AS_1_Exploration.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import data_utils as du\n",
    "\n",
    "path_to_media = './data/extra_infos.json'\n",
    "path_to_train = './data/train.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alb_title</th>\n",
       "      <th>art_name</th>\n",
       "      <th>media_id</th>\n",
       "      <th>sng_title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>El Son de Cuba</td>\n",
       "      <td>Septeto Nacional De Ignacio Pineiro</td>\n",
       "      <td>213952</td>\n",
       "      <td>Maria Cristina</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Sounds from the fourth world</td>\n",
       "      <td>Calvin Russell</td>\n",
       "      <td>223014</td>\n",
       "      <td>Love stealer</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Le meilleur de Malavoi</td>\n",
       "      <td>Malavoi</td>\n",
       "      <td>237198</td>\n",
       "      <td>Caressé moin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Les titres essentiels Cuba</td>\n",
       "      <td>Soneros De Verdad</td>\n",
       "      <td>238380</td>\n",
       "      <td>A Buena Vista</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Dantzig Twist</td>\n",
       "      <td>Marquis de Sade</td>\n",
       "      <td>244881</td>\n",
       "      <td>Nacht und Nebel</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      alb_title                             art_name  \\\n",
       "0                El Son de Cuba  Septeto Nacional De Ignacio Pineiro   \n",
       "1  Sounds from the fourth world                       Calvin Russell   \n",
       "2        Le meilleur de Malavoi                              Malavoi   \n",
       "3    Les titres essentiels Cuba                    Soneros De Verdad   \n",
       "4                 Dantzig Twist                      Marquis de Sade   \n",
       "\n",
       "   media_id        sng_title  \n",
       "0    213952   Maria Cristina  \n",
       "1    223014     Love stealer  \n",
       "2    237198     Caressé moin  \n",
       "3    238380    A Buena Vista  \n",
       "4    244881  Nacht und Nebel  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "media_info = du.json_to_pandas(path_to_media)\n",
    "media_info.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>genre_id</th>\n",
       "      <th>ts_listen</th>\n",
       "      <th>media_id</th>\n",
       "      <th>album_id</th>\n",
       "      <th>context_type</th>\n",
       "      <th>release_date</th>\n",
       "      <th>platform_name</th>\n",
       "      <th>platform_family</th>\n",
       "      <th>media_duration</th>\n",
       "      <th>listen_type</th>\n",
       "      <th>user_gender</th>\n",
       "      <th>user_id</th>\n",
       "      <th>artist_id</th>\n",
       "      <th>user_age</th>\n",
       "      <th>is_listened</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>25471</td>\n",
       "      <td>1480597215</td>\n",
       "      <td>222606</td>\n",
       "      <td>41774</td>\n",
       "      <td>12</td>\n",
       "      <td>20040704</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>223</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9241</td>\n",
       "      <td>55164</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>25571</td>\n",
       "      <td>1480544735</td>\n",
       "      <td>250467</td>\n",
       "      <td>43941</td>\n",
       "      <td>0</td>\n",
       "      <td>20060301</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>171</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>16547</td>\n",
       "      <td>55830</td>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>16</td>\n",
       "      <td>1479563953</td>\n",
       "      <td>305197</td>\n",
       "      <td>48078</td>\n",
       "      <td>1</td>\n",
       "      <td>20140714</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>7665</td>\n",
       "      <td>2704</td>\n",
       "      <td>29</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>1480152098</td>\n",
       "      <td>900502</td>\n",
       "      <td>71521</td>\n",
       "      <td>0</td>\n",
       "      <td>20001030</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>240</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1580</td>\n",
       "      <td>938</td>\n",
       "      <td>30</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>7</td>\n",
       "      <td>1478368974</td>\n",
       "      <td>542335</td>\n",
       "      <td>71718</td>\n",
       "      <td>0</td>\n",
       "      <td>20080215</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>150</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1812</td>\n",
       "      <td>2939</td>\n",
       "      <td>24</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   genre_id   ts_listen  media_id  album_id  context_type  release_date  \\\n",
       "0     25471  1480597215    222606     41774            12      20040704   \n",
       "1     25571  1480544735    250467     43941             0      20060301   \n",
       "2        16  1479563953    305197     48078             1      20140714   \n",
       "3         7  1480152098    900502     71521             0      20001030   \n",
       "4         7  1478368974    542335     71718             0      20080215   \n",
       "\n",
       "   platform_name  platform_family  media_duration  listen_type  user_gender  \\\n",
       "0              1                0             223            0            0   \n",
       "1              2                1             171            0            0   \n",
       "2              2                1             149            1            1   \n",
       "3              0                0             240            0            1   \n",
       "4              0                0             150            0            1   \n",
       "\n",
       "   user_id  artist_id  user_age  is_listened  \n",
       "0     9241      55164        29            0  \n",
       "1    16547      55830        30            1  \n",
       "2     7665       2704        29            1  \n",
       "3     1580        938        30            0  \n",
       "4     1812       2939        24            1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train = pd.read_csv(path_to_train, encoding='utf-8')\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "67142"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "by_artist = train.groupby(['artist_id'], as_index=False).agg({'media_id' : 'count'})\n",
    "len(by_artist)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Є три типи колонок. Є інформація про користувача: `user_gender`, `user_age`. Є інформація про пісню: назва, альбом, виконавець, `genre_id`, `release_date`, `media_duration`. Є інформація про конкретне прослуховування: `ts_listen`, `platform_name`, `platform_family`, `listen_type`.\n",
    "\n",
    "Що звідки ми можемо витягнути? Час доби. Чи слухає користувач flow взагалі. \n",
    "\n",
    "Як можна кластеризувати користувачів? Скільки разів він слухав кожен з жанрів (порахувати кількість жанрів). В який час любить слухати музику (вечір, робочий час, вихідні). В якому контексті скільки разів він слухав музику. З якої платформи скільки разів він слухав музику. Чи використовує він Flow. І, найскладніше, за спільними музичними уподобаннями.\n",
    "\n",
    "Для останнього можна спробувати аналіз графа."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Питання 1\n",
    "Чи є розшифровка закодованних полів? (`platform_name`, `platform_family` etc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Feature idea 1 (+++)\n",
    "Можна з назви пісні, альбома, імені виконавця витащити мову, на якій ця пісня. 100% це буде важлива фіча.\n",
    "#### Feature idea 2 (+)\n",
    "Можна взяти як фічу час доби.\n",
    "#### Feature idea 3 (++++++)\n",
    "Порівнювати музику за допомогою DBPedia https://link.springer.com/chapter/10.1007%2F978-3-642-13489-0_34?LI=true\n",
    "#### Feature idea 4 (++)\n",
    "Побудувати граф по цих данних і визначати відстань від пісні до користувача.\n",
    "#### Feature idea 5 (+++++)\n",
    "Витащити з DBPedia жанри цих пісень і натренувати нейронку визначати жанр. Потім дані з останнього шару нейронки використовувати як фічі."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train = pd.read_csv(path_to_train, encoding='utf-8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2922"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[\"genre_id\"].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19918"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[\"user_id\"].unique())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "74"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[\"context_type\"].unique()) # цікаво, що це за контексти?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[\"platform_name\"].unique()) # PC, smartphone, tablet?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[\"platform_family\"].unique()) # Windows, android, iOS?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcf83855a90>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAD8CAYAAABQFVIjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcfce713320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "train[\"media_duration\"].hist(bins=100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Дуже дивно виглядає `media_duration`, наче є якісь супер аутлаєри. Може це якісь аудіо книги?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    7.558834e+06\n",
       "mean     2.311527e+02\n",
       "std      7.999463e+01\n",
       "min      0.000000e+00\n",
       "25%      1.960000e+02\n",
       "50%      2.220000e+02\n",
       "75%      2.540000e+02\n",
       "max      6.553500e+04\n",
       "Name: media_duration, dtype: float64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[\"media_duration\"].describe() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcf80b7e668>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAD8CAYAAACLrvgBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcf817ef5c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "limit = 1000\n",
    "outliers = train[train['media_duration'] < limit]\n",
    "outliers['media_duration'].hist(bins=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0          223\n",
       "1          171\n",
       "2          149\n",
       "3          240\n",
       "4          150\n",
       "5          150\n",
       "6          150\n",
       "7          150\n",
       "8          150\n",
       "9          150\n",
       "10         150\n",
       "11         150\n",
       "12         188\n",
       "13         150\n",
       "14         150\n",
       "15         150\n",
       "16         150\n",
       "17         150\n",
       "18         150\n",
       "19         150\n",
       "20         150\n",
       "21         150\n",
       "22         150\n",
       "23         150\n",
       "24         150\n",
       "25         197\n",
       "26         210\n",
       "27         188\n",
       "28         150\n",
       "29         150\n",
       "          ... \n",
       "7558804    268\n",
       "7558805    268\n",
       "7558806    268\n",
       "7558807    268\n",
       "7558808    268\n",
       "7558809    268\n",
       "7558810    268\n",
       "7558811    268\n",
       "7558812    268\n",
       "7558813    268\n",
       "7558814    268\n",
       "7558815    268\n",
       "7558816    268\n",
       "7558817    268\n",
       "7558818    268\n",
       "7558819    268\n",
       "7558820    268\n",
       "7558821    268\n",
       "7558822    268\n",
       "7558823    268\n",
       "7558824    268\n",
       "7558825    268\n",
       "7558826    268\n",
       "7558827    268\n",
       "7558828    268\n",
       "7558829    268\n",
       "7558830    974\n",
       "7558831    175\n",
       "7558832    126\n",
       "7558833    177\n",
       "Name: media_duration, dtype: int64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "outliers['media_duration']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7556062"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(outliers['media_duration']) # це явно аудіо книги"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Anatoliy\\Anaconda2\\envs\\py35\\lib\\site-packages\\ipykernel\\__main__.py:1: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcf81d94ac8>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAD8CAYAAACRkhiPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcf81c47be0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "train[train['media_duration'] > 500][train['media_duration'] < 2000][\"media_duration\"].hist(bins=200)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "genre_id                  109\n",
       "ts_listen          1478809731\n",
       "media_id             12092135\n",
       "album_id              1109711\n",
       "context_type                0\n",
       "release_date         20051231\n",
       "platform_name               1\n",
       "platform_family             0\n",
       "media_duration           2822\n",
       "listen_type                 0\n",
       "user_gender                 1\n",
       "user_id                 13389\n",
       "artist_id              191083\n",
       "user_age                   25\n",
       "is_listened                 0\n",
       "Name: 8341, dtype: int64"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# подивимося назву\n",
    "media1 = train[train['media_duration'] > 1000].iloc[0]\n",
    "media1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alb_title</th>\n",
       "      <th>art_name</th>\n",
       "      <th>media_id</th>\n",
       "      <th>sng_title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>157043</th>\n",
       "      <td>Opera in the Outback</td>\n",
       "      <td>Nessun Dorma</td>\n",
       "      <td>12092135</td>\n",
       "      <td>Nessun Dorma</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   alb_title      art_name  media_id     sng_title\n",
       "157043  Opera in the Outback  Nessun Dorma  12092135  Nessun Dorma"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "media_info[media_info[\"media_id\"] == media1[\"media_id\"]] # ні, це опера!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>alb_title</th>\n",
       "      <th>art_name</th>\n",
       "      <th>media_id</th>\n",
       "      <th>sng_title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>377284</th>\n",
       "      <td>Fear And Loathing</td>\n",
       "      <td>Luke Slater</td>\n",
       "      <td>38545391</td>\n",
       "      <td>Fear And Loathing - Part 1 (Continuous DJ Mix)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                alb_title     art_name  media_id  \\\n",
       "377284  Fear And Loathing  Luke Slater  38545391   \n",
       "\n",
       "                                             sng_title  \n",
       "377284  Fear And Loathing - Part 1 (Continuous DJ Mix)  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "media2 = train[train['media_duration'] > 3000].iloc[0]\n",
    "media_info[media_info[\"media_id\"] == media2[\"media_id\"]] # а це вже аудіокниги, страх та ненависть"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "27"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "media2[\"genre_id\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "187946"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[train[\"genre_id\"] == 27])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xcf8308ad68>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAELCAYAAAA/cjqaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/x8GemQx1zVJ2wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xcf82cf0d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "by_user = train.groupby(\"user_id\").agg({'genre_id': 'count'})\n",
    "by_user['count'] = by_user['genre_id']\n",
    "del by_user['genre_id']\n",
    "by_user.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1815"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(by_user[by_user['count'] < 20])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6240"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train[train[\"user_id\"]==1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}