoyWpaWex94RMfAfSn7ctEkM1xvpW+ZdwHf011LNWmNqatJ28olNC9+j613mszHC95i2Z9/y0d/nc6wCXtinHg+9cOGb/Lz7auWUL/1SBY4N9L84Xy2Gj8J4+QLyDQNY+ReBzHn599kxO4HkmnaivXvv82Ox54Vy/vYgtt0Xjm2JWm+cuhnVMcxg8RoiaGrSVdbC/OmXYNx0rfJNG3F2ENOYsql09j7ojsZMmIU/uObH58KurpoXjSXMYd8gX0uuoO6hiaW/PFBAHY46gz2+d6dGCd9i8VP3cuOx3+d5a+4zLv/OhY/e3/Ub6en5cAdcQ5YjNQG3HC9D4BY/4WrXVumIdLfh+6uTuZNu4ZRBx7DtlOmAjBk+ChUXQZVV8eYgyzW++9s9ryGbcbQMHIMW+80GYBt9zuC5kWbtqYL/xzQNNZg1ewXmHjmVbR9tJjW5R9E+ZZ6ujafy+pe/GQzqQ14wTWAXEBRpLa66AIeBAELH76JprE7Me6I0zY+3r72o433V705g6HjdtnsuUOGj6Jh5Bhal4WdtNbOfY2msZsux7/oqXvZ8fhzws/k3eHJAKUU3R2xdNZ6lwRuvaG0g2xVx3C9933LvBn4ke5aqkF7JrqmBx8veJOPXn2aoeN25a2b/xkIT4mtnPUcLYvDTtAN245j51PDwybta1aw4JH/YI9zfwrATqdcyPwHfkLQ1UHjdjuwy2mfLCS06s0ZDJuwJw0jRwMwbMeJvPWzbzB03G4M2zGWBWJ/mLTP3htU/bLJA/EtczjwHjBWdy1Jd86hV766pmH4P+iuo8rMyOeyU3UX0Ze076JjuN464GrddVSDTiVdTcqQ011Af1If8IK7gLd0F5F0nXWZBt01VJmH87msthVTi1ETATdcr4twJRrRjy5VJwEvXjtVcGynJgIOYLjes8DduutIsm4lXU1KcHU+l52nu4iB1EzACy4m7MQitiBAScCL8xJwo+4iilFTATdcby3wTd11JFUgTQ+K8THwtXwum4R+AAOqqYADGK73BHCf7jqSJoAAJVvwIlycz2Xn6y6iWDUX8IKLgEW6i0iStrDpgeifm89l79RdRClqMuCG660CTgcSOftIh9aMBHwAH1GFXWxrMuAAhuu9hObGiUkiXU0GdH4+l12iu4hS1WzAAQzXu5mEXaCvS0umMZarMqrUz/O5bFX+ntR0wAv+ifBqoJomXU369Djh6dWqVPMBL8xVPxVI3LW8cWrJNEnANzcL+HK1nBLbkpoPOIDhem8ShrxmV4Bprm+s2ffehw+BE3V3Bx0sCXiB4XpPA+eyaQ/0mtGcaZIzCp9oBr6Qz2WrftajBLwHw/WmAZfprkMHaXqwUUA4U83TXUglSMB7MVzvBuB23XXETQK+0Q/zuexvdBdRKRLwLfsu8L+6i4hTc6apJj+a9HJVPpetiotIiiUB3wLD9bqBrxF2K60JLfURdjWpDlfmc9l/1V1EpUnA+1AI+TeAX+iuJQ6Rti1Kvsvzuez1uouIQqpXVR0sw/UC4Nu+ZbaR8r7jrZnGyJseJFAAXJTPZW/VXUhUZAteBMP1vsfmPdlSpTWGriYJ0wmcneZwgwS8aIbrXQ6cR0onw7RmGjK6a4hRC3BqPpedpruQqEnAS2C43t3AcYSXDqZKe92QWvldmAd8Jp/LPlbqE1Woqv6eqqrYJDBc73ngYOBt3bVUUntddF1NEsQFzHwu+0axT1BK7aKU+ptS6j7gTeAepZSnlHpLKXVtj59boJS6USk1Wyk1Uym1ewT1l0wCXgbD9eYBhwBP6a6lUjrqUt30ICDsU3dSPpddXcbzJwH/FQTBPsDFQRCYwH7AkUqp/Xr83JogCKYQTpT6+aAqrhAJeJkM11sDfB64Eqj6WWCd6Q34KsKLRq7N57LlngpcGATBhgYHtlLqVeA1YB9g7x4/90CPr4eUOVZFScAHwXC9bsP1rgeOAt7XXM6gdKaz6cEswl3yJwb5OusBlFK7ApcAxwRBsB/hLn/PhSqDPu5rIwGvAMP1ZhDuslXt9NYuVdeou4YK6gJuAg6p8AqoIwjDvkYptT3wuV7fP73H15crOG7ZauHASiwKu+xn+pb5e+BWqqybaXd6Av5X4Lx8LvtapV84CIJZSqnXgHcAn7ABQk/bKqXeANqAL1d6/HKkvn2wDr5lbgv8lPC8eVXsJX3pyH9bi1IjdNcxCOsJj4fcms9lYz8mopRaAJhBEKyIe+z+yBY8AoVlmc/3LfNewrnsB2ouqRjV3NXkD8C38rnsQt2FJI1swSPmW2YG+BfgOsLPcInTpeo6Tzvyhmr8z34J8L18Lvug7kKSqip2H6uZ4XpdhuvdAuxG2LCuWXNJm2mtG1JtTQ+WEh7Nnijh7p9swWPmW+b2hMtCfRNIxIGtlQ0jln/j0CvG6K6jCEuBfwd+kc9lE/cfZRJJwDXxLXMC4UGhs9Ec9EVDR/sXHvwDQ2cNA5Bgl6kaP3elguF6HwDf9C3zCsKt+fnAeB21tCa3q8ki4GYk2GWTgGtmuN5y4HrfMm8AvgR8Bzgszhqak9XVpBt4ErgTeFzHKa80kYAnhOF6nYADOL5lHgCcBdjAhKjHbsk0JeEa97nA/cAv87lsVU/7TRIJeAIZrvc68LpvmZcAhwNnAKcBkRwIa67XFvBlwMPAtHwu+4qmGlJNAp5ghTXhXgRe9C3zO8DRwAnAsYRz3yuyzFKMXU1agBnA04XbrEFc4SWKIAGvEobrdQHPFG74ljkGyALHFG67lfvaETY9CAgvq9wQ6Jfyuaz0IY+RBLxKFQ7OPVS44VvmaOCAXrc9KeLfuLm+sRLdM5cDc3rdZuVz2dQtb1VNJOApYbjeCnps4QF8y2wi3LIbW7jtQDh1dng/B9m6gDXA6l63NcBK4D0KYc7nsom6yEKEZKKLAODT//5cHdAADCl8bav21rlCAi5EqsnFJkKkmARciBSTgAuRYhJwIVJMAi5EiknAhUgxCbgQKSYBFyLFJOBCpJgEXIgUk4ALkWIScCFSTAIuRIpJwIVIMQm4ECkmARcixSTgQqSYBFyIFJOAC5FiEnAhUkwCLkSKScCFSDEJuBApJgEXIsUk4EKkmARciBSTgAuRYv8PRZTPBTJJH2EAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_small['is_rock'].value_counts(normalize=True).plot(kind='pie', autopct='%1.1f%%'); plt.show()\n",
    "\n",
    "df_small['music_style'].value_counts(normalize=True).plot(kind='pie', autopct='%1.1f%%'); plt.show()\n",
    "\n",
    "df_small['music_style'].value_counts(normalize=True).plot(kind='bar'); plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "urban-conditioning",
   "metadata": {},
   "source": [
    "### Когда категорий больше двух"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "integral-victor",
   "metadata": {},
   "source": [
    "Вернемся к нашему полному датафрейму. В нем music_style принимает не 2 значения, а целых 7. Что делать в таком случае?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "crude-photography",
   "metadata": {},
   "source": [
    "**Наивный метод:** Давайте каждому значению из music_style сопоставим какое-нибудь число. Такой метод называется Label Encoding. Заранее подумайте, какие проблемы может создать такой способ?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "9548cf01",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>2маши</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )</td>\n",
       "      <td>NaN</td>\n",
       "      <td>690.0</td>\n",
       "      <td>10532.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>2221419.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3217</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       title  commentCount  dislikeCount  \\\n",
       "0                  \"#2Маши \"Свободные линии\"          34.0          43.0   \n",
       "1  \"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )           NaN         690.0   \n",
       "\n",
       "   likeCount music_style performer  viewCount  commentators_uniq  \\\n",
       "0     1402.0       popsa     2маши   214301.0               24.0   \n",
       "1    10532.0       popsa   лазарев  2221419.0                NaN   \n",
       "\n",
       "   comments_obscene_cnt  video_age  \n",
       "0                   0.0       1141  \n",
       "1                   NaN       3217  "
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "id": "advisory-channel",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['popsa', 'rock', 'classic', 'indi', 'estrada', 'shanson', 'rap'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 272,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.music_style.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "id": "spectacular-northeast",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'popsa': 0, 'rock': 1, 'classic': 2, 'indi': 3, 'estrada': 4, 'shanson': 5, 'rap': 6}\n"
     ]
    }
   ],
   "source": [
    "# создадим словарь\n",
    "d = dict(zip(df.music_style.unique(), np.arange(7)))\n",
    "print(d)\n",
    "\n",
    "# сделаем преобразование\n",
    "df['music_style_int'] = df['music_style'].map(d)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "authentic-junior",
   "metadata": {},
   "source": [
    "Однако проблема такого способа в том, что мы непроизвольным образом установили **порядок** на наших категориях:\n",
    "$6 > 5 > 4 > 3 > 2 > 1 > 0 \\Rightarrow rap > shanson > estrada > indi > classic > rock > popsa$ \n",
    "\n",
    "\n",
    "Чтобы такого избежать, чаще используется другой метод: **One Hot Encoding**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "pursuant-twist",
   "metadata": {},
   "source": [
    "<center>\n",
    "<img src=\"https://4.bp.blogspot.com/-iC7OQxsaPY0/XMdB9Q9QahI/AAAAAAAAGoY/qEStugMphDIbUIabd_taWCzTyscXzxOPgCLcBGAs/s1600/human_transpose.JPG\" height=\"200\" width=\"600\">\n",
    "</center>   "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "apart-hotel",
   "metadata": {},
   "source": [
    "В pandas такое преобразование можно сделать с помощью функции `get_dummies()`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "e55bfab8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>music_style</th>\n",
       "      <th>classic</th>\n",
       "      <th>estrada</th>\n",
       "      <th>indi</th>\n",
       "      <th>popsa</th>\n",
       "      <th>rap</th>\n",
       "      <th>rock</th>\n",
       "      <th>shanson</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>popsa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>popsa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>popsa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>popsa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>popsa</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19119</th>\n",
       "      <td>shanson</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19120</th>\n",
       "      <td>classic</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19121</th>\n",
       "      <td>rap</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19122</th>\n",
       "      <td>rock</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19123</th>\n",
       "      <td>shanson</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19124 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      music_style  classic  estrada  indi  popsa  rap  rock  shanson\n",
       "0           popsa        0        0     0      1    0     0        0\n",
       "1           popsa        0        0     0      1    0     0        0\n",
       "2           popsa        0        0     0      1    0     0        0\n",
       "3           popsa        0        0     0      1    0     0        0\n",
       "4           popsa        0        0     0      1    0     0        0\n",
       "...           ...      ...      ...   ...    ...  ...   ...      ...\n",
       "19119     shanson        0        0     0      0    0     0        1\n",
       "19120     classic        1        0     0      0    0     0        0\n",
       "19121         rap        0        0     0      0    1     0        0\n",
       "19122        rock        0        0     0      0    0     1        0\n",
       "19123     shanson        0        0     0      0    0     0        1\n",
       "\n",
       "[19124 rows x 8 columns]"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([df['music_style'], pd.get_dummies(df['music_style'])], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "intense-heritage",
   "metadata": {},
   "source": [
    "Такой метод тоже имеет свои проблемы, которые выходят за рамки нашего курса. Например, это полная мультиколлинеарность (или линейная зависимость) столбцов. Подробнее об этом можно посмотреть здесь: https://www.youtube.com/watch?v=k9PPuTmTLMk"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "continued-endorsement",
   "metadata": {},
   "source": [
    "## Работа с датой и временем"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "moved-bahrain",
   "metadata": {},
   "source": [
    "Для даты и времени в python есть свой тип переменных: datetime / timestamp. Все полезные методы, оперирующие над датами можно найти либо в пакете pandas, либо в дополнительном пакете `datetime`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "variable-clark",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2020-01-01 00:00:00')"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# получим дату из строки\n",
    "a = pd.to_datetime('01/01/20')\n",
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "deb1facb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Wednesday'"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a.day_name()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "grand-liberal",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "datetime.datetime(2021, 4, 8, 20, 33, 54, 74690)"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# можно получить сегодняшнее число\n",
    "from datetime import datetime\n",
    "\n",
    "datetime.today()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "id": "hispanic-indianapolis",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2021"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# какой сегодня год?\n",
    "datetime.today().year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "id": "nasty-hands",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'08/04/2021'"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# преобразуем в строку удобного нам формата\n",
    "datetime.today().strftime('%d/%m/%Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "id": "correct-spirit",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "datetime.datetime(2021, 2, 17, 20, 34, 24, 968917)"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from datetime import timedelta\n",
    "\n",
    "# какой месяц был 50 дней назад?\n",
    "(datetime.today() - timedelta(days=50))#.month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "id": "673f275d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(datetime.today() - timedelta(days=50)).month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "id": "0415ec37",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4320000.0"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "timedelta(days=50).total_seconds()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "accomplished-conditioning",
   "metadata": {},
   "source": [
    "### Задача №4: \n",
    "\n",
    "В нашем датасете есть колонка `video_age`, которая показывает нам возраст нашего клипа в днях с 15 марта 2019 года. Опираясь на нее, получите для каждого клипа дату его публикации и сохраните в новую колонку. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "df5cfc42",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2019-03-15 00:00:00')"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_datetime('2019-03-15')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "id": "4935fcdf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "datetime.timedelta(days=3)"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "timedelta(days=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "id": "7c9e3fe7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       2016-01-29\n",
       "1       2010-05-24\n",
       "2       2009-06-29\n",
       "3       2010-06-21\n",
       "4       2009-12-31\n",
       "           ...    \n",
       "19119   2019-01-30\n",
       "19120   2018-07-19\n",
       "19121   2018-04-10\n",
       "19122   2016-06-21\n",
       "19123   2019-01-09\n",
       "Name: video_age, Length: 19124, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datetime(2019,3,15) - df['video_age'].apply(lambda x: timedelta(days=x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "ethical-paste",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0   2016-01-29\n",
       "1   2010-05-24\n",
       "2   2009-06-29\n",
       "3   2010-06-21\n",
       "4   2009-12-31\n",
       "Name: date, dtype: datetime64[ns]"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Ваш код\n",
    "df['date'] = pd.to_datetime('2019-03-15') - df['video_age'].apply(lambda x: timedelta(days=x))\n",
    "df['date'].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "trying-courage",
   "metadata": {},
   "source": [
    "### Задача №5:\n",
    "\n",
    "Какое распределение кол-ва выпущенных клипов по годам? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "id": "fundamental-conference",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>2маши</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "      <td>2016-01-29</td>\n",
       "      <td>2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )</td>\n",
       "      <td>NaN</td>\n",
       "      <td>690.0</td>\n",
       "      <td>10532.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>2221419.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3217</td>\n",
       "      <td>2010-05-24</td>\n",
       "      <td>2010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       title  commentCount  dislikeCount  \\\n",
       "0                  \"#2Маши \"Свободные линии\"          34.0          43.0   \n",
       "1  \"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )           NaN         690.0   \n",
       "\n",
       "   likeCount music_style performer  viewCount  commentators_uniq  \\\n",
       "0     1402.0       popsa     2маши   214301.0               24.0   \n",
       "1    10532.0       popsa   лазарев  2221419.0                NaN   \n",
       "\n",
       "   comments_obscene_cnt  video_age       date  year  \n",
       "0                   0.0       1141 2016-01-29  2016  \n",
       "1                   NaN       3217 2010-05-24  2010  "
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Ваш код\n",
    "df['year'] = df['date'].dt.year\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "id": "be74dff7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year\n",
       "2006       3\n",
       "2007      63\n",
       "2008     133\n",
       "2009     448\n",
       "2010    1223\n",
       "2011    1564\n",
       "2012    1176\n",
       "2013    1250\n",
       "2014    1980\n",
       "2015    1857\n",
       "2016    1885\n",
       "2017    2153\n",
       "2018    4968\n",
       "2019     421\n",
       "Name: year, dtype: int64"
      ]
     },
     "execution_count": 180,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby('year')['year'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "id": "4bf3223a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['year'].value_counts().sort_index().plot(kind='bar')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "numerical-teacher",
   "metadata": {},
   "source": [
    "### Задача №6:\n",
    "\n",
    "Выведите месяц, в течение которого на ютуб опубликовали больше всего клипов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "id": "virtual-passage",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Ваш код\n",
    "\n",
    "df['month'] = df['date'].dt.month_name()\n",
    "\n",
    "df['month'].value_counts().sort_values().plot(kind='barh')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "id": "e61e0dde",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df['year-month'] = df['date'].dt.year.astype(str) +'-'+ df['date'].dt.month.astype(str) + '-1'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "6331d96e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>year-month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>2маши</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "      <td>2016-01-29</td>\n",
       "      <td>2016</td>\n",
       "      <td>January</td>\n",
       "      <td>2016-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )</td>\n",
       "      <td>NaN</td>\n",
       "      <td>690.0</td>\n",
       "      <td>10532.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>2221419.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3217</td>\n",
       "      <td>2010-05-24</td>\n",
       "      <td>2010</td>\n",
       "      <td>May</td>\n",
       "      <td>2010-05-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>\"EARTH SONG\"  SERGEY LAZAREV. DEDICATED TO MIC...</td>\n",
       "      <td>26.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>269.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>40327.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3546</td>\n",
       "      <td>2009-06-29</td>\n",
       "      <td>2009</td>\n",
       "      <td>June</td>\n",
       "      <td>2009-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>\"Earth song\" премия МУЗ-тВ 2010 (LIVE)</td>\n",
       "      <td>12.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>163.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>11907.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3189</td>\n",
       "      <td>2010-06-21</td>\n",
       "      <td>2010</td>\n",
       "      <td>June</td>\n",
       "      <td>2010-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>\"FAKE\" СЕРГЕЙ ЛАЗАРЕВ (SERGEY)</td>\n",
       "      <td>46.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>597.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>91483.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3361</td>\n",
       "      <td>2009-12-31</td>\n",
       "      <td>2009</td>\n",
       "      <td>December</td>\n",
       "      <td>2009-12-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               title  commentCount  \\\n",
       "0                          \"#2Маши \"Свободные линии\"          34.0   \n",
       "1          \"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )           NaN   \n",
       "2  \"EARTH SONG\"  SERGEY LAZAREV. DEDICATED TO MIC...          26.0   \n",
       "3             \"Earth song\" премия МУЗ-тВ 2010 (LIVE)          12.0   \n",
       "4                     \"FAKE\" СЕРГЕЙ ЛАЗАРЕВ (SERGEY)          46.0   \n",
       "\n",
       "   dislikeCount  likeCount music_style performer  viewCount  \\\n",
       "0          43.0     1402.0       popsa     2маши   214301.0   \n",
       "1         690.0    10532.0       popsa   лазарев  2221419.0   \n",
       "2          10.0      269.0       popsa   лазарев    40327.0   \n",
       "3           7.0      163.0       popsa   лазарев    11907.0   \n",
       "4          19.0      597.0       popsa   лазарев    91483.0   \n",
       "\n",
       "   commentators_uniq  comments_obscene_cnt  video_age       date  year  \\\n",
       "0               24.0                   0.0       1141 2016-01-29  2016   \n",
       "1                NaN                   NaN       3217 2010-05-24  2010   \n",
       "2               23.0                   0.0       3546 2009-06-29  2009   \n",
       "3               11.0                   0.0       3189 2010-06-21  2010   \n",
       "4               33.0                   1.0       3361 2009-12-31  2009   \n",
       "\n",
       "      month year-month  \n",
       "0   January 2016-01-01  \n",
       "1       May 2010-05-01  \n",
       "2      June 2009-06-01  \n",
       "3      June 2010-06-01  \n",
       "4  December 2009-12-01  "
      ]
     },
     "execution_count": 217,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['year-month'] = pd.to_datetime(df['year-month'])\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "id": "9a905ab0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['year-month'].value_counts().sort_index().plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "completed-restoration",
   "metadata": {},
   "source": [
    "## Пропущенные значения"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "documented-investigator",
   "metadata": {},
   "source": [
    "Пропуски в данных $-$ это противная проблема, с которой можно столкнуться при работе. Пропуски возникают практически всегда. Причинами пропусков может быть, например, технические возможности вашего парсера, который не смог достать часть информации из сайта, либо, например, если вы Юрий Дудь и опрашиваете людей об их заработке, собирая статистику для дальнейшего исследования, то вряд ли каждый человек захочет вам ответить. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "certified-discretion",
   "metadata": {},
   "source": [
    "### Выявление"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "id": "sitting-commitment",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 19124 entries, 0 to 19123\n",
      "Data columns (total 14 columns):\n",
      " #   Column                Non-Null Count  Dtype         \n",
      "---  ------                --------------  -----         \n",
      " 0   title                 19124 non-null  object        \n",
      " 1   commentCount          18914 non-null  float64       \n",
      " 2   dislikeCount          18850 non-null  float64       \n",
      " 3   likeCount             18850 non-null  float64       \n",
      " 4   music_style           19124 non-null  object        \n",
      " 5   performer             19124 non-null  object        \n",
      " 6   viewCount             19123 non-null  float64       \n",
      " 7   commentators_uniq     12314 non-null  float64       \n",
      " 8   comments_obscene_cnt  12314 non-null  float64       \n",
      " 9   video_age             19124 non-null  int64         \n",
      " 10  date                  19124 non-null  datetime64[ns]\n",
      " 11  year                  19124 non-null  int64         \n",
      " 12  month                 19124 non-null  object        \n",
      " 13  year-month            19124 non-null  datetime64[ns]\n",
      "dtypes: datetime64[ns](2), float64(6), int64(2), object(4)\n",
      "memory usage: 2.0+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "arbitrary-medication",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title                      0\n",
       "music_style                0\n",
       "performer                  0\n",
       "video_age                  0\n",
       "date                       0\n",
       "year                       0\n",
       "month                      0\n",
       "year-month                 0\n",
       "viewCount                  1\n",
       "commentCount             210\n",
       "dislikeCount             274\n",
       "likeCount                274\n",
       "commentators_uniq       6810\n",
       "comments_obscene_cnt    6810\n",
       "dtype: int64"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum().sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "id": "perceived-vermont",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(20,12))\n",
    "sns_heatmap = sns.heatmap(df.isnull(), yticklabels=False, cbar=False, cmap='viridis')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "wireless-minister",
   "metadata": {},
   "source": [
    "### Удаление"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "id": "0003d7c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna().isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "id": "d6302088",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>commentCount</th>\n",
       "      <th>dislikeCount</th>\n",
       "      <th>likeCount</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>viewCount</th>\n",
       "      <th>commentators_uniq</th>\n",
       "      <th>comments_obscene_cnt</th>\n",
       "      <th>video_age</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>year-month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>34.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>1402.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>2маши</td>\n",
       "      <td>214301.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1141</td>\n",
       "      <td>2016-01-29</td>\n",
       "      <td>2016</td>\n",
       "      <td>January</td>\n",
       "      <td>2016-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>\"EARTH SONG\"  SERGEY LAZAREV. DEDICATED TO MIC...</td>\n",
       "      <td>26.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>269.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>40327.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3546</td>\n",
       "      <td>2009-06-29</td>\n",
       "      <td>2009</td>\n",
       "      <td>June</td>\n",
       "      <td>2009-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>\"Earth song\" премия МУЗ-тВ 2010 (LIVE)</td>\n",
       "      <td>12.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>163.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>11907.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3189</td>\n",
       "      <td>2010-06-21</td>\n",
       "      <td>2010</td>\n",
       "      <td>June</td>\n",
       "      <td>2010-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>\"FAKE\" СЕРГЕЙ ЛАЗАРЕВ (SERGEY)</td>\n",
       "      <td>46.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>597.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>91483.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3361</td>\n",
       "      <td>2009-12-31</td>\n",
       "      <td>2009</td>\n",
       "      <td>December</td>\n",
       "      <td>2009-12-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>\"Feelin High\" Сергей Лазарев Sergey Lazarev live</td>\n",
       "      <td>7.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>9217.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3179</td>\n",
       "      <td>2010-07-01</td>\n",
       "      <td>2010</td>\n",
       "      <td>July</td>\n",
       "      <td>2010-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19116</th>\n",
       "      <td>♫ Музыка Моцарта для детей (Mozart's music for...</td>\n",
       "      <td>41.0</td>\n",
       "      <td>142.0</td>\n",
       "      <td>914.0</td>\n",
       "      <td>classic</td>\n",
       "      <td>classical_channel</td>\n",
       "      <td>308721.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1748</td>\n",
       "      <td>2014-06-01</td>\n",
       "      <td>2014</td>\n",
       "      <td>June</td>\n",
       "      <td>2014-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19117</th>\n",
       "      <td>♫♡ Сборник Лучших Мелодий - Наслаждайтесь Восх...</td>\n",
       "      <td>2881.0</td>\n",
       "      <td>4108.0</td>\n",
       "      <td>58429.0</td>\n",
       "      <td>classic</td>\n",
       "      <td>classical_channel</td>\n",
       "      <td>9577433.0</td>\n",
       "      <td>1696.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>538</td>\n",
       "      <td>2017-09-23</td>\n",
       "      <td>2017</td>\n",
       "      <td>September</td>\n",
       "      <td>2017-09-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19120</th>\n",
       "      <td>❤ 2 ЧАСА ❤ МОЦАРТ Для Младенцев - Колыбельная ...</td>\n",
       "      <td>94.0</td>\n",
       "      <td>431.0</td>\n",
       "      <td>1837.0</td>\n",
       "      <td>classic</td>\n",
       "      <td>classical_channel</td>\n",
       "      <td>391836.0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>239</td>\n",
       "      <td>2018-07-19</td>\n",
       "      <td>2018</td>\n",
       "      <td>July</td>\n",
       "      <td>2018-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19121</th>\n",
       "      <td>モスクワライブショー ' 17</td>\n",
       "      <td>4872.0</td>\n",
       "      <td>15494.0</td>\n",
       "      <td>30603.0</td>\n",
       "      <td>rap</td>\n",
       "      <td>элджей</td>\n",
       "      <td>1187340.0</td>\n",
       "      <td>1788.0</td>\n",
       "      <td>352.0</td>\n",
       "      <td>339</td>\n",
       "      <td>2018-04-10</td>\n",
       "      <td>2018</td>\n",
       "      <td>April</td>\n",
       "      <td>2018-04-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19122</th>\n",
       "      <td>舞梅鬼   -  新年好宝贝 / Mumiy Troll - Happy New Year ...</td>\n",
       "      <td>59.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>260.0</td>\n",
       "      <td>rock</td>\n",
       "      <td>мумий_тролль</td>\n",
       "      <td>10147.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>997</td>\n",
       "      <td>2016-06-21</td>\n",
       "      <td>2016</td>\n",
       "      <td>June</td>\n",
       "      <td>2016-06-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12116 rows × 14 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   title  commentCount  \\\n",
       "0                              \"#2Маши \"Свободные линии\"          34.0   \n",
       "2      \"EARTH SONG\"  SERGEY LAZAREV. DEDICATED TO MIC...          26.0   \n",
       "3                 \"Earth song\" премия МУЗ-тВ 2010 (LIVE)          12.0   \n",
       "4                         \"FAKE\" СЕРГЕЙ ЛАЗАРЕВ (SERGEY)          46.0   \n",
       "5       \"Feelin High\" Сергей Лазарев Sergey Lazarev live           7.0   \n",
       "...                                                  ...           ...   \n",
       "19116  ♫ Музыка Моцарта для детей (Mozart's music for...          41.0   \n",
       "19117  ♫♡ Сборник Лучших Мелодий - Наслаждайтесь Восх...        2881.0   \n",
       "19120  ❤ 2 ЧАСА ❤ МОЦАРТ Для Младенцев - Колыбельная ...          94.0   \n",
       "19121                                    モスクワライブショー ' 17        4872.0   \n",
       "19122  舞梅鬼   -  新年好宝贝 / Mumiy Troll - Happy New Year ...          59.0   \n",
       "\n",
       "       dislikeCount  likeCount music_style          performer  viewCount  \\\n",
       "0              43.0     1402.0       popsa              2маши   214301.0   \n",
       "2              10.0      269.0       popsa            лазарев    40327.0   \n",
       "3               7.0      163.0       popsa            лазарев    11907.0   \n",
       "4              19.0      597.0       popsa            лазарев    91483.0   \n",
       "5               7.0       87.0       popsa            лазарев     9217.0   \n",
       "...             ...        ...         ...                ...        ...   \n",
       "19116         142.0      914.0     classic  classical_channel   308721.0   \n",
       "19117        4108.0    58429.0     classic  classical_channel  9577433.0   \n",
       "19120         431.0     1837.0     classic  classical_channel   391836.0   \n",
       "19121       15494.0    30603.0         rap             элджей  1187340.0   \n",
       "19122          10.0      260.0        rock       мумий_тролль    10147.0   \n",
       "\n",
       "       commentators_uniq  comments_obscene_cnt  video_age       date  year  \\\n",
       "0                   24.0                   0.0       1141 2016-01-29  2016   \n",
       "2                   23.0                   0.0       3546 2009-06-29  2009   \n",
       "3                   11.0                   0.0       3189 2010-06-21  2010   \n",
       "4                   33.0                   1.0       3361 2009-12-31  2009   \n",
       "5                    6.0                   0.0       3179 2010-07-01  2010   \n",
       "...                  ...                   ...        ...        ...   ...   \n",
       "19116               22.0                   0.0       1748 2014-06-01  2014   \n",
       "19117             1696.0                   5.0        538 2017-09-23  2017   \n",
       "19120               37.0                   1.0        239 2018-07-19  2018   \n",
       "19121             1788.0                 352.0        339 2018-04-10  2018   \n",
       "19122               19.0                   0.0        997 2016-06-21  2016   \n",
       "\n",
       "           month year-month  \n",
       "0        January 2016-01-01  \n",
       "2           June 2009-06-01  \n",
       "3           June 2010-06-01  \n",
       "4       December 2009-12-01  \n",
       "5           July 2010-07-01  \n",
       "...          ...        ...  \n",
       "19116       June 2014-06-01  \n",
       "19117  September 2017-09-01  \n",
       "19120       July 2018-07-01  \n",
       "19121      April 2018-04-01  \n",
       "19122       June 2016-06-01  \n",
       "\n",
       "[12116 rows x 14 columns]"
      ]
     },
     "execution_count": 239,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "id": "inside-queensland",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(19124, 14)\n",
      "(12116, 14)\n",
      "(19124, 8)\n"
     ]
    }
   ],
   "source": [
    "print(df.shape)\n",
    "\n",
    "# по умолчанию удаляет строки, где есть na\n",
    "print(df.dropna().shape)\n",
    "\n",
    "# можем удалять столбцы, где есть хотя бы один пропуск\n",
    "print(df.dropna(axis=1).shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "id": "1c7b366d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>music_style</th>\n",
       "      <th>performer</th>\n",
       "      <th>video_age</th>\n",
       "      <th>date</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>year-month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>\"#2Маши \"Свободные линии\"</td>\n",
       "      <td>popsa</td>\n",
       "      <td>2маши</td>\n",
       "      <td>1141</td>\n",
       "      <td>2016-01-29</td>\n",
       "      <td>2016</td>\n",
       "      <td>January</td>\n",
       "      <td>2016-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>\"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>3217</td>\n",
       "      <td>2010-05-24</td>\n",
       "      <td>2010</td>\n",
       "      <td>May</td>\n",
       "      <td>2010-05-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>\"EARTH SONG\"  SERGEY LAZAREV. DEDICATED TO MIC...</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>3546</td>\n",
       "      <td>2009-06-29</td>\n",
       "      <td>2009</td>\n",
       "      <td>June</td>\n",
       "      <td>2009-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>\"Earth song\" премия МУЗ-тВ 2010 (LIVE)</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>3189</td>\n",
       "      <td>2010-06-21</td>\n",
       "      <td>2010</td>\n",
       "      <td>June</td>\n",
       "      <td>2010-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>\"FAKE\" СЕРГЕЙ ЛАЗАРЕВ (SERGEY)</td>\n",
       "      <td>popsa</td>\n",
       "      <td>лазарев</td>\n",
       "      <td>3361</td>\n",
       "      <td>2009-12-31</td>\n",
       "      <td>2009</td>\n",
       "      <td>December</td>\n",
       "      <td>2009-12-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19119</th>\n",
       "      <td>♫♪Мы расстаемся ♫♪  Бутырка new2019</td>\n",
       "      <td>shanson</td>\n",
       "      <td>бутырка</td>\n",
       "      <td>44</td>\n",
       "      <td>2019-01-30</td>\n",
       "      <td>2019</td>\n",
       "      <td>January</td>\n",
       "      <td>2019-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19120</th>\n",
       "      <td>❤ 2 ЧАСА ❤ МОЦАРТ Для Младенцев - Колыбельная ...</td>\n",
       "      <td>classic</td>\n",
       "      <td>classical_channel</td>\n",
       "      <td>239</td>\n",
       "      <td>2018-07-19</td>\n",
       "      <td>2018</td>\n",
       "      <td>July</td>\n",
       "      <td>2018-07-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19121</th>\n",
       "      <td>モスクワライブショー ' 17</td>\n",
       "      <td>rap</td>\n",
       "      <td>элджей</td>\n",
       "      <td>339</td>\n",
       "      <td>2018-04-10</td>\n",
       "      <td>2018</td>\n",
       "      <td>April</td>\n",
       "      <td>2018-04-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19122</th>\n",
       "      <td>舞梅鬼   -  新年好宝贝 / Mumiy Troll - Happy New Year ...</td>\n",
       "      <td>rock</td>\n",
       "      <td>мумий_тролль</td>\n",
       "      <td>997</td>\n",
       "      <td>2016-06-21</td>\n",
       "      <td>2016</td>\n",
       "      <td>June</td>\n",
       "      <td>2016-06-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19123</th>\n",
       "      <td>🎄#премьера песни!!! 2019г(вы не спешите говори...</td>\n",
       "      <td>shanson</td>\n",
       "      <td>бутырка</td>\n",
       "      <td>65</td>\n",
       "      <td>2019-01-09</td>\n",
       "      <td>2019</td>\n",
       "      <td>January</td>\n",
       "      <td>2019-01-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19124 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   title music_style  \\\n",
       "0                              \"#2Маши \"Свободные линии\"       popsa   \n",
       "1              \"ALARM\" СЕРГЕЙ ЛАЗАРЕВ ( SERGEY LAZAREV )       popsa   \n",
       "2      \"EARTH SONG\"  SERGEY LAZAREV. DEDICATED TO MIC...       popsa   \n",
       "3                 \"Earth song\" премия МУЗ-тВ 2010 (LIVE)       popsa   \n",
       "4                         \"FAKE\" СЕРГЕЙ ЛАЗАРЕВ (SERGEY)       popsa   \n",
       "...                                                  ...         ...   \n",
       "19119                ♫♪Мы расстаемся ♫♪  Бутырка new2019     shanson   \n",
       "19120  ❤ 2 ЧАСА ❤ МОЦАРТ Для Младенцев - Колыбельная ...     classic   \n",
       "19121                                    モスクワライブショー ' 17         rap   \n",
       "19122  舞梅鬼   -  新年好宝贝 / Mumiy Troll - Happy New Year ...        rock   \n",
       "19123  🎄#премьера песни!!! 2019г(вы не спешите говори...     shanson   \n",
       "\n",
       "               performer  video_age       date  year     month year-month  \n",
       "0                  2маши       1141 2016-01-29  2016   January 2016-01-01  \n",
       "1                лазарев       3217 2010-05-24  2010       May 2010-05-01  \n",
       "2                лазарев       3546 2009-06-29  2009      June 2009-06-01  \n",
       "3                лазарев       3189 2010-06-21  2010      June 2010-06-01  \n",
       "4                лазарев       3361 2009-12-31  2009  December 2009-12-01  \n",
       "...                  ...        ...        ...   ...       ...        ...  \n",
       "19119            бутырка         44 2019-01-30  2019   January 2019-01-01  \n",
       "19120  classical_channel        239 2018-07-19  2018      July 2018-07-01  \n",
       "19121             элджей        339 2018-04-10  2018     April 2018-04-01  \n",
       "19122       мумий_тролль        997 2016-06-21  2016      June 2016-06-01  \n",
       "19123            бутырка         65 2019-01-09  2019   January 2019-01-01  \n",
       "\n",
       "[19124 rows x 8 columns]"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna(axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "legislative-theology",
   "metadata": {},
   "source": [
    "Часто удаление слишком дорогая операция, так как избавляемся от кучи данных, которых и так частенько бывает мало"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "standard-cowboy",
   "metadata": {},
   "source": [
    "### Заполнение"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "id": "taken-vietnamese",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    24.0\n",
       "1     NaN\n",
       "2    27.0\n",
       "3     2.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# было\n",
    "ser = pd.Series([24, np.nan, 27, 2])\n",
    "ser"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "id": "appreciated-pittsburgh",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    24.0\n",
       "1     0.0\n",
       "2    27.0\n",
       "3     2.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 236,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# заполняем нулем\n",
    "ser.fillna(0).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "id": "5493726c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    24.000000\n",
       "1    17.666667\n",
       "2    27.000000\n",
       "3     2.000000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ser.fillna(ser.mean()).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "id": "weird-brass",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    24.0\n",
       "1    24.0\n",
       "2    27.0\n",
       "3     2.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 289,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# заполняем предыдущим значением\n",
    "ser.fillna(method='ffill').head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "id": "waiting-theater",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    24.0\n",
       "1    27.0\n",
       "2    27.0\n",
       "3     2.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# заполняем последующим значением\n",
    "ser.fillna(method='bfill').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "serious-signature",
   "metadata": {},
   "source": [
    "### Задача №7:\n",
    "\n",
    "Заполните пропуски в столбцах таблицы по клипам средними"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "id": "unlike-measurement",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-244-208a407bf89c>:2: FutureWarning: DataFrame.mean and DataFrame.median with numeric_only=None will include datetime64 and datetime64tz columns in a future version.\n",
      "  df.fillna(df.mean(), inplace=True)\n"
     ]
    }
   ],
   "source": [
    "## Ваш код\n",
    "df.fillna(df.mean(), inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "qualified-netscape",
   "metadata": {},
   "source": [
    "## Обнаружение аномалий"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "centered-annual",
   "metadata": {},
   "source": [
    "Аномалии в данных $-$ это столь же частая проблема, что и пропущенные значения, однако более серьезная. Посмотрим на примере, как аномалии могут все испортить:  \n",
    "\n",
    "### Оцениваем средний рост людей в школьном классе:\n",
    "\n",
    "Представим, что мы хотим оценить средний рост людей в школьном классе. Имеем следующую таблицу:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "id": "fallen-protest",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>student</th>\n",
       "      <th>height_cm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Max</td>\n",
       "      <td>170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ahmed</td>\n",
       "      <td>165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Nastya</td>\n",
       "      <td>171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Vlad</td>\n",
       "      <td>168</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  student  height_cm\n",
       "0     Max        170\n",
       "1   Ahmed        165\n",
       "2  Nastya        171\n",
       "3    Vlad        168"
      ]
     },
     "execution_count": 252,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "students = pd.DataFrame({'student': ['Max', 'Ahmed', 'Nastya', 'Vlad'],\n",
    "                        'height_cm': [170, 165, 171, 168]})\n",
    "\n",
    "students"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "about-viking",
   "metadata": {},
   "source": [
    "Найдем среднее:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "id": "frozen-federal",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "168.5"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "students.height_cm.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "smaller-transition",
   "metadata": {},
   "source": [
    "Теперь представим, что в наш класс пришел новенький Филипп, который остался на второй год из-за того, что играл в баскетбол, а не учился. У нас в таблице появилось новое наблюдение, попробуем оценить рост теперь. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "id": "exclusive-eagle",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>student</th>\n",
       "      <th>height_cm</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Max</td>\n",
       "      <td>170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Ahmed</td>\n",
       "      <td>165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Nastya</td>\n",
       "      <td>171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Vlad</td>\n",
       "      <td>168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Phil</td>\n",
       "      <td>210</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  student  height_cm\n",
       "0     Max        170\n",
       "1   Ahmed        165\n",
       "2  Nastya        171\n",
       "3    Vlad        168\n",
       "4    Phil        210"
      ]
     },
     "execution_count": 254,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "students.loc[4] = ['Phil', 210]\n",
    "students"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "id": "descending-twins",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "176.8"
      ]
     },
     "execution_count": 255,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "students.height_cm.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "logical-plaza",
   "metadata": {},
   "source": [
    "Видим, что значение среднего роста *сильно изменилось* из-за одного человека, имеющего неспецифичный (*аномальный*) рост для нашей выборки. Связано это с тем, что Филипп пришел к нам в класс *из другого распределения* $-$ распределения баскетболистов. Именно поэтому значение нашего среднего теперь *сильно завышено*, и мы таким образом можем собрать *ошибочную статистику*."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "common-china",
   "metadata": {},
   "source": [
    "Возможное решение: использовать для агрегированной статистики квантили (например, медиану)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "id": "seventh-spending",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "170.0"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "students.height_cm.median()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "documentary-diamond",
   "metadata": {},
   "source": [
    "**ЗАПОМНИТЬ: Медиана в отличие от среднего устойчива к выбросам!**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "painful-sussex",
   "metadata": {},
   "source": [
    "### Выявление и удаление выбросов на примере клипов:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "id": "restricted-hindu",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5735.014482758621"
      ]
     },
     "execution_count": 297,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.likeCount.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "id": "worldwide-running",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "100.0"
      ]
     },
     "execution_count": 298,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.likeCount.median()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "automatic-leonard",
   "metadata": {},
   "source": [
    "Видим сильное завышение среднего по лайкам в отличие от медианы. Причина - аномалии. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "opposite-celebration",
   "metadata": {},
   "source": [
    "На распределении виден длинный хвост:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "id": "greater-metallic",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.likeCount.hist(bins=30);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "limiting-peoples",
   "metadata": {},
   "source": [
    "Если не особо виден, то можно посмотреть на лог шкалу:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "id": "shaped-stations",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.likeCount.hist(bins=30, log=True);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "drawn-multiple",
   "metadata": {},
   "source": [
    "То есть у нас в выборке есть клипы с аномальным кол-вом лайков - так называемые хиты. Они то нам и портят всю статистику.  \n",
    "\n",
    "Попробуем от них избавиться:\n",
    "\n",
    "**Способ №1:** Откинем $1\\%$ нашей выборки клипов, у которых больше всего лайков. Воспользуемся для этого $99\\%$ персентилем."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "id": "da369d2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>performer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1412</th>\n",
       "      <td>Four Seasons ~ Vivaldi</td>\n",
       "      <td>classical_channel</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17571</th>\n",
       "      <td>Тимати feat. Егор Крид - Гучи (премьера клипа,...</td>\n",
       "      <td>тимати</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8738</th>\n",
       "      <td>Егор Крид feat. Филипп Киркоров - Цвет настрое...</td>\n",
       "      <td>крид</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2635</th>\n",
       "      <td>Miyagi &amp; Эндшпиль feat. Рем Дигга - I Got Love...</td>\n",
       "      <td>miyagi&amp;эндшпиль</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18649</th>\n",
       "      <td>Элджей &amp; Feduk - Розовое вино</td>\n",
       "      <td>feduk</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   title          performer\n",
       "1412                              Four Seasons ~ Vivaldi  classical_channel\n",
       "17571  Тимати feat. Егор Крид - Гучи (премьера клипа,...             тимати\n",
       "8738   Егор Крид feat. Филипп Киркоров - Цвет настрое...               крид\n",
       "2635   Miyagi & Эндшпиль feat. Рем Дигга - I Got Love...    miyagi&эндшпиль\n",
       "18649                      Элджей & Feduk - Розовое вино              feduk"
      ]
     },
     "execution_count": 271,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.likeCount >= df.likeCount.quantile(0.9995)].sort_values('likeCount').tail()[['title','performer']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "id": "bulgarian-grounds",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold = df.likeCount.quantile(0.99)\n",
    "df_clear = df[df.likeCount <= threshold]\n",
    "\n",
    "# визуализируем\n",
    "df_clear.likeCount.hist(bins=30);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ordinary-arabic",
   "metadata": {},
   "source": [
    "**Способ №2:** Считаем аномалиями все наблюдения, которые оказались за пределами трех межквартильных отклонений.  \n",
    "\n",
    "### Задача №8:\n",
    "\n",
    "Реализовать этот метод. Межквартильное отклонение - это разница между 25-м и 75-м перцентилями. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "id": "domestic-warning",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Ваш код\n",
    "\n",
    "q1 = df.likeCount.quantile(0.25)\n",
    "q2 = df.likeCount.quantile(0.75)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "id": "d771742d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.query(f'{q1}<likeCount<{q2}')['likeCount'].plot(kind='hist', bins=20);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "decent-moisture",
   "metadata": {},
   "source": [
    "Стоит отметить, что иногда искать аномалии требуется не только для обработки данных. Порой, поиск аномалий $-$ это и есть задача. Так, например, аномалиями являются мошенники банка, чье поведение отличается от обычных пользователей. И поиск таких людей и является нашей задачей. "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.4"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": true,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}