ml_18-Analysis text data part 1 Practice.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "bc784456",
   "metadata": {},
   "source": [
    "# Memology - DataScience & Memes\n",
    "![](http://orig07.deviantart.net/3e62/f/2015/267/4/1/meme_team_logo_png_smol_by_kuiwi-d9ass0h.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "498419fb",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import json\n",
    "\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "le = LabelEncoder()\n",
    "\n",
    "import seaborn as sns\n",
    "from pylab import rcParams\n",
    "rcParams['figure.figsize']=8, 8\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('ggplot')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9682ee14",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import StratifiedKFold, train_test_split, RandomizedSearchCV\n",
    "from sklearn.metrics import classification_report, confusion_matrix, accuracy_score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d50a3341",
   "metadata": {},
   "source": [
    "### Часть 1. Описание набора данных и признаков"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b04ccb32",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes = pd.read_csv('datasets/MEMES_new_short.csv', index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "12b692fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4206, 14)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1d88924d",
   "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>status</th>\n",
       "      <th>type</th>\n",
       "      <th>origin_year</th>\n",
       "      <th>origin_place</th>\n",
       "      <th>date_added</th>\n",
       "      <th>views</th>\n",
       "      <th>videos</th>\n",
       "      <th>photos</th>\n",
       "      <th>comments</th>\n",
       "      <th>tags</th>\n",
       "      <th>about</th>\n",
       "      <th>origin</th>\n",
       "      <th>other_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Muvvafukka</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Viral Video</td>\n",
       "      <td>2019</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>2019-10-24T04:05:03-04:00</td>\n",
       "      <td>13613</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>jay-lee hilzinger &amp; maddie flynn, eshayz, davi...</td>\n",
       "      <td>Muvvafukka refers to a viral video of two teen...</td>\n",
       "      <td>While it appears the original video is deleted...</td>\n",
       "      <td>Butler also made a screenshot of Hilzinger's f...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Martyrdom (Drop a Live Grenade When Killed)</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Reaction</td>\n",
       "      <td>2007</td>\n",
       "      <td>Call of Duty 4: Modern Warfare</td>\n",
       "      <td>2019-10-23T12:13:13-04:00</td>\n",
       "      <td>13973</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>8</td>\n",
       "      <td>perk, martyrdom, live grenade, reaction image,...</td>\n",
       "      <td>Martyrdom (Drop a Live Grenade When Killed) re...</td>\n",
       "      <td>On April 12th, 2007, first person shooter Call...</td>\n",
       "      <td>Spread The exact first instance of use of the ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Wet Kylo Ren</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Image Macro</td>\n",
       "      <td>2019</td>\n",
       "      <td>Star Wars: The Rise of Skywalker (Movie Trailer)</td>\n",
       "      <td>2019-10-23T11:56:19-04:00</td>\n",
       "      <td>5339</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>kylo ren, star wars the rise of skywalker, ada...</td>\n",
       "      <td>Wet Kylo Ren is an image macro series based on...</td>\n",
       "      <td>On October 21st, 2019, the final trailer for t...</td>\n",
       "      <td>Shortly after the trailer's release, some on T...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>You're Invited</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Parody</td>\n",
       "      <td>2019</td>\n",
       "      <td>TikTok</td>\n",
       "      <td>2019-10-23T10:32:04-04:00</td>\n",
       "      <td>2249</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>byob, byow, party, flyer, charge, door</td>\n",
       "      <td>You're Invited is a series of TikTok videos pa...</td>\n",
       "      <td>On October 13th, 2019, TikTok user @coolguycol...</td>\n",
       "      <td>Spread On October 15th, 2019, @moneyxmo upload...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Companies Posting Their Own Names</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Participatory Media</td>\n",
       "      <td>2019</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>2019-10-25T12:42:22-04:00</td>\n",
       "      <td>5662</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>33</td>\n",
       "      <td>social media campaign, brands, companies, max ...</td>\n",
       "      <td>Companies Posting Their Own Names refers to a ...</td>\n",
       "      <td>On October 18th, 2019, Australian musician and...</td>\n",
       "      <td>Before October 2019, a number of companies and...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          name      status  \\\n",
       "0                                   Muvvafukka  Submission   \n",
       "1  Martyrdom (Drop a Live Grenade When Killed)  Submission   \n",
       "2                                 Wet Kylo Ren  Submission   \n",
       "3                               You're Invited  Submission   \n",
       "4            Companies Posting Their Own Names  Submission   \n",
       "\n",
       "                  type origin_year  \\\n",
       "0          Viral Video        2019   \n",
       "1             Reaction        2007   \n",
       "2          Image Macro        2019   \n",
       "3               Parody        2019   \n",
       "4  Participatory Media        2019   \n",
       "\n",
       "                                       origin_place  \\\n",
       "0                                          Facebook   \n",
       "1                    Call of Duty 4: Modern Warfare   \n",
       "2  Star Wars: The Rise of Skywalker (Movie Trailer)   \n",
       "3                                            TikTok   \n",
       "4                                          Facebook   \n",
       "\n",
       "                  date_added  views  videos  photos  comments  \\\n",
       "0  2019-10-24T04:05:03-04:00  13613       0       7         8   \n",
       "1  2019-10-23T12:13:13-04:00  13973       0      17         8   \n",
       "2  2019-10-23T11:56:19-04:00   5339       0      12         7   \n",
       "3  2019-10-23T10:32:04-04:00   2249       0       0         0   \n",
       "4  2019-10-25T12:42:22-04:00   5662       0      16        33   \n",
       "\n",
       "                                                tags  \\\n",
       "0  jay-lee hilzinger & maddie flynn, eshayz, davi...   \n",
       "1  perk, martyrdom, live grenade, reaction image,...   \n",
       "2  kylo ren, star wars the rise of skywalker, ada...   \n",
       "3             byob, byow, party, flyer, charge, door   \n",
       "4  social media campaign, brands, companies, max ...   \n",
       "\n",
       "                                               about  \\\n",
       "0  Muvvafukka refers to a viral video of two teen...   \n",
       "1  Martyrdom (Drop a Live Grenade When Killed) re...   \n",
       "2  Wet Kylo Ren is an image macro series based on...   \n",
       "3  You're Invited is a series of TikTok videos pa...   \n",
       "4  Companies Posting Their Own Names refers to a ...   \n",
       "\n",
       "                                              origin  \\\n",
       "0  While it appears the original video is deleted...   \n",
       "1  On April 12th, 2007, first person shooter Call...   \n",
       "2  On October 21st, 2019, the final trailer for t...   \n",
       "3  On October 13th, 2019, TikTok user @coolguycol...   \n",
       "4  On October 18th, 2019, Australian musician and...   \n",
       "\n",
       "                                          other_text  \n",
       "0  Butler also made a screenshot of Hilzinger's f...  \n",
       "1  Spread The exact first instance of use of the ...  \n",
       "2  Shortly after the trailer's release, some on T...  \n",
       "3  Spread On October 15th, 2019, @moneyxmo upload...  \n",
       "4  Before October 2019, a number of companies and...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ea40f8a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes.reset_index(drop=True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "23f18a9e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4206 entries, 0 to 4205\n",
      "Data columns (total 14 columns):\n",
      " #   Column        Non-Null Count  Dtype \n",
      "---  ------        --------------  ----- \n",
      " 0   name          4206 non-null   object\n",
      " 1   status        4206 non-null   object\n",
      " 2   type          1786 non-null   object\n",
      " 3   origin_year   4206 non-null   object\n",
      " 4   origin_place  4205 non-null   object\n",
      " 5   date_added    4206 non-null   object\n",
      " 6   views         4206 non-null   int64 \n",
      " 7   videos        4206 non-null   int64 \n",
      " 8   photos        4206 non-null   int64 \n",
      " 9   comments      4206 non-null   int64 \n",
      " 10  tags          4205 non-null   object\n",
      " 11  about         3798 non-null   object\n",
      " 12  origin        2524 non-null   object\n",
      " 13  other_text    3508 non-null   object\n",
      "dtypes: int64(4), object(10)\n",
      "memory usage: 460.2+ KB\n"
     ]
    }
   ],
   "source": [
    "memes.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b22ca035",
   "metadata": {},
   "source": [
    "Итак, пройдемся по переменным:\n",
    "\n",
    "- name - название мема\n",
    "- added - дата добавления\n",
    "- views - число просмотров на сайте\n",
    "- comments - число комментариев\n",
    "- status - категориальная переменная, статус присваивается на сайте в зависимости от \"одобрения\" мема сообществом\n",
    "- year - год, к которому относится первоисточник появления картинки (например, год выхода определенной серии Симпсонов)\n",
    "- tags - тэги, также являются категориями\n",
    "- about - описание мема\n",
    "- origin - описание происхождения\n",
    "- spread - где приобрел особенную популярность"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "eae647a0",
   "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>status</th>\n",
       "      <th>type</th>\n",
       "      <th>origin_year</th>\n",
       "      <th>origin_place</th>\n",
       "      <th>date_added</th>\n",
       "      <th>views</th>\n",
       "      <th>videos</th>\n",
       "      <th>photos</th>\n",
       "      <th>comments</th>\n",
       "      <th>tags</th>\n",
       "      <th>about</th>\n",
       "      <th>origin</th>\n",
       "      <th>other_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Muvvafukka</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Viral Video</td>\n",
       "      <td>2019</td>\n",
       "      <td>Facebook</td>\n",
       "      <td>2019-10-24T04:05:03-04:00</td>\n",
       "      <td>13613</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>8</td>\n",
       "      <td>jay-lee hilzinger &amp; maddie flynn, eshayz, davi...</td>\n",
       "      <td>Muvvafukka refers to a viral video of two teen...</td>\n",
       "      <td>While it appears the original video is deleted...</td>\n",
       "      <td>Butler also made a screenshot of Hilzinger's f...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Martyrdom (Drop a Live Grenade When Killed)</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Reaction</td>\n",
       "      <td>2007</td>\n",
       "      <td>Call of Duty 4: Modern Warfare</td>\n",
       "      <td>2019-10-23T12:13:13-04:00</td>\n",
       "      <td>13973</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>8</td>\n",
       "      <td>perk, martyrdom, live grenade, reaction image,...</td>\n",
       "      <td>Martyrdom (Drop a Live Grenade When Killed) re...</td>\n",
       "      <td>On April 12th, 2007, first person shooter Call...</td>\n",
       "      <td>Spread The exact first instance of use of the ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Wet Kylo Ren</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Image Macro</td>\n",
       "      <td>2019</td>\n",
       "      <td>Star Wars: The Rise of Skywalker (Movie Trailer)</td>\n",
       "      <td>2019-10-23T11:56:19-04:00</td>\n",
       "      <td>5339</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>kylo ren, star wars the rise of skywalker, ada...</td>\n",
       "      <td>Wet Kylo Ren is an image macro series based on...</td>\n",
       "      <td>On October 21st, 2019, the final trailer for t...</td>\n",
       "      <td>Shortly after the trailer's release, some on T...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          name      status         type  \\\n",
       "0                                   Muvvafukka  Submission  Viral Video   \n",
       "1  Martyrdom (Drop a Live Grenade When Killed)  Submission     Reaction   \n",
       "2                                 Wet Kylo Ren  Submission  Image Macro   \n",
       "\n",
       "  origin_year                                      origin_place  \\\n",
       "0        2019                                          Facebook   \n",
       "1        2007                    Call of Duty 4: Modern Warfare   \n",
       "2        2019  Star Wars: The Rise of Skywalker (Movie Trailer)   \n",
       "\n",
       "                  date_added  views  videos  photos  comments  \\\n",
       "0  2019-10-24T04:05:03-04:00  13613       0       7         8   \n",
       "1  2019-10-23T12:13:13-04:00  13973       0      17         8   \n",
       "2  2019-10-23T11:56:19-04:00   5339       0      12         7   \n",
       "\n",
       "                                                tags  \\\n",
       "0  jay-lee hilzinger & maddie flynn, eshayz, davi...   \n",
       "1  perk, martyrdom, live grenade, reaction image,...   \n",
       "2  kylo ren, star wars the rise of skywalker, ada...   \n",
       "\n",
       "                                               about  \\\n",
       "0  Muvvafukka refers to a viral video of two teen...   \n",
       "1  Martyrdom (Drop a Live Grenade When Killed) re...   \n",
       "2  Wet Kylo Ren is an image macro series based on...   \n",
       "\n",
       "                                              origin  \\\n",
       "0  While it appears the original video is deleted...   \n",
       "1  On April 12th, 2007, first person shooter Call...   \n",
       "2  On October 21st, 2019, the final trailer for t...   \n",
       "\n",
       "                                          other_text  \n",
       "0  Butler also made a screenshot of Hilzinger's f...  \n",
       "1  Spread The exact first instance of use of the ...  \n",
       "2  Shortly after the trailer's release, some on T...  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "cba60d53",
   "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>status</th>\n",
       "      <th>type</th>\n",
       "      <th>origin_year</th>\n",
       "      <th>origin_place</th>\n",
       "      <th>date_added</th>\n",
       "      <th>views</th>\n",
       "      <th>videos</th>\n",
       "      <th>photos</th>\n",
       "      <th>comments</th>\n",
       "      <th>tags</th>\n",
       "      <th>about</th>\n",
       "      <th>origin</th>\n",
       "      <th>other_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4203</th>\n",
       "      <td>Because of Parkinson's</td>\n",
       "      <td>Confirmed</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2006</td>\n",
       "      <td>YTMND</td>\n",
       "      <td>2013-09-26T16:26:49-04:00</td>\n",
       "      <td>205323</td>\n",
       "      <td>2</td>\n",
       "      <td>101</td>\n",
       "      <td>61</td>\n",
       "      <td>michael j. fox, joke, image macro, 4chan, cele...</td>\n",
       "      <td>Because of Parkinson's (also known as \"Michael...</td>\n",
       "      <td>Michel J. Fox is an Canadian–American actor an...</td>\n",
       "      <td>The first instance of Michael J. Fox's equatio...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4204</th>\n",
       "      <td>Paranoid Parrot</td>\n",
       "      <td>Confirmed</td>\n",
       "      <td>Image Macro</td>\n",
       "      <td>2009</td>\n",
       "      <td>r9k</td>\n",
       "      <td>2019-10-31T03:47:02-04:00</td>\n",
       "      <td>416158</td>\n",
       "      <td>2</td>\n",
       "      <td>218</td>\n",
       "      <td>51</td>\n",
       "      <td>4chan, parrot, paranoia, paranoid parrot, advi...</td>\n",
       "      <td>Paranoid Parrot is an advice animal image macr...</td>\n",
       "      <td>While the photographer is unknown, the origina...</td>\n",
       "      <td>In 2007, the same image was used to illustrate...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4205</th>\n",
       "      <td>I Believe You Have My Stapler</td>\n",
       "      <td>Confirmed</td>\n",
       "      <td>Pop Culture Reference</td>\n",
       "      <td>2004</td>\n",
       "      <td>YTMND</td>\n",
       "      <td>2018-04-09T16:28:49-04:00</td>\n",
       "      <td>109512</td>\n",
       "      <td>1</td>\n",
       "      <td>25</td>\n",
       "      <td>29</td>\n",
       "      <td>movie, catchphrase, film, office space, movie ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Office Space was a 1999 comedy by Beavis and B...</td>\n",
       "      <td>Milton Waddams (played by actor Stephen Root) ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               name     status                   type  \\\n",
       "4203         Because of Parkinson's  Confirmed                    NaN   \n",
       "4204                Paranoid Parrot  Confirmed            Image Macro   \n",
       "4205  I Believe You Have My Stapler  Confirmed  Pop Culture Reference   \n",
       "\n",
       "     origin_year origin_place                 date_added   views  videos  \\\n",
       "4203        2006        YTMND  2013-09-26T16:26:49-04:00  205323       2   \n",
       "4204        2009          r9k  2019-10-31T03:47:02-04:00  416158       2   \n",
       "4205        2004        YTMND  2018-04-09T16:28:49-04:00  109512       1   \n",
       "\n",
       "      photos  comments                                               tags  \\\n",
       "4203     101        61  michael j. fox, joke, image macro, 4chan, cele...   \n",
       "4204     218        51  4chan, parrot, paranoia, paranoid parrot, advi...   \n",
       "4205      25        29  movie, catchphrase, film, office space, movie ...   \n",
       "\n",
       "                                                  about  \\\n",
       "4203  Because of Parkinson's (also known as \"Michael...   \n",
       "4204  Paranoid Parrot is an advice animal image macr...   \n",
       "4205                                                NaN   \n",
       "\n",
       "                                                 origin  \\\n",
       "4203  Michel J. Fox is an Canadian–American actor an...   \n",
       "4204  While the photographer is unknown, the origina...   \n",
       "4205  Office Space was a 1999 comedy by Beavis and B...   \n",
       "\n",
       "                                             other_text  \n",
       "4203  The first instance of Michael J. Fox's equatio...  \n",
       "4204  In 2007, the same image was used to illustrate...  \n",
       "4205  Milton Waddams (played by actor Stephen Root) ...  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.tail(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb1cb148",
   "metadata": {},
   "source": [
    "### Часть 2. первичный анализ признаков\n",
    "\n",
    "Конечно же в первую очередь хочется посмотреть на самый популярный мем"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3a12eb6b",
   "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>status</th>\n",
       "      <th>type</th>\n",
       "      <th>origin_year</th>\n",
       "      <th>origin_place</th>\n",
       "      <th>date_added</th>\n",
       "      <th>views</th>\n",
       "      <th>videos</th>\n",
       "      <th>photos</th>\n",
       "      <th>comments</th>\n",
       "      <th>tags</th>\n",
       "      <th>about</th>\n",
       "      <th>origin</th>\n",
       "      <th>other_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3958</th>\n",
       "      <td>Me Gusta</td>\n",
       "      <td>Confirmed</td>\n",
       "      <td>Character</td>\n",
       "      <td>2010</td>\n",
       "      <td>4chan</td>\n",
       "      <td>2018-09-27T16:03:51-04:00</td>\n",
       "      <td>9095857</td>\n",
       "      <td>21</td>\n",
       "      <td>518</td>\n",
       "      <td>142</td>\n",
       "      <td>4chan, rageguy, spinoff, 4 panel comic, me gus...</td>\n",
       "      <td>Me Gusta is a rage comic face that is typicall...</td>\n",
       "      <td>The colloquial Spanish phrase \"Me Gusta\" was f...</td>\n",
       "      <td>Spread In December 2010, the single topic blog...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          name     status       type origin_year origin_place  \\\n",
       "3958  Me Gusta  Confirmed  Character        2010        4chan   \n",
       "\n",
       "                     date_added    views  videos  photos  comments  \\\n",
       "3958  2018-09-27T16:03:51-04:00  9095857      21     518       142   \n",
       "\n",
       "                                                   tags  \\\n",
       "3958  4chan, rageguy, spinoff, 4 panel comic, me gus...   \n",
       "\n",
       "                                                  about  \\\n",
       "3958  Me Gusta is a rage comic face that is typicall...   \n",
       "\n",
       "                                                 origin  \\\n",
       "3958  The colloquial Spanish phrase \"Me Gusta\" was f...   \n",
       "\n",
       "                                             other_text  \n",
       "3958  Spread In December 2010, the single topic blog...  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes[memes.views==memes.views.max()]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "722b409a",
   "metadata": {},
   "source": [
    "### Время добавления\n",
    "Так как в нашем распоряжении есть переменная added, то можно поиграться с ней, преобразовав в понятный питону формат\n",
    "и получить час, день, месяц, год добавления"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b5a556ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes['Timestamp'] = pd.to_datetime(memes['date_added'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "58d657a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4206 entries, 0 to 4205\n",
      "Data columns (total 20 columns):\n",
      " #   Column        Non-Null Count  Dtype \n",
      "---  ------        --------------  ----- \n",
      " 0   name          4206 non-null   object\n",
      " 1   status        4206 non-null   object\n",
      " 2   type          1786 non-null   object\n",
      " 3   origin_year   4206 non-null   object\n",
      " 4   origin_place  4205 non-null   object\n",
      " 5   date_added    4206 non-null   object\n",
      " 6   views         4206 non-null   int64 \n",
      " 7   videos        4206 non-null   int64 \n",
      " 8   photos        4206 non-null   int64 \n",
      " 9   comments      4206 non-null   int64 \n",
      " 10  tags          4205 non-null   object\n",
      " 11  about         3798 non-null   object\n",
      " 12  origin        2524 non-null   object\n",
      " 13  other_text    3508 non-null   object\n",
      " 14  Timestamp     4206 non-null   object\n",
      " 15  Year          4206 non-null   int64 \n",
      " 16  Month         4206 non-null   int64 \n",
      " 17  Day           4206 non-null   int64 \n",
      " 18  Hour          4206 non-null   int64 \n",
      " 19  Weekday       4206 non-null   int64 \n",
      "dtypes: int64(9), object(11)\n",
      "memory usage: 657.3+ KB\n"
     ]
    }
   ],
   "source": [
    "memes['Year'] = memes.Timestamp.apply(lambda x: x.year)\n",
    "memes['Month'] = memes.Timestamp.apply(lambda x: x.month)\n",
    "memes['Day'] = memes.Timestamp.apply(lambda x: x.day)\n",
    "memes['Hour'] = memes.Timestamp.apply(lambda x: x.hour)\n",
    "memes['Weekday'] = memes.Timestamp.apply(lambda x: x.weekday())\n",
    "\n",
    "memes.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c3693b8",
   "metadata": {},
   "source": [
    "Так же пригодится переменная, отвечающая за число дней, прошедших с момента досбавления на сайт"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fdc1ee00",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes['days_from_creation'] = (memes.Timestamp.max() - memes.Timestamp) / np.timedelta64(1, 'D')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "002d6e25",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2019-10-24T04:05:03-04:00'"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.date_added[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "5f1ba4a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.to_datetime(memes.date_added[0]).weekday()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9b4f820",
   "metadata": {},
   "source": [
    "А теперь, для наших созданных переменных построим несколько гистограмм"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f907a369",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, axarr = plt.subplots(2, 2, figsize=(15, 10))\n",
    "times = ['Month', 'Day', 'Hour', 'Weekday']\n",
    "bins = [12, 31, 24, 7]\n",
    "index = 0\n",
    "for i in range(2):\n",
    "    for j in range(2):\n",
    "        axarr[i, j].hist(memes[times[index]], bins=bins[index])\n",
    "        axarr[i, j].set_title(f'Histogram of {times[index]}')\n",
    "\n",
    "        index += 1\n",
    "\n",
    "f.subplots_adjust(hspace=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "58da3c2d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "memes.Year.value_counts().plot(kind='bar', figsize=(10, 5))\n",
    "plt.title('Histogram of the year')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "5ebd58ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2019    27.41\n",
       "2018    15.93\n",
       "2017    11.70\n",
       "2016     9.77\n",
       "2015     7.66\n",
       "2014     6.92\n",
       "2011     5.37\n",
       "2012     5.33\n",
       "2010     4.97\n",
       "2013     4.33\n",
       "2009     0.62\n",
       "Name: Year, dtype: object"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.Year.value_counts(normalize=True).apply(lambda x: f'{x*100:.2f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "5c0682bf",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-19-ef6432d72ec0>:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\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",
      "  memes.tags[memes.tags.isin(['none...', 'none', 'Unknown', 'none... Unknown', 'None'])] = 'unknown'\n"
     ]
    }
   ],
   "source": [
    "pd.set_option('mode.chained_assignment','warn')\n",
    "memes.tags[memes.tags.isin(['none...', 'none', 'Unknown', 'none... Unknown', 'None'])] = 'unknown'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "21540d7d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count        4206\n",
       "unique        117\n",
       "top       Unknown\n",
       "freq          349\n",
       "Name: origin_year, dtype: object"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.origin_year.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d079e7ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-21-6a39d0d8ae4e>:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\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",
      "  memes.origin_year[memes.origin_year=='Unknown'] = None\n"
     ]
    }
   ],
   "source": [
    "memes.origin_year[memes.origin_year=='Unknown'] = None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "58c3d649",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes.origin_year.fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "8661952d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-23-251f79d4276b>:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\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",
      "  memes.origin_year[~memes.origin_year.apply(lambda x: str(x).isdigit())] = 0\n"
     ]
    }
   ],
   "source": [
    "memes.origin_year[~memes.origin_year.apply(lambda x: str(x).isdigit())] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "65f44f2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       True\n",
       "1       True\n",
       "2       True\n",
       "3       True\n",
       "4       True\n",
       "        ... \n",
       "4201    True\n",
       "4202    True\n",
       "4203    True\n",
       "4204    True\n",
       "4205    True\n",
       "Name: origin_year, Length: 4206, dtype: bool"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.origin_year.apply(lambda x: str(x).isdigit())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "584527a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# memes.origin_year[(memes.origin_year>2000)&(memes.origin_year<=2020)].hist(bins=200)\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "950d4fb3",
   "metadata": {},
   "source": [
    "### Число просмотров"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "d9a1a8a7",
   "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": [
    "memes.views[memes.views>0].hist(bins=10)\n",
    "plt.title('Hist of Views')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "cde314f4",
   "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": [
    "memes.views[memes.views>0].apply(np.log).hist(bins=10)\n",
    "plt.title('Hist of log Views')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "251f9372",
   "metadata": {},
   "source": [
    "### Популярность"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6a6a8b97",
   "metadata": {},
   "source": [
    "Метрикой оценки популярности, в данном случае будет оценка среднего числа просмотров в день. Для этого разделим\n",
    "фактическое число просмотров на число дней, прошедших со дня публикации."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "66b7923f",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes['average_views'] = memes.views/memes.days_from_creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "4556b5ee",
   "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>status</th>\n",
       "      <th>type</th>\n",
       "      <th>origin_year</th>\n",
       "      <th>origin_place</th>\n",
       "      <th>date_added</th>\n",
       "      <th>views</th>\n",
       "      <th>videos</th>\n",
       "      <th>photos</th>\n",
       "      <th>comments</th>\n",
       "      <th>...</th>\n",
       "      <th>origin</th>\n",
       "      <th>other_text</th>\n",
       "      <th>Timestamp</th>\n",
       "      <th>Year</th>\n",
       "      <th>Month</th>\n",
       "      <th>Day</th>\n",
       "      <th>Hour</th>\n",
       "      <th>Weekday</th>\n",
       "      <th>days_from_creation</th>\n",
       "      <th>average_views</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>831</th>\n",
       "      <td>Flex Tape</td>\n",
       "      <td>Submission</td>\n",
       "      <td>Advertisement</td>\n",
       "      <td>2017</td>\n",
       "      <td>Flex Seal</td>\n",
       "      <td>2019-11-06T14:33:40-05:00</td>\n",
       "      <td>92958</td>\n",
       "      <td>27</td>\n",
       "      <td>77</td>\n",
       "      <td>22</td>\n",
       "      <td>...</td>\n",
       "      <td>On January 12th, 2017, Flex Seal uploaded a co...</td>\n",
       "      <td>Spread On September 30th, 2017, popular YouTub...</td>\n",
       "      <td>2019-11-06 14:33:40-05:00</td>\n",
       "      <td>2019</td>\n",
       "      <td>11</td>\n",
       "      <td>6</td>\n",
       "      <td>14</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>inf</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          name      status           type origin_year origin_place  \\\n",
       "831  Flex Tape  Submission  Advertisement        2017    Flex Seal   \n",
       "\n",
       "                    date_added  views  videos  photos  comments  ...  \\\n",
       "831  2019-11-06T14:33:40-05:00  92958      27      77        22  ...   \n",
       "\n",
       "                                                origin  \\\n",
       "831  On January 12th, 2017, Flex Seal uploaded a co...   \n",
       "\n",
       "                                            other_text  \\\n",
       "831  Spread On September 30th, 2017, popular YouTub...   \n",
       "\n",
       "                     Timestamp  Year Month  Day  Hour  Weekday  \\\n",
       "831  2019-11-06 14:33:40-05:00  2019    11    6    14        2   \n",
       "\n",
       "     days_from_creation  average_views  \n",
       "831                 0.0            inf  \n",
       "\n",
       "[1 rows x 22 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes[memes['average_views'] == memes['average_views'].max()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "18528768",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-30-33d403789fc9>:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\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",
      "  memes['average_views'][index] = memes['views'][index]\n"
     ]
    }
   ],
   "source": [
    "index = memes['average_views'] == memes['average_views'].max()\n",
    "memes['average_views'][index] = memes['views'][index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "2d6fd610",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Hist of log average views')"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(memes.average_views.apply(np.log))\n",
    "plt.title('Hist of log average views')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a46814b3",
   "metadata": {},
   "source": [
    "### Посмотрим на самые популярные"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "a03ea1d1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Meme name: Colonel Meow\n",
      "Average views (per day) 536753\n",
      "================\n",
      "\n",
      "Meme name: Distracted Boyfriend\n",
      "Average views (per day) 466923\n",
      "================\n",
      "\n",
      "Meme name: Derp\n",
      "Average views (per day) 300836\n",
      "================\n",
      "\n",
      "Meme name: Need for Speed\n",
      "Average views (per day) 246737\n",
      "================\n",
      "\n",
      "Meme name: Shitposting\n",
      "Average views (per day) 210984\n",
      "================\n",
      "\n",
      "Meme name: Starter Packs\n",
      "Average views (per day) 167624\n",
      "================\n",
      "\n",
      "Meme name: Double D's Facts Book\n",
      "Average views (per day) 166858\n",
      "================\n",
      "\n"
     ]
    }
   ],
   "source": [
    "popular = memes.sort_values(by='average_views', ascending=False)\n",
    "for i in popular.index.tolist()[:7]:\n",
    "    print(f'Meme name: {popular.loc[i, \"name\"]}')\n",
    "    print(f'Average views (per day) {round(popular.loc[i, \"average_views\"])}')\n",
    "    print('================\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49862476",
   "metadata": {},
   "source": [
    "### Комментарии\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "bcd84f40",
   "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": [
    "memes.comments.hist()\n",
    "plt.title('Hist of comments')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c6f74a0",
   "metadata": {},
   "source": [
    "### Посмотрим на самые комментируемы"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "f21993a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes['average_comments'] = memes.comments/memes.days_from_creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "e87896a2",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-35-f5a439e103ca>:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\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",
      "  memes['average_comments'][index] = memes['comments'][index]\n"
     ]
    }
   ],
   "source": [
    "index = memes['average_comments']==memes['average_comments'].max()\n",
    "memes['average_comments'][index] = memes['comments'][index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "c9c40a4b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Meme name: Need for Speed\n",
      "Average comments (per day) 989\n",
      "================\n",
      "\n",
      "Meme name: Hero Cat\n",
      "Average comments (per day) 308\n",
      "================\n",
      "\n",
      "Meme name: Colonel Meow\n",
      "Average comments (per day) 302\n",
      "================\n",
      "\n",
      "Meme name: Miley Cyrus\n",
      "Average comments (per day) 99\n",
      "================\n",
      "\n",
      "Meme name: McDonald’s \"Happy\" Mascot\n",
      "Average comments (per day) 97\n",
      "================\n",
      "\n",
      "Meme name: Metro Man Sees Megamind and Roxie\n",
      "Average comments (per day) 87\n",
      "================\n",
      "\n",
      "Meme name: Double D's Facts Book\n",
      "Average comments (per day) 83\n",
      "================\n",
      "\n"
     ]
    }
   ],
   "source": [
    "popular = memes.sort_values(by='average_comments', ascending=False)\n",
    "for i in popular.index.tolist()[:7]:\n",
    "    print(f'Meme name: {popular.loc[i, \"name\"]}')\n",
    "    print(f'Average comments (per day) {round(popular.loc[i, \"average_comments\"])}')\n",
    "    print('================\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1a09c911",
   "metadata": {},
   "source": [
    "### Статус"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "aa68bbf4",
   "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": [
    "memes.status.value_counts().plot(kind='barh')\n",
    "plt.title('Hist of status')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "670743cf",
   "metadata": {},
   "source": [
    "А вот здесь можно попутно решать задачу множественной классификации - для каждого мема есть 3 состояния\n",
    "\n",
    "- Submission - добавлен недавно или ещё не стал \"признанным в обществе\"\n",
    "- Deadpool - если умер и не используется\n",
    "- Confirmed - активно используется и получил статус \"подтверждённого\" мема"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c644d714",
   "metadata": {},
   "source": [
    "### ML"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "c4a4e999",
   "metadata": {},
   "outputs": [],
   "source": [
    "import nltk\n",
    "from nltk.corpus import stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "8bec60c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "stop = stopwords.words('english')\n",
    "stemmer = nltk.stem.snowball.EnglishStemmer('english')\n",
    "\n",
    "def data_stemmer(text):\n",
    "    return ' '.join([stemmer.stem(word) for word in text.split(' ') if word not in stop])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "e8c20211",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes['description'] = memes.tags.fillna('')+' '+\\\n",
    "                       memes.about.fillna('')+' '+\\\n",
    "                       memes.origin.fillna('')+' '+\\\n",
    "                       memes.other_text.fillna('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "4a79518a",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes.description = memes.description.apply(lambda x: x.lower().replace('unknown', ''))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "28c6e903",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'jay-lee hilzinger & maddie flynn, eshayz, david jones, freakout, jay-lee hilzinger, maddie flynn, frazzle tazz, caiden butler, declan, zinger burgers, simbra, sir david attenborough, daniel sanders muvvafukka refers to a viral video of two teenage girls, jay-lee hilzinger and maddie flynn, having a tantrum in australian department store david jones. quotes from the video, as well as a face hilzinger made in the video, have been parodied in memes in australian areas of the internet. while it appears the original video is deleted, it appears to have been first posted to facebook on october 20th, 2019. the person holding the camera says its \"a sunday,\" indicating the likely date the tantrum was filmed was sunday, october 20th. a screen-capture of the video when it was live on facebook was posted to youtube by user frazzle tazz on the 21st.[1] the screen capture is taken from a video posted by facebook user caiden butler. butler has also posted the video to youtube (shown below). butler also made a screenshot of hilzinger\\'s face in the video his cover photo on the 20th.[2] the face (shown below), would be used in future photoshop memes.     spread after the video grew popular on facebook and youtube, australian internet users began joking about it with various memes. youtuber simbra edited clips of the two girls into a commercial for zinger burgers, making a pun on hilzinger\\'s name, gaining over 5,700 views (shown below, left). user declan made a video in which sir david attenborough\\'s planet earth narration plays over the clip (shown below, right).     the video was also the source of dozens of photoshops, many of which used the face butler screenshot for his cover photo. facebook user daniel sanders[3] posted a collection of photoshops, gaining over 1,400 reactions (examples shown below).     various examples          search interest   trends.embed.renderexplorewidget(\"timeseries\", {\"comparisonitem\":[{\"keyword\":\"muvvafukka\",\"geo\":\"\",\"time\":\"now 7-d\"}],\"category\":0,\"property\":\"\"}, {\"explorequery\":\"date=now%207-d&q=muvvafukka\",\"guestpath\":\"https://trends.google.com:443/trends/embed/\"}); know your meme store external references  [1] youtube – muvvafukka eshay loses her mind! muvaf*ka crazy eshayz bother people. jay lee hilzinger/maddie flynn [2] facebook – caiden butler [3] facebook – daniel sanders'"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "memes.description[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "a07f224f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'jay-le hilzing & maddi flynn, eshayz, david jones, freakout, jay-le hilzinger, maddi flynn, frazzl tazz, caiden butler, declan, zinger burgers, simbra, sir david attenborough, daniel sander muvvafukka refer viral video two teenag girls, jay-le hilzing maddi flynn, tantrum australian depart store david jones. quot video, well face hilzing made video, parodi meme australian area internet. appear origin video deleted, appear first post facebook octob 20th, 2019. person hold camera say \"a sunday,\" indic like date tantrum film sunday, octob 20th. screen-captur video live facebook post youtub user frazzl tazz 21st.[1] screen captur taken video post facebook user caiden butler. butler also post video youtub (shown below). butler also made screenshot hilzing face video cover photo 20th.[2] face (shown below), would use futur photoshop memes.     spread video grew popular facebook youtube, australian internet user began joke various memes. youtub simbra edit clip two girl commerci zinger burgers, make pun hilzing name, gain 5,700 view (shown below, left). user declan made video sir david attenborough planet earth narrat play clip (shown below, right).     video also sourc dozen photoshops, mani use face butler screenshot cover photo. facebook user daniel sanders[3] post collect photoshops, gain 1,400 reaction (exampl shown below).     various exampl          search interest   trends.embed.renderexplorewidget(\"timeseries\", {\"comparisonitem\":[{\"keyword\":\"muvvafukka\",\"geo\":\"\",\"time\":\"now 7-d\"}],\"category\":0,\"property\":\"\"}, {\"explorequery\":\"date=now%207-d&q=muvvafukka\",\"guestpath\":\"https://trends.google.com:443/trends/embed/\"}); know meme store extern refer  [1] youtub – muvvafukka eshay lose mind! muvaf*ka crazi eshayz bother people. jay lee hilzinger/maddi flynn [2] facebook – caiden butler [3] facebook – daniel sander'"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_stemmer(memes.description[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "d4835ae5",
   "metadata": {},
   "outputs": [],
   "source": [
    "memes.description = memes.description.apply(lambda x: data_stemmer(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "be506121",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = memes.description\n",
    "y = memes.average_views"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67c8ac86",
   "metadata": {},
   "source": [
    "Для TF-IDF попробуем использовать биграммы, а чтобы словарь не раздувался до неадекватных размеров, ограничим \n",
    "минимальную частоту содержания слов в документе 0.1%. Т.е. будем выкидывать такие слова и биграммы, \n",
    "которые встречаются реже, чем в 0.1% длокументов."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "e5fc7b9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfTransformer, TfidfVectorizer\n",
    "\n",
    "tf_vectorizer = TfidfVectorizer(stop_words=stop, ngram_range=(1, 2), min_df=0.001, norm='l2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "d68361c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "Xtr, Xval, ytr, yval = train_test_split(X, y, test_size=.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "8ae0e906",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2944, 38513), (1262, 38513))"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Xtr_tfidf = tf_vectorizer.fit_transform(Xtr)\n",
    "Xval_tfidf = tf_vectorizer.transform(Xval)\n",
    "Xtr_tfidf.shape, Xval_tfidf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "d957cb27",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<2944x38513 sparse matrix of type '<class 'numpy.float64'>'\n",
       "\twith 664372 stored elements in Compressed Sparse Row format>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Xtr_tfidf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed09a98d",
   "metadata": {},
   "source": [
    "Целевую переменную (среднее сичло просмотров в день), разделим на 5 групп по квантилям, одновременно получая наивный baseline - если мы по всем объектам примем один класс,то итоговая точность составит 1/5 = 20%. еще один подход плюс\n",
    "подхода - дисбаланса классов у нас точно не будет."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "a2d36aaf",
   "metadata": {},
   "outputs": [],
   "source": [
    "ytr_quantile = pd.qcut(ytr, q=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "67b42373",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.096, 7.114]          982\n",
       "(7.114, 58.965]         981\n",
       "(58.965, 536753.118]    981\n",
       "Name: average_views, dtype: int64"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ytr_quantile.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18ab0453",
   "metadata": {},
   "source": [
    "Для будущих нужд запомним имеющиеся интервалы"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "09313628",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<IntervalArray>\n",
       "[(0.096, 7.114], (7.114, 58.965], (58.965, 536753.118]]\n",
       "Length: 3, closed: right, dtype: interval[float64]"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intervals = ytr_quantile.unique().categories.values\n",
    "intervals"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3c9918a",
   "metadata": {},
   "source": [
    "Так как на текстовой выборке мы аналогичное преобразование с квантилем выполнять не можем (группы не будут совпадать)\n",
    "запишем функцию для преобразования тестовых значений"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "837cc298",
   "metadata": {},
   "outputs": [],
   "source": [
    "def quantile(y):\n",
    "    for interval in intervals:\n",
    "        if y <= interval.right:\n",
    "            return interval\n",
    "            break\n",
    "    return interval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "8071eccd",
   "metadata": {},
   "outputs": [],
   "source": [
    "yval_quantile = yval.apply(quantile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "15eda69c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(58.965, 536753.118]    436\n",
       "(7.114, 58.965]         426\n",
       "(0.096, 7.114]          400\n",
       "Name: average_views, dtype: int64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yval_quantile.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56d60eb7",
   "metadata": {},
   "source": [
    "Наконец для обучения зщакодируем интервалы при помощи LabelEncoder()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "125617a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "ytr_quantile = le.fit_transform(ytr_quantile)\n",
    "yval_quantile = le.transform(yval_quantile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "b806966a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 0, ..., 1, 0, 1])"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ytr_quantile"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "37ca2fae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([Interval(0.096, 7.114, closed='right'),\n",
       "       Interval(7.114, 58.965, closed='right'),\n",
       "       Interval(0.096, 7.114, closed='right'), ...,\n",
       "       Interval(7.114, 58.965, closed='right'),\n",
       "       Interval(0.096, 7.114, closed='right'),\n",
       "       Interval(7.114, 58.965, closed='right')], dtype=object)"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "le.inverse_transform(ytr_quantile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "21153434",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((2944, 38513), (1262, 38513))"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Xtr_tfidf.shape, Xval_tfidf.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "469d12b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import SGDClassifier, LogisticRegressionCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "9d940573",
   "metadata": {},
   "outputs": [],
   "source": [
    "logit_sgd = SGDClassifier(\n",
    "    loss='log',\n",
    "    shuffle=True,\n",
    "    n_iter_no_change=10,\n",
    "    max_iter=1000,\n",
    "    penalty='l1',\n",
    "    random_state=42\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "989b9a67",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8474864130434783\n",
      "0.6109350237717908\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.81      0.72      0.76       451\n",
      "           1       0.43      0.51      0.47       363\n",
      "           2       0.60      0.58      0.59       448\n",
      "\n",
      "    accuracy                           0.61      1262\n",
      "   macro avg       0.62      0.60      0.61      1262\n",
      "weighted avg       0.63      0.61      0.62      1262\n",
      "\n"
     ]
    }
   ],
   "source": [
    "logit_sgd.fit(Xtr_tfidf, ytr_quantile)\n",
    "print(accuracy_score(logit_sgd.predict(Xtr_tfidf), ytr_quantile))\n",
    "print(accuracy_score(logit_sgd.predict(Xval_tfidf), yval_quantile))\n",
    "\n",
    "print(classification_report(logit_sgd.predict(Xval_tfidf), yval_quantile))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "441a78f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegressionCV(n_jobs=-1)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logit_cv = LogisticRegressionCV(Cs=10, n_jobs=-1)\n",
    "logit_cv.fit(Xtr_tfidf, ytr_quantile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "db7aba34",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9847146739130435\n",
      "0.6275752773375595\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.83      0.75      0.79       439\n",
      "           1       0.45      0.50      0.48       383\n",
      "           2       0.61      0.61      0.61       440\n",
      "\n",
      "    accuracy                           0.63      1262\n",
      "   macro avg       0.63      0.62      0.63      1262\n",
      "weighted avg       0.64      0.63      0.63      1262\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(accuracy_score(logit_cv.predict(Xtr_tfidf), ytr_quantile))\n",
    "print(accuracy_score(logit_cv.predict(Xval_tfidf), yval_quantile))\n",
    "\n",
    "print(classification_report(logit_cv.predict(Xval_tfidf), yval_quantile))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "9839e35b",
   "metadata": {},
   "outputs": [],
   "source": [
    "def visualize_coefficients(coef, feature_names, title, plot_num, n_top_features=25):\n",
    "    interesting_coefficients = np.argsort(coef)[-n_top_features:]\n",
    "    \n",
    "    plt.figure(figsize=(15, 15))\n",
    "    plt.subplot(510+plot_num+1)\n",
    "    colors = ['lightblue' for c in coef[interesting_coefficients]]\n",
    "    plt.bar(np.arange(n_top_features), coef[interesting_coefficients], color=colors)\n",
    "    feature_names = np.array(feature_names)\n",
    "    plt.xticks(np.arange(1, 1 + n_top_features),\n",
    "               feature_names[interesting_coefficients], rotation=60, ha='right', fontsize=14);\n",
    "    plt.title(title)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "3225d39b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, 38513)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logit_sgd.coef_.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d86107e9",
   "metadata": {},
   "source": [
    "На графиках отображены top 25 слов, которые (по мнению логистической регрессии), наиболее сильно увеличивают \n",
    "вроятность попадания в соответствующий класс популярности"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "830e1a6a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "titles = [\"Very unpopular\", 'Average', 'Very popular'] # \"Unpopular\",'Popular', \n",
    "for i in range(3):\n",
    "    visualize_coefficients(logit_sgd.coef_[i,:], tf_vectorizer.get_feature_names(),\n",
    "                      title=titles[i], plot_num=i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "5b65bbff",
   "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": [
    "plt.hist(np.max(logit_sgd.predict_proba(Xval_tfidf), axis=1));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "d4af349f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.00306058, 0.48962858, 0.50731084],\n",
       "       [0.47211532, 0.27780688, 0.2500778 ],\n",
       "       [0.71323172, 0.18345106, 0.10331722],\n",
       "       ...,\n",
       "       [0.57889472, 0.29564875, 0.12545653],\n",
       "       [0.34525323, 0.41686866, 0.2378781 ],\n",
       "       [0.37461185, 0.38739071, 0.23799745]])"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logit_sgd.predict_proba(Xval_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "28af57b7",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = logit_sgd.predict(Xval_tfidf)\n",
    "certainty_mask = np.max(logit_sgd.predict_proba(Xval_tfidf), axis=1)>.6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "001122db",
   "metadata": {},
   "outputs": [],
   "source": [
    "meme_slice = memes.loc[Xval.index[certainty_mask]]\n",
    "meme_slice['actual_popularity'] = le.inverse_transform(predictions[certainty_mask]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "6a2fc1da",
   "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>origin_year</th>\n",
       "      <th>date_added</th>\n",
       "      <th>views</th>\n",
       "      <th>description</th>\n",
       "      <th>average_views</th>\n",
       "      <th>actual_popularity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1377</th>\n",
       "      <td>TheLegend27</td>\n",
       "      <td>2016</td>\n",
       "      <td>2019-10-07T05:37:23-04:00</td>\n",
       "      <td>1284582</td>\n",
       "      <td>video, ad, game war, viral, flater, grandayy, ...</td>\n",
       "      <td>42236.416865</td>\n",
       "      <td>(58.965, 536753.118]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3360</th>\n",
       "      <td>Zyzz</td>\n",
       "      <td>2010</td>\n",
       "      <td>2019-08-21T10:45:21-04:00</td>\n",
       "      <td>917088</td>\n",
       "      <td>bodybuilding, fitness, aesthetics, zyzz, good ...</td>\n",
       "      <td>11879.344400</td>\n",
       "      <td>(58.965, 536753.118]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1659</th>\n",
       "      <td>Yes - Roundabout / To Be Continued</td>\n",
       "      <td>2016</td>\n",
       "      <td>2019-07-10T16:40:07-04:00</td>\n",
       "      <td>910219</td>\n",
       "      <td>jjba, jojo bizzar adventure, exploitable, vine...</td>\n",
       "      <td>7651.866401</td>\n",
       "      <td>(58.965, 536753.118]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3790</th>\n",
       "      <td>He-Man Sings / HEYYEYAAEYAAAEYAEYAA</td>\n",
       "      <td>2005</td>\n",
       "      <td>2019-01-30T09:08:57-05:00</td>\n",
       "      <td>883569</td>\n",
       "      <td>roll, rick roll, he-man, superhero, heman, hey...</td>\n",
       "      <td>3153.064255</td>\n",
       "      <td>(58.965, 536753.118]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1491</th>\n",
       "      <td>Stefán Karl Stefánsson / Robbie Rotten</td>\n",
       "      <td>2016</td>\n",
       "      <td>2019-07-10T11:28:40-04:00</td>\n",
       "      <td>693765</td>\n",
       "      <td>robbi rotten, lazi town, irl, stefán karl stef...</td>\n",
       "      <td>5821.634568</td>\n",
       "      <td>(58.965, 536753.118]</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1477</th>\n",
       "      <td>#Tierry</td>\n",
       "      <td>2016</td>\n",
       "      <td>2016-10-18T14:17:54-04:00</td>\n",
       "      <td>165</td>\n",
       "      <td>badass, cool, hip, swag, yolo, tierri tierry. ...</td>\n",
       "      <td>0.148108</td>\n",
       "      <td>(0.096, 7.114]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>949</th>\n",
       "      <td>Paul Zante</td>\n",
       "      <td>2015</td>\n",
       "      <td>2017-08-29T18:45:04-04:00</td>\n",
       "      <td>158</td>\n",
       "      <td>donald love billi jone admir paul david zant ...</td>\n",
       "      <td>0.197780</td>\n",
       "      <td>(0.096, 7.114]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1022</th>\n",
       "      <td>Bob gets jolted</td>\n",
       "      <td>0</td>\n",
       "      <td>2018-08-24T22:22:18-04:00</td>\n",
       "      <td>157</td>\n",
       "      <td>funny, electric, shocking, sarah, bob novemb 2...</td>\n",
       "      <td>0.357862</td>\n",
       "      <td>(0.096, 7.114]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>Wesley Ford (Also known as Wesley_Ford, or u/W...</td>\n",
       "      <td>2019</td>\n",
       "      <td>2019-10-20T12:12:23-04:00</td>\n",
       "      <td>146</td>\n",
       "      <td>troll, reddit, hypixel, minecraft wesley ford ...</td>\n",
       "      <td>8.518196</td>\n",
       "      <td>(58.965, 536753.118]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1173</th>\n",
       "      <td>JackJoseva</td>\n",
       "      <td>2017</td>\n",
       "      <td>2017-04-13T11:53:52-04:00</td>\n",
       "      <td>145</td>\n",
       "      <td>csgo, counter strike, global offens user call ...</td>\n",
       "      <td>0.154724</td>\n",
       "      <td>(0.096, 7.114]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>642 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   name origin_year  \\\n",
       "1377                                        TheLegend27        2016   \n",
       "3360                                               Zyzz        2010   \n",
       "1659                 Yes - Roundabout / To Be Continued        2016   \n",
       "3790                He-Man Sings / HEYYEYAAEYAAAEYAEYAA        2005   \n",
       "1491             Stefán Karl Stefánsson / Robbie Rotten        2016   \n",
       "...                                                 ...         ...   \n",
       "1477                                            #Tierry        2016   \n",
       "949                                          Paul Zante        2015   \n",
       "1022                                    Bob gets jolted           0   \n",
       "33    Wesley Ford (Also known as Wesley_Ford, or u/W...        2019   \n",
       "1173                                         JackJoseva        2017   \n",
       "\n",
       "                     date_added    views  \\\n",
       "1377  2019-10-07T05:37:23-04:00  1284582   \n",
       "3360  2019-08-21T10:45:21-04:00   917088   \n",
       "1659  2019-07-10T16:40:07-04:00   910219   \n",
       "3790  2019-01-30T09:08:57-05:00   883569   \n",
       "1491  2019-07-10T11:28:40-04:00   693765   \n",
       "...                         ...      ...   \n",
       "1477  2016-10-18T14:17:54-04:00      165   \n",
       "949   2017-08-29T18:45:04-04:00      158   \n",
       "1022  2018-08-24T22:22:18-04:00      157   \n",
       "33    2019-10-20T12:12:23-04:00      146   \n",
       "1173  2017-04-13T11:53:52-04:00      145   \n",
       "\n",
       "                                            description  average_views  \\\n",
       "1377  video, ad, game war, viral, flater, grandayy, ...   42236.416865   \n",
       "3360  bodybuilding, fitness, aesthetics, zyzz, good ...   11879.344400   \n",
       "1659  jjba, jojo bizzar adventure, exploitable, vine...    7651.866401   \n",
       "3790  roll, rick roll, he-man, superhero, heman, hey...    3153.064255   \n",
       "1491  robbi rotten, lazi town, irl, stefán karl stef...    5821.634568   \n",
       "...                                                 ...            ...   \n",
       "1477  badass, cool, hip, swag, yolo, tierri tierry. ...       0.148108   \n",
       "949    donald love billi jone admir paul david zant ...       0.197780   \n",
       "1022  funny, electric, shocking, sarah, bob novemb 2...       0.357862   \n",
       "33    troll, reddit, hypixel, minecraft wesley ford ...       8.518196   \n",
       "1173  csgo, counter strike, global offens user call ...       0.154724   \n",
       "\n",
       "         actual_popularity  \n",
       "1377  (58.965, 536753.118]  \n",
       "3360  (58.965, 536753.118]  \n",
       "1659  (58.965, 536753.118]  \n",
       "3790  (58.965, 536753.118]  \n",
       "1491  (58.965, 536753.118]  \n",
       "...                    ...  \n",
       "1477        (0.096, 7.114]  \n",
       "949         (0.096, 7.114]  \n",
       "1022        (0.096, 7.114]  \n",
       "33    (58.965, 536753.118]  \n",
       "1173        (0.096, 7.114]  \n",
       "\n",
       "[642 rows x 7 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "meme_slice[\n",
    "    ['name', 'origin_year', 'date_added', 'views', 'description', 'average_views', 'actual_popularity'] # \n",
    "    \n",
    "].sort_values(by='views', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24424c96",
   "metadata": {},
   "source": [
    "### А теперь можно взглянуть на confision matrix и посмотреть, где же косячат модели"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "13469a89",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion(predictions, yval, title):\n",
    "    array = confusion_matrix(yval, predictions)\n",
    "    df_cm = pd.DataFrame(array, index=[i for i in titles],\n",
    "                         columns=[i for i in titles])\n",
    "    plt.figure(figsize=(7,7))\n",
    "    sns.heatmap(df_cm, annot=True, fmt='g', cbar=False)\n",
    "    plt.title(title);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "83dbd00a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion(logit_sgd.predict(Xval_tfidf), yval_quantile, 'logit_regression')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "39ae9f49",
   "metadata": {},
   "outputs": [],
   "source": [
    "from wordcloud import WordCloud # визуализаирует облако слов"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bf83339e",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}