data_wrangling.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "11548bdd-daab-4b3f-a9b4-b2c12d508335",
   "metadata": {},
   "source": [
    "# Data wrangling"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "86e13632-82ca-4bff-9602-b3fa4e9b4989",
   "metadata": {},
   "source": [
    "### Summary\n",
    "\n",
    "To ensure the dataset was clean and suitable for analysis, several preprocessing steps were taken:\n",
    "\n",
    "#### anime_df\n",
    "- Data type conversion: Converted features to their appropriate data types, such as converting object types to strings for names and categorical data for types.\n",
    "- Handling missing values:\n",
    "  - Ratings: Dropped rows without ratings, as they are the target variable and are essential for the analysis.\n",
    "  - Genre: Filled missing genre values by referencing information from similar entries within the same series, identified through name similarities.\n",
    "  - Number of episodes: For all types except TV and ONA, null episode values were replaced with the mean episode count, as these types typically have episode numbers within 4 standard deviations.\n",
    "  - Remaining missing values: Dropped records with remaining missing values, which accounted for about 1% of the total dataset, to maintain data quality.\n",
    "- Check for duplicates: Verified that there were no duplicate records in the dataset for unique name and type pairs.\n",
    "- Check for outliers: The distributions of member count and episode count were left-skewed with a few large values. Spot checks confirmed the accuracy of these counts, as they correspond to popular and well-known anime titles.\n",
    "- Genre splitting: Since the genre column contains arrays of values, each unique genre was split into its own column with boolean values indicating the presence or absence of that genre for each anime.\n",
    "\n",
    "\n",
    "#### rating_df\n",
    "\n",
    "- Check for missing values: Confirmed that there are no missing values in the dataset.\n",
    "- Remove duplicates: Removed duplicate records based on the user_id and anime_id pair to ensure unique user-anime interactions.\n",
    "- Check rating distributions: Observed that most ratings are 6 or above. Some records had a rating of -1, indicating that the user watched the anime but did not provide a rating. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 517,
   "id": "dcaa4b22-a2b3-4453-8754-bcd05270ca71",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os\n",
    "import difflib"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42cb0b00-0538-48df-bced-b8bc13f87a5e",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Data Collection<a id='data_collection'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 518,
   "id": "0cd55f56-4271-4470-8181-29e93146d3a0",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32281</td>\n",
       "      <td>Kimi no Na wa.</td>\n",
       "      <td>Drama, Romance, School, Supernatural</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>9.37</td>\n",
       "      <td>200630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5114</td>\n",
       "      <td>Fullmetal Alchemist: Brotherhood</td>\n",
       "      <td>Action, Adventure, Drama, Fantasy, Magic, Mili...</td>\n",
       "      <td>TV</td>\n",
       "      <td>64</td>\n",
       "      <td>9.26</td>\n",
       "      <td>793665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>28977</td>\n",
       "      <td>Gintama°</td>\n",
       "      <td>Action, Comedy, Historical, Parody, Samurai, S...</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.25</td>\n",
       "      <td>114262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9253</td>\n",
       "      <td>Steins;Gate</td>\n",
       "      <td>Sci-Fi, Thriller</td>\n",
       "      <td>TV</td>\n",
       "      <td>24</td>\n",
       "      <td>9.17</td>\n",
       "      <td>673572</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9969</td>\n",
       "      <td>Gintama&amp;#039;</td>\n",
       "      <td>Action, Comedy, Historical, Parody, Samurai, S...</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.16</td>\n",
       "      <td>151266</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   anime_id                              name  \\\n",
       "0     32281                    Kimi no Na wa.   \n",
       "1      5114  Fullmetal Alchemist: Brotherhood   \n",
       "2     28977                          Gintama°   \n",
       "3      9253                       Steins;Gate   \n",
       "4      9969                     Gintama&#039;   \n",
       "\n",
       "                                               genre   type episodes  rating  \\\n",
       "0               Drama, Romance, School, Supernatural  Movie        1    9.37   \n",
       "1  Action, Adventure, Drama, Fantasy, Magic, Mili...     TV       64    9.26   \n",
       "2  Action, Comedy, Historical, Parody, Samurai, S...     TV       51    9.25   \n",
       "3                                   Sci-Fi, Thriller     TV       24    9.17   \n",
       "4  Action, Comedy, Historical, Parody, Samurai, S...     TV       51    9.16   \n",
       "\n",
       "   members  \n",
       "0   200630  \n",
       "1   793665  \n",
       "2   114262  \n",
       "3   673572  \n",
       "4   151266  "
      ]
     },
     "execution_count": 518,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df = pd.read_csv('data/anime.csv')\n",
    "anime_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08365416-7fed-41f7-a7b6-4a82177895b1",
   "metadata": {},
   "source": [
    "- anime_id - myanimelist.net's unique id identifying an anime.\n",
    "- name - full name of anime.\n",
    "- genre - comma separated list of genres for this anime.\n",
    "- type - movie, TV, OVA, etc.\n",
    "- episodes - how many episodes in this show. (1 if movie).\n",
    "- rating - average rating out of 10 for this anime.\n",
    "- members - number of community members that are in this anime's \"group\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 519,
   "id": "549f9598-ec8e-417e-8de2-45c81a67e257",
   "metadata": {
    "tags": []
   },
   "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>user_id</th>\n",
       "      <th>anime_id</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>20</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>24</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>79</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>226</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>241</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  anime_id  rating\n",
       "0        1        20      -1\n",
       "1        1        24      -1\n",
       "2        1        79      -1\n",
       "3        1       226      -1\n",
       "4        1       241      -1"
      ]
     },
     "execution_count": 519,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rating_df = pd.read_csv('data/rating.csv')\n",
    "rating_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de8b4ec1-b056-445a-846f-f032eaa6c656",
   "metadata": {},
   "source": [
    "- user_id - non identifiable randomly generated user id.\n",
    "- anime_id - the anime that this user has rated.\n",
    "- rating - rating out of 10 this user has assigned (-1 if the user watched it but didn't assign a rating)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f21f594b-f7fd-4383-b685-ad6cbf513e9f",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Data Cleaning<a id='data_cleaning'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 520,
   "id": "c8f75622-fff3-434d-9e18-6da2d426a634",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 12294 entries, 0 to 12293\n",
      "Data columns (total 7 columns):\n",
      " #   Column    Non-Null Count  Dtype  \n",
      "---  ------    --------------  -----  \n",
      " 0   anime_id  12294 non-null  int64  \n",
      " 1   name      12294 non-null  object \n",
      " 2   genre     12232 non-null  object \n",
      " 3   type      12269 non-null  object \n",
      " 4   episodes  12294 non-null  object \n",
      " 5   rating    12064 non-null  float64\n",
      " 6   members   12294 non-null  int64  \n",
      "dtypes: float64(1), int64(2), object(4)\n",
      "memory usage: 672.5+ KB\n"
     ]
    }
   ],
   "source": [
    "anime_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 521,
   "id": "e595e6f8-a28f-4162-802e-03287352d0ce",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id      0\n",
       "name          0\n",
       "genre        62\n",
       "type         25\n",
       "episodes      0\n",
       "rating      230\n",
       "members       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 521,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 522,
   "id": "de0a82ba-e520-43c7-8b40-50ccde8a3950",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id    False\n",
       "name        False\n",
       "genre       False\n",
       "type        False\n",
       "episodes     True\n",
       "rating      False\n",
       "members     False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 522,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(anime_df== 'Unknown').any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 523,
   "id": "7973488b-0978-454c-8126-57e07649eef4",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id    False\n",
       "name        False\n",
       "genre       False\n",
       "type        False\n",
       "episodes    False\n",
       "rating      False\n",
       "members     False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 523,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Replace string unknown with NaN type\n",
    "anime_df.replace('Unknown', np.NaN, inplace=True)\n",
    "(anime_df== 'Unknown').any()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b45c6f9d-1cb3-4959-a875-d8e1deea99e9",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Drop rows with missing `rating` (which is our target)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 524,
   "id": "f52a200d-bbb2-4dd8-b1a5-093d48217848",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "anime_df = anime_df[~anime_df['rating'].isnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 525,
   "id": "59781f08-199a-47a4-91d1-257d79cb1f72",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id      0\n",
       "name          0\n",
       "genre        47\n",
       "type          0\n",
       "episodes    188\n",
       "rating        0\n",
       "members       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 525,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7af7e0f9-5e7d-47bd-8af1-ea43011ea70a",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Fill missing value for `genre` [anime_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 526,
   "id": "79e61d85-1e6b-4c09-a205-f82cf624a145",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/xk/5gx1tq55563djqn2f91dfp040000gn/T/ipykernel_50884/1998551640.py:6: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  anime_has_null_genre['matched_name'] = anime_has_null_genre['name'].apply(lambda x: difflib.get_close_matches(x, names_without_null_genre, n=1))\n"
     ]
    },
    {
     "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "      <th>matched_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2844</th>\n",
       "      <td>33242</td>\n",
       "      <td>IS: Infinite Stratos 2 - Infinite Wedding</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Special</td>\n",
       "      <td>1</td>\n",
       "      <td>7.15</td>\n",
       "      <td>6604</td>\n",
       "      <td>[IS: Infinite Stratos 2]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3541</th>\n",
       "      <td>33589</td>\n",
       "      <td>ViVid Strike!</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TV</td>\n",
       "      <td>12</td>\n",
       "      <td>6.96</td>\n",
       "      <td>12345</td>\n",
       "      <td>[]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6040</th>\n",
       "      <td>29765</td>\n",
       "      <td>Metropolis (2009)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>6.27</td>\n",
       "      <td>313</td>\n",
       "      <td>[Metropolis]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6646</th>\n",
       "      <td>32695</td>\n",
       "      <td>Match Shoujo</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ONA</td>\n",
       "      <td>1</td>\n",
       "      <td>6.02</td>\n",
       "      <td>242</td>\n",
       "      <td>[Match Uri no Shoujo]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7018</th>\n",
       "      <td>33187</td>\n",
       "      <td>Katsudou Shashin</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>5.79</td>\n",
       "      <td>607</td>\n",
       "      <td>[Sakura Taisen: Katsudou Shashin]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      anime_id                                       name genre     type  \\\n",
       "2844     33242  IS: Infinite Stratos 2 - Infinite Wedding   NaN  Special   \n",
       "3541     33589                              ViVid Strike!   NaN       TV   \n",
       "6040     29765                          Metropolis (2009)   NaN    Movie   \n",
       "6646     32695                               Match Shoujo   NaN      ONA   \n",
       "7018     33187                           Katsudou Shashin   NaN    Movie   \n",
       "\n",
       "     episodes  rating  members                       matched_name  \n",
       "2844        1    7.15     6604           [IS: Infinite Stratos 2]  \n",
       "3541       12    6.96    12345                                 []  \n",
       "6040        1    6.27      313                       [Metropolis]  \n",
       "6646        1    6.02      242              [Match Uri no Shoujo]  \n",
       "7018        1    5.79      607  [Sakura Taisen: Katsudou Shashin]  "
      ]
     },
     "execution_count": 526,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check if anime that has genre null is from the same anime but different episode title or type. \n",
    "# Check by finding if title has similarity to other animes. \n",
    "anime_has_null_genre = anime_df[anime_df['genre'].isnull()]\n",
    "names_without_null_genre = anime_df[~anime_df['genre'].isnull()]['name']\n",
    "\n",
    "anime_has_null_genre['matched_name'] = anime_has_null_genre['name'].apply(lambda x: difflib.get_close_matches(x, names_without_null_genre, n=1))\n",
    "anime_has_null_genre.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 527,
   "id": "1568a23a-7034-4afb-ba6e-b696ddeae8b5",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 47 entries, 2844 to 10863\n",
      "Data columns (total 8 columns):\n",
      " #   Column        Non-Null Count  Dtype  \n",
      "---  ------        --------------  -----  \n",
      " 0   anime_id      47 non-null     int64  \n",
      " 1   name          47 non-null     object \n",
      " 2   genre         0 non-null      object \n",
      " 3   type          47 non-null     object \n",
      " 4   episodes      46 non-null     object \n",
      " 5   rating        47 non-null     float64\n",
      " 6   members       47 non-null     int64  \n",
      " 7   matched_name  47 non-null     object \n",
      "dtypes: float64(1), int64(2), object(5)\n",
      "memory usage: 3.3+ KB\n"
     ]
    }
   ],
   "source": [
    "anime_has_null_genre.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 528,
   "id": "ae4be1b2-cc1b-4744-96dc-b5977d0d157c",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "      <th>matched_name</th>\n",
       "      <th>matched_genre</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2844</th>\n",
       "      <td>33242</td>\n",
       "      <td>IS: Infinite Stratos 2 - Infinite Wedding</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Special</td>\n",
       "      <td>1</td>\n",
       "      <td>7.15</td>\n",
       "      <td>6604</td>\n",
       "      <td>IS: Infinite Stratos 2</td>\n",
       "      <td>Action, Comedy, Harem, Mecha, Sci-Fi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3541</th>\n",
       "      <td>33589</td>\n",
       "      <td>ViVid Strike!</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TV</td>\n",
       "      <td>12</td>\n",
       "      <td>6.96</td>\n",
       "      <td>12345</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6040</th>\n",
       "      <td>29765</td>\n",
       "      <td>Metropolis (2009)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>6.27</td>\n",
       "      <td>313</td>\n",
       "      <td>Metropolis</td>\n",
       "      <td>Adventure, Drama, Police, Romance, Sci-Fi, Sho...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6646</th>\n",
       "      <td>32695</td>\n",
       "      <td>Match Shoujo</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ONA</td>\n",
       "      <td>1</td>\n",
       "      <td>6.02</td>\n",
       "      <td>242</td>\n",
       "      <td>Match Uri no Shoujo</td>\n",
       "      <td>Drama, Kids</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7018</th>\n",
       "      <td>33187</td>\n",
       "      <td>Katsudou Shashin</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>5.79</td>\n",
       "      <td>607</td>\n",
       "      <td>Sakura Taisen: Katsudou Shashin</td>\n",
       "      <td>Adventure, Mecha, Sci-Fi, Shounen</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      anime_id                                       name genre     type  \\\n",
       "2844     33242  IS: Infinite Stratos 2 - Infinite Wedding   NaN  Special   \n",
       "3541     33589                              ViVid Strike!   NaN       TV   \n",
       "6040     29765                          Metropolis (2009)   NaN    Movie   \n",
       "6646     32695                               Match Shoujo   NaN      ONA   \n",
       "7018     33187                           Katsudou Shashin   NaN    Movie   \n",
       "\n",
       "     episodes  rating  members                     matched_name  \\\n",
       "2844        1    7.15     6604           IS: Infinite Stratos 2   \n",
       "3541       12    6.96    12345                              NaN   \n",
       "6040        1    6.27      313                       Metropolis   \n",
       "6646        1    6.02      242              Match Uri no Shoujo   \n",
       "7018        1    5.79      607  Sakura Taisen: Katsudou Shashin   \n",
       "\n",
       "                                          matched_genre  \n",
       "2844               Action, Comedy, Harem, Mecha, Sci-Fi  \n",
       "3541                                                NaN  \n",
       "6040  Adventure, Drama, Police, Romance, Sci-Fi, Sho...  \n",
       "6646                                        Drama, Kids  \n",
       "7018                  Adventure, Mecha, Sci-Fi, Shounen  "
      ]
     },
     "execution_count": 528,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fill the null genre with the matched anime's genre\n",
    "df_has_null_genre = anime_has_null_genre.copy()\n",
    "df_has_null_genre['matched_name'] = anime_has_null_genre['matched_name'].apply(lambda x: x[0] if len(x) > 0 else np.NaN)\n",
    "df_has_null_genre['matched_genre'] = df_has_null_genre['matched_name'].apply(lambda x: anime_df[anime_df['name']==x]['genre'].values)\n",
    "df_matched_genre = df_has_null_genre.copy()\n",
    "df_matched_genre['matched_genre'] = df_has_null_genre['matched_genre'].apply(lambda x: x[0] if len(x) > 0 else np.NaN)\n",
    "df_matched_genre.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 529,
   "id": "04f91b73-48fd-4cf2-a6d5-4da0870796c3",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "genre null count before:  47\n",
      "genre null count after:  10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/xk/5gx1tq55563djqn2f91dfp040000gn/T/ipykernel_50884/2974107618.py:4: FutureWarning: A value is trying to be set on a copy of a DataFrame or Series through chained assignment using an inplace method.\n",
      "The behavior will change in pandas 3.0. This inplace method will never work because the intermediate object on which we are setting values always behaves as a copy.\n",
      "\n",
      "For example, when doing 'df[col].method(value, inplace=True)', try using 'df.method({col: value}, inplace=True)' or df[col] = df[col].method(value) instead, to perform the operation inplace on the original object.\n",
      "\n",
      "\n",
      "  anime_df['genre'].fillna(df_matched_genre['matched_genre'], inplace=True)\n"
     ]
    }
   ],
   "source": [
    "# Apply the filled data to anime_df\n",
    "print('genre null count before: ', anime_df['genre'].isna().sum())\n",
    "df_matched_genre[['anime_id', 'matched_genre']]\n",
    "anime_df['genre'].fillna(df_matched_genre['matched_genre'], inplace=True)\n",
    "print('genre null count after: ', anime_df['genre'].isna().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 530,
   "id": "8a943f26-8a82-452d-9cd9-a706649e8d40",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id      0\n",
       "name          0\n",
       "genre        10\n",
       "type          0\n",
       "episodes    188\n",
       "rating        0\n",
       "members       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 530,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c15bc63d-e292-4482-8682-faafd767f00c",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Fill missing value for `episodes` [anime_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 531,
   "id": "59dc28b7-b935-4daa-8c3b-f532aea46b39",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing episodes types: ['TV' 'Special' 'ONA' 'Music' 'Movie' 'OVA']\n"
     ]
    },
    {
     "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>mean</th>\n",
       "      <th>median</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Movie</th>\n",
       "      <td>1.104530</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.171605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Music</th>\n",
       "      <td>1.131417</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.293852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ONA</th>\n",
       "      <td>6.862295</td>\n",
       "      <td>2.0</td>\n",
       "      <td>11.408264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OVA</th>\n",
       "      <td>2.422236</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.214364</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Special</th>\n",
       "      <td>2.566026</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.704655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TV</th>\n",
       "      <td>35.969468</td>\n",
       "      <td>24.0</td>\n",
       "      <td>80.722257</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              mean  median        std\n",
       "type                                 \n",
       "Movie     1.104530     1.0   2.171605\n",
       "Music     1.131417     1.0   1.293852\n",
       "ONA       6.862295     2.0  11.408264\n",
       "OVA       2.422236     2.0   3.214364\n",
       "Special   2.566026     1.0   3.704655\n",
       "TV       35.969468    24.0  80.722257"
      ]
     },
     "execution_count": 531,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The hypothesis is that certain types will exhibit specific episode counts. For instance, movies typically have just one episode. \n",
    "# If we verify that there is a \"typical\" number of episodes associated with each type and this number shows little variation, \n",
    "# we will use it to impute missing episode values based on the type.\n",
    "anime_df['episodes'] = pd.to_numeric(anime_df['episodes'], errors='coerce')\n",
    "\n",
    "print('Missing episodes types:', anime_df[anime_df['episodes'].isna()]['type'].unique())\n",
    "mean = anime_df[['type', 'episodes']].groupby('type').mean().rename(columns = {'episodes': 'mean'})\n",
    "median = anime_df[['type', 'episodes']].groupby('type').median().rename(columns = {'episodes': 'median'})\n",
    "std = anime_df[['type', 'episodes']].groupby('type').std().rename(columns = {'episodes': 'std'})\n",
    "pd.concat((mean, median, std), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 532,
   "id": "17c06744-2182-40a4-a9f9-52718d67e6e2",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "anime_df[['type', 'episodes']].groupby('type').boxplot(subplots=False, rot=45, fontsize=12)\n",
    "plt.yscale('log')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f1e2550-c81d-4583-a9b0-406e1ea8e4d9",
   "metadata": {
    "tags": []
   },
   "source": [
    "All types except `TV` and `ONA` has less than 4 Standard Deviation for episode numbers. Will replace the null episode value with the rounded mean values for all types except `TV` and `ONA`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 533,
   "id": "7d9ee666-14ad-40cf-94a9-6878f7484a9a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "avg_episodes_by_type = anime_df.groupby('type')['episodes'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 534,
   "id": "40880aaf-ffec-449f-b1c8-f9a36f7f6d94",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "anime_df.loc[(anime_df['type'] == 'Movie'), 'episodes'] = anime_df.loc[(anime_df['type'] == 'Movie'), 'episodes'].fillna(value=round(avg_episodes_by_type['Movie']))\n",
    "anime_df.loc[(anime_df['type'] == 'Music'), 'episodes'] = anime_df.loc[(anime_df['type'] == 'Music'), 'episodes'].fillna(value=round(avg_episodes_by_type['Music']))\n",
    "anime_df.loc[(anime_df['type'] == 'OVA'), 'episodes'] = anime_df.loc[(anime_df['type'] == 'OVA'), 'episodes'].fillna(value=round(avg_episodes_by_type['OVA']))\n",
    "anime_df.loc[(anime_df['type'] == 'Special'), 'episodes'] = anime_df.loc[(anime_df['type'] == 'Special'), 'episodes'].fillna(value=round(avg_episodes_by_type['Special']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 535,
   "id": "f82b44dd-88fe-4bb2-be50-92b01ca71a30",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id      0\n",
       "name          0\n",
       "genre        10\n",
       "type          0\n",
       "episodes    143\n",
       "rating        0\n",
       "members       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 535,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ada5c8a4-f039-49d6-ab96-9c5ee0058dfa",
   "metadata": {
    "tags": []
   },
   "source": [
    "**Decreased missing counts of episodes from 188 to 143.**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17b9560d-92d7-4526-aa90-01d70b753d6a",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Handle the rest of the missing values [anime_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 536,
   "id": "05d483d9-edd4-425e-b05c-641639c0b041",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>21</td>\n",
       "      <td>One Piece</td>\n",
       "      <td>Action, Adventure, Comedy, Drama, Fantasy, Sho...</td>\n",
       "      <td>TV</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.58</td>\n",
       "      <td>504862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>252</th>\n",
       "      <td>235</td>\n",
       "      <td>Detective Conan</td>\n",
       "      <td>Adventure, Comedy, Mystery, Police, Shounen</td>\n",
       "      <td>TV</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.25</td>\n",
       "      <td>114702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>615</th>\n",
       "      <td>1735</td>\n",
       "      <td>Naruto: Shippuuden</td>\n",
       "      <td>Action, Comedy, Martial Arts, Shounen, Super P...</td>\n",
       "      <td>TV</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.94</td>\n",
       "      <td>533578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>991</th>\n",
       "      <td>966</td>\n",
       "      <td>Crayon Shin-chan</td>\n",
       "      <td>Comedy, Ecchi, Kids, School, Shounen, Slice of...</td>\n",
       "      <td>TV</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.73</td>\n",
       "      <td>26267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1272</th>\n",
       "      <td>21639</td>\n",
       "      <td>Yu☆Gi☆Oh! Arc-V</td>\n",
       "      <td>Action, Fantasy, Game, Shounen</td>\n",
       "      <td>TV</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.61</td>\n",
       "      <td>17571</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      anime_id                name  \\\n",
       "74          21           One Piece   \n",
       "252        235     Detective Conan   \n",
       "615       1735  Naruto: Shippuuden   \n",
       "991        966    Crayon Shin-chan   \n",
       "1272     21639     Yu☆Gi☆Oh! Arc-V   \n",
       "\n",
       "                                                  genre type  episodes  \\\n",
       "74    Action, Adventure, Comedy, Drama, Fantasy, Sho...   TV       NaN   \n",
       "252         Adventure, Comedy, Mystery, Police, Shounen   TV       NaN   \n",
       "615   Action, Comedy, Martial Arts, Shounen, Super P...   TV       NaN   \n",
       "991   Comedy, Ecchi, Kids, School, Shounen, Slice of...   TV       NaN   \n",
       "1272                     Action, Fantasy, Game, Shounen   TV       NaN   \n",
       "\n",
       "      rating  members  \n",
       "74      8.58   504862  \n",
       "252     8.25   114702  \n",
       "615     7.94   533578  \n",
       "991     7.73    26267  \n",
       "1272    7.61    17571  "
      ]
     },
     "execution_count": 536,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df[anime_df['episodes'].isna()].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "78f7195a-d190-4fff-9830-cce3d69dc180",
   "metadata": {},
   "source": [
    "It appears that when the episode count is missing for a TV anime, the series is still ongoing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 537,
   "id": "454c7154-6542-45ad-889b-50213a0eae0f",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3541</th>\n",
       "      <td>33589</td>\n",
       "      <td>ViVid Strike!</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TV</td>\n",
       "      <td>12.0</td>\n",
       "      <td>6.96</td>\n",
       "      <td>12345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7198</th>\n",
       "      <td>30862</td>\n",
       "      <td>Yubi wo Nusunda Onna</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.65</td>\n",
       "      <td>223</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7349</th>\n",
       "      <td>19219</td>\n",
       "      <td>Modern No.2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.52</td>\n",
       "      <td>374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7685</th>\n",
       "      <td>31760</td>\n",
       "      <td>Tsuru Shitae Waka Kan</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.15</td>\n",
       "      <td>195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7824</th>\n",
       "      <td>28655</td>\n",
       "      <td>PiKA PiKA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.92</td>\n",
       "      <td>289</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      anime_id                   name genre   type  episodes  rating  members\n",
       "3541     33589          ViVid Strike!   NaN     TV      12.0    6.96    12345\n",
       "7198     30862   Yubi wo Nusunda Onna   NaN  Movie       1.0    5.65      223\n",
       "7349     19219            Modern No.2   NaN  Movie       1.0    5.52      374\n",
       "7685     31760  Tsuru Shitae Waka Kan   NaN  Movie       1.0    5.15      195\n",
       "7824     28655              PiKA PiKA   NaN  Movie       1.0    4.92      289"
      ]
     },
     "execution_count": 537,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df[anime_df['genre'].isna()].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 538,
   "id": "df16114e-57e3-4e82-a59f-ca30e15f291c",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.01268236074270557"
      ]
     },
     "execution_count": 538,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(anime_df[(anime_df['genre'].isna()) | (anime_df['episodes'].isna())]['anime_id'].count())/ (anime_df['anime_id'].count())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a31f98c0-9aec-4e19-8186-1597e3bd1edb",
   "metadata": {
    "tags": []
   },
   "source": [
    "The rest of the missing value is about 1% of the total data. Will drop the records."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 539,
   "id": "d8def72f-2ce3-4b38-ac57-10b2469aa495",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id    0\n",
       "name        0\n",
       "genre       0\n",
       "type        0\n",
       "episodes    0\n",
       "rating      0\n",
       "members     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 539,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df.dropna(inplace=True)\n",
    "anime_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "076baa58-2a70-4fa2-bcd9-17be807cf30a",
   "metadata": {},
   "source": [
    "### Convert Data type [anime_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 540,
   "id": "699322bf-4f3d-4c3c-a5e9-71a47a19e54b",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 11911 entries, 0 to 12293\n",
      "Data columns (total 7 columns):\n",
      " #   Column    Non-Null Count  Dtype   \n",
      "---  ------    --------------  -----   \n",
      " 0   anime_id  11911 non-null  int64   \n",
      " 1   name      11911 non-null  string  \n",
      " 2   genre     11911 non-null  string  \n",
      " 3   type      11911 non-null  category\n",
      " 4   episodes  11911 non-null  int64   \n",
      " 5   rating    11911 non-null  float64 \n",
      " 6   members   11911 non-null  int64   \n",
      "dtypes: category(1), float64(1), int64(3), string(2)\n",
      "memory usage: 663.2 KB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "# Convert to expected datatype\n",
    "anime_df['name'] = anime_df['name'].astype('string')\n",
    "anime_df['type'] = anime_df['type'].astype('category')\n",
    "anime_df['genre'] = anime_df['genre'].astype('string')\n",
    "anime_df['episodes'] = anime_df['episodes'].astype('int')\n",
    "print(anime_df.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "165876a3-f043-4e1a-88b3-29014c4f181f",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Check duplicates [anime_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 541,
   "id": "0fbb6116-8530-4334-b5cd-37f7105a6e66",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 541,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check duplicates\n",
    "anime_df[['name', 'type']].duplicated().any()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e54d2dc6-76a1-4c9a-aed8-5d73d4b6c9fd",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Check if any abnormal feature values [anime_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 542,
   "id": "34c240dc-9cee-4ba2-8811-ae3906b6ba28",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 11911 entries, 0 to 12293\n",
      "Data columns (total 7 columns):\n",
      " #   Column    Non-Null Count  Dtype   \n",
      "---  ------    --------------  -----   \n",
      " 0   anime_id  11911 non-null  int64   \n",
      " 1   name      11911 non-null  string  \n",
      " 2   genre     11911 non-null  string  \n",
      " 3   type      11911 non-null  category\n",
      " 4   episodes  11911 non-null  int64   \n",
      " 5   rating    11911 non-null  float64 \n",
      " 6   members   11911 non-null  int64   \n",
      "dtypes: category(1), float64(1), int64(3), string(2)\n",
      "memory usage: 663.2 KB\n"
     ]
    }
   ],
   "source": [
    "anime_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 543,
   "id": "2dc735dc-d87c-42d1-9ebe-e96f36be4b57",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check feature outliers\n",
    "fig, (ax1, ax2) = plt.subplots(1, 2, constrained_layout=True)\n",
    "anime_df['type'].value_counts().plot(kind='barh', ax=ax1, xlabel='counts')\n",
    "anime_df['rating'].plot(kind='hist', ax = ax2, bins=10, xlabel='rating', ylabel='counts')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52ba98f7-1bfb-41f1-84b7-b9fb22d014f3",
   "metadata": {
    "tags": []
   },
   "source": [
    "Most of the anime are type OVA and TV. And most of the rating are between 5 to 7."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 544,
   "id": "9dc6890c-f2a9-4b39-9ffd-410d68069fd7",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>11911.000000</td>\n",
       "      <td>11911.000000</td>\n",
       "      <td>11911.000000</td>\n",
       "      <td>1.191100e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>13526.981026</td>\n",
       "      <td>12.416506</td>\n",
       "      <td>6.481722</td>\n",
       "      <td>1.839268e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>11173.497214</td>\n",
       "      <td>46.945266</td>\n",
       "      <td>1.021110</td>\n",
       "      <td>5.520147e+04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.670000</td>\n",
       "      <td>1.200000e+01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>3364.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.890000</td>\n",
       "      <td>2.295000e+02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>9895.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.570000</td>\n",
       "      <td>1.575000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>23577.000000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>7.185000</td>\n",
       "      <td>9.694000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>34519.000000</td>\n",
       "      <td>1818.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>1.013917e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           anime_id      episodes        rating       members\n",
       "count  11911.000000  11911.000000  11911.000000  1.191100e+04\n",
       "mean   13526.981026     12.416506      6.481722  1.839268e+04\n",
       "std    11173.497214     46.945266      1.021110  5.520147e+04\n",
       "min        1.000000      1.000000      1.670000  1.200000e+01\n",
       "25%     3364.500000      1.000000      5.890000  2.295000e+02\n",
       "50%     9895.000000      2.000000      6.570000  1.575000e+03\n",
       "75%    23577.000000     12.000000      7.185000  9.694000e+03\n",
       "max    34519.000000   1818.000000     10.000000  1.013917e+06"
      ]
     },
     "execution_count": 544,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 545,
   "id": "47aa906b-b488-4fad-b2f3-880ef8bede1d",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "anime_df['episodes'].plot(kind='hist', bins=20, log= True, xlabel='number of episodes', ylabel='counts (log)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90f3b6db-e99b-4694-8ab6-74850e380f4f",
   "metadata": {
    "tags": []
   },
   "source": [
    "Most of the anime are under 250 episodes. And we have some outliers that has 500 or more episodes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 546,
   "id": "d243f350-68cb-4fa5-a7f2-282274c57e83",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "anime_df['members'].plot(kind='hist', bins=20, log=True, xlabel='members count', ylabel='counts (log)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b181b78d-6262-41be-94bb-f50b846afaa6",
   "metadata": {},
   "source": [
    "Most anime has lower number of member counts and the graph skew to the right."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 547,
   "id": "12e0df9c-2cfb-4a17-9cbd-50c3bdde7fa7",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>929</th>\n",
       "      <td>2471</td>\n",
       "      <td>Doraemon (1979)</td>\n",
       "      <td>Adventure, Comedy, Fantasy, Kids, Sci-Fi, Shounen</td>\n",
       "      <td>TV</td>\n",
       "      <td>1787</td>\n",
       "      <td>7.76</td>\n",
       "      <td>14233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5360</th>\n",
       "      <td>6277</td>\n",
       "      <td>Manga Nippon Mukashibanashi (1976)</td>\n",
       "      <td>Fantasy, Historical</td>\n",
       "      <td>TV</td>\n",
       "      <td>1471</td>\n",
       "      <td>6.48</td>\n",
       "      <td>406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6296</th>\n",
       "      <td>12393</td>\n",
       "      <td>Oyako Club</td>\n",
       "      <td>Comedy, Slice of Life</td>\n",
       "      <td>TV</td>\n",
       "      <td>1818</td>\n",
       "      <td>6.18</td>\n",
       "      <td>160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6580</th>\n",
       "      <td>8213</td>\n",
       "      <td>Hoka Hoka Kazoku</td>\n",
       "      <td>Comedy</td>\n",
       "      <td>TV</td>\n",
       "      <td>1428</td>\n",
       "      <td>6.05</td>\n",
       "      <td>194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9242</th>\n",
       "      <td>32448</td>\n",
       "      <td>Kirin Ashita no Calendar</td>\n",
       "      <td>Historical, Kids</td>\n",
       "      <td>TV</td>\n",
       "      <td>1306</td>\n",
       "      <td>6.43</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      anime_id                                name  \\\n",
       "929       2471                     Doraemon (1979)   \n",
       "5360      6277  Manga Nippon Mukashibanashi (1976)   \n",
       "6296     12393                          Oyako Club   \n",
       "6580      8213                    Hoka Hoka Kazoku   \n",
       "9242     32448            Kirin Ashita no Calendar   \n",
       "\n",
       "                                                  genre type  episodes  \\\n",
       "929   Adventure, Comedy, Fantasy, Kids, Sci-Fi, Shounen   TV      1787   \n",
       "5360                                Fantasy, Historical   TV      1471   \n",
       "6296                              Comedy, Slice of Life   TV      1818   \n",
       "6580                                             Comedy   TV      1428   \n",
       "9242                                   Historical, Kids   TV      1306   \n",
       "\n",
       "      rating  members  \n",
       "929     7.76    14233  \n",
       "5360    6.48      406  \n",
       "6296    6.18      160  \n",
       "6580    6.05      194  \n",
       "9242    6.43       59  "
      ]
     },
     "execution_count": 547,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df[anime_df['episodes'] > 1000].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "28edade3-c2f2-44ec-8dba-461708811bfb",
   "metadata": {
    "tags": []
   },
   "source": [
    "Upon spot-checking anime with over 1,000 episodes, the episode counts appear to be accurate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 548,
   "id": "cc76def2-0e8e-4bde-8c26-56540f5d033c",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>genre</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32281</td>\n",
       "      <td>Kimi no Na wa.</td>\n",
       "      <td>Drama, Romance, School, Supernatural</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>9.37</td>\n",
       "      <td>200630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5114</td>\n",
       "      <td>Fullmetal Alchemist: Brotherhood</td>\n",
       "      <td>Action, Adventure, Drama, Fantasy, Magic, Mili...</td>\n",
       "      <td>TV</td>\n",
       "      <td>64</td>\n",
       "      <td>9.26</td>\n",
       "      <td>793665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>28977</td>\n",
       "      <td>Gintama°</td>\n",
       "      <td>Action, Comedy, Historical, Parody, Samurai, S...</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.25</td>\n",
       "      <td>114262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9253</td>\n",
       "      <td>Steins;Gate</td>\n",
       "      <td>Sci-Fi, Thriller</td>\n",
       "      <td>TV</td>\n",
       "      <td>24</td>\n",
       "      <td>9.17</td>\n",
       "      <td>673572</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9969</td>\n",
       "      <td>Gintama&amp;#039;</td>\n",
       "      <td>Action, Comedy, Historical, Parody, Samurai, S...</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.16</td>\n",
       "      <td>151266</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   anime_id                              name  \\\n",
       "0     32281                    Kimi no Na wa.   \n",
       "1      5114  Fullmetal Alchemist: Brotherhood   \n",
       "2     28977                          Gintama°   \n",
       "3      9253                       Steins;Gate   \n",
       "4      9969                     Gintama&#039;   \n",
       "\n",
       "                                               genre   type  episodes  rating  \\\n",
       "0               Drama, Romance, School, Supernatural  Movie         1    9.37   \n",
       "1  Action, Adventure, Drama, Fantasy, Magic, Mili...     TV        64    9.26   \n",
       "2  Action, Comedy, Historical, Parody, Samurai, S...     TV        51    9.25   \n",
       "3                                   Sci-Fi, Thriller     TV        24    9.17   \n",
       "4  Action, Comedy, Historical, Parody, Samurai, S...     TV        51    9.16   \n",
       "\n",
       "   members  \n",
       "0   200630  \n",
       "1   793665  \n",
       "2   114262  \n",
       "3   673572  \n",
       "4   151266  "
      ]
     },
     "execution_count": 548,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df[anime_df['members'] > 100000].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49679cac-a972-45c2-ba1c-0ebf8649f3ad",
   "metadata": {},
   "source": [
    "Upon spot-checking anime with over 100,000 members, the number appear to be accurate as well since those are very popular and well know anime."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da4409bd-4589-4bcd-9a25-88bad098ff20",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Split each genre into columns [anime_df]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6c47ca6-88f3-43bf-9022-c506d0c9a4fc",
   "metadata": {},
   "source": [
    "Since the genre column contains arrays of values, each unique genre was split into its own column with boolean values indicating the presence or absence of that genre for each anime."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 549,
   "id": "724cd1a2-4d47-4817-915a-e6b06b3990fa",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "      <th>Action</th>\n",
       "      <th>Adventure</th>\n",
       "      <th>Cars</th>\n",
       "      <th>Comedy</th>\n",
       "      <th>...</th>\n",
       "      <th>Shounen Ai</th>\n",
       "      <th>Slice of Life</th>\n",
       "      <th>Space</th>\n",
       "      <th>Sports</th>\n",
       "      <th>Super Power</th>\n",
       "      <th>Supernatural</th>\n",
       "      <th>Thriller</th>\n",
       "      <th>Vampire</th>\n",
       "      <th>Yaoi</th>\n",
       "      <th>Yuri</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32281</td>\n",
       "      <td>Kimi no Na wa.</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>9.37</td>\n",
       "      <td>200630</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5114</td>\n",
       "      <td>Fullmetal Alchemist: Brotherhood</td>\n",
       "      <td>TV</td>\n",
       "      <td>64</td>\n",
       "      <td>9.26</td>\n",
       "      <td>793665</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>28977</td>\n",
       "      <td>Gintama°</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.25</td>\n",
       "      <td>114262</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9253</td>\n",
       "      <td>Steins;Gate</td>\n",
       "      <td>TV</td>\n",
       "      <td>24</td>\n",
       "      <td>9.17</td>\n",
       "      <td>673572</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9969</td>\n",
       "      <td>Gintama&amp;#039;</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.16</td>\n",
       "      <td>151266</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12289</th>\n",
       "      <td>9316</td>\n",
       "      <td>Toushindai My Lover: Minami tai Mecha-Minami</td>\n",
       "      <td>OVA</td>\n",
       "      <td>1</td>\n",
       "      <td>4.15</td>\n",
       "      <td>211</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12290</th>\n",
       "      <td>5543</td>\n",
       "      <td>Under World</td>\n",
       "      <td>OVA</td>\n",
       "      <td>1</td>\n",
       "      <td>4.28</td>\n",
       "      <td>183</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12291</th>\n",
       "      <td>5621</td>\n",
       "      <td>Violence Gekiga David no Hoshi</td>\n",
       "      <td>OVA</td>\n",
       "      <td>4</td>\n",
       "      <td>4.88</td>\n",
       "      <td>219</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12292</th>\n",
       "      <td>6133</td>\n",
       "      <td>Violence Gekiga Shin David no Hoshi: Inma Dens...</td>\n",
       "      <td>OVA</td>\n",
       "      <td>1</td>\n",
       "      <td>4.98</td>\n",
       "      <td>175</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12293</th>\n",
       "      <td>26081</td>\n",
       "      <td>Yasuji no Pornorama: Yacchimae!!</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>5.46</td>\n",
       "      <td>142</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11911 rows × 49 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       anime_id                                               name   type  \\\n",
       "0         32281                                     Kimi no Na wa.  Movie   \n",
       "1          5114                   Fullmetal Alchemist: Brotherhood     TV   \n",
       "2         28977                                           Gintama°     TV   \n",
       "3          9253                                        Steins;Gate     TV   \n",
       "4          9969                                      Gintama&#039;     TV   \n",
       "...         ...                                                ...    ...   \n",
       "12289      9316       Toushindai My Lover: Minami tai Mecha-Minami    OVA   \n",
       "12290      5543                                        Under World    OVA   \n",
       "12291      5621                     Violence Gekiga David no Hoshi    OVA   \n",
       "12292      6133  Violence Gekiga Shin David no Hoshi: Inma Dens...    OVA   \n",
       "12293     26081                   Yasuji no Pornorama: Yacchimae!!  Movie   \n",
       "\n",
       "       episodes  rating  members  Action  Adventure  Cars  Comedy  ...  \\\n",
       "0             1    9.37   200630       0          0     0       0  ...   \n",
       "1            64    9.26   793665       1          1     0       0  ...   \n",
       "2            51    9.25   114262       1          0     0       1  ...   \n",
       "3            24    9.17   673572       0          0     0       0  ...   \n",
       "4            51    9.16   151266       1          0     0       1  ...   \n",
       "...         ...     ...      ...     ...        ...   ...     ...  ...   \n",
       "12289         1    4.15      211       0          0     0       0  ...   \n",
       "12290         1    4.28      183       0          0     0       0  ...   \n",
       "12291         4    4.88      219       0          0     0       0  ...   \n",
       "12292         1    4.98      175       0          0     0       0  ...   \n",
       "12293         1    5.46      142       0          0     0       0  ...   \n",
       "\n",
       "       Shounen Ai  Slice of Life  Space  Sports  Super Power  Supernatural  \\\n",
       "0               0              0      0       0            0             1   \n",
       "1               0              0      0       0            0             0   \n",
       "2               0              0      0       0            0             0   \n",
       "3               0              0      0       0            0             0   \n",
       "4               0              0      0       0            0             0   \n",
       "...           ...            ...    ...     ...          ...           ...   \n",
       "12289           0              0      0       0            0             0   \n",
       "12290           0              0      0       0            0             0   \n",
       "12291           0              0      0       0            0             0   \n",
       "12292           0              0      0       0            0             0   \n",
       "12293           0              0      0       0            0             0   \n",
       "\n",
       "       Thriller  Vampire  Yaoi  Yuri  \n",
       "0             0        0     0     0  \n",
       "1             0        0     0     0  \n",
       "2             0        0     0     0  \n",
       "3             1        0     0     0  \n",
       "4             0        0     0     0  \n",
       "...         ...      ...   ...   ...  \n",
       "12289         0        0     0     0  \n",
       "12290         0        0     0     0  \n",
       "12291         0        0     0     0  \n",
       "12292         0        0     0     0  \n",
       "12293         0        0     0     0  \n",
       "\n",
       "[11911 rows x 49 columns]"
      ]
     },
     "execution_count": 549,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "genre_columns = anime_df['genre'].str.get_dummies(sep=', ')\n",
    "anime_df = pd.concat([anime_df, genre_columns], axis=1)\n",
    "anime_df = anime_df.drop('genre', axis=1)\n",
    "anime_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "06b936c0-1614-423b-8fd7-803213968689",
   "metadata": {},
   "source": [
    "## Quick look at rating_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4a1684a-ec14-46fd-9b3c-be91114a5278",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Check missing value [rating_df]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 550,
   "id": "9b4af48a-0933-4966-b146-16834480da0a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 7813737 entries, 0 to 7813736\n",
      "Data columns (total 3 columns):\n",
      " #   Column    Dtype\n",
      "---  ------    -----\n",
      " 0   user_id   int64\n",
      " 1   anime_id  int64\n",
      " 2   rating    int64\n",
      "dtypes: int64(3)\n",
      "memory usage: 178.8 MB\n",
      "None\n"
     ]
    },
    {
     "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>user_id</th>\n",
       "      <th>anime_id</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>20</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>24</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>79</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>226</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>241</td>\n",
       "      <td>-1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  anime_id  rating\n",
       "0        1        20      -1\n",
       "1        1        24      -1\n",
       "2        1        79      -1\n",
       "3        1       226      -1\n",
       "4        1       241      -1"
      ]
     },
     "execution_count": 550,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Cleaning for rating\n",
    "print(rating_df.info())\n",
    "rating_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 551,
   "id": "feef644b-74d0-41ba-96cd-d429280681f9",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id     0\n",
       "anime_id    0\n",
       "rating      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 551,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rating_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "677cabac-b361-43cd-b687-72350f609fe6",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Drop duplicates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 552,
   "id": "7f5ea4c8-251a-4ac5-818e-f46a5d9d39c0",
   "metadata": {
    "tags": []
   },
   "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>user_id</th>\n",
       "      <th>anime_id</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4499258</th>\n",
       "      <td>42653</td>\n",
       "      <td>16498</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4499316</th>\n",
       "      <td>42653</td>\n",
       "      <td>16498</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         user_id  anime_id  rating\n",
       "4499258    42653     16498       8\n",
       "4499316    42653     16498       8"
      ]
     },
     "execution_count": 552,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rating_df[rating_df.duplicated(keep=False)]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "543a389a-d67b-4f72-9f10-4bca1516eb66",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Check feature value distributions [rating_df]¶"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 553,
   "id": "4027e171-740c-465a-b081-5781f36e0db0",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rating_df['rating'].plot(kind='hist', xlabel='rating', ylabel='count')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01c34638-e6ee-4046-a5fd-81422530e258",
   "metadata": {},
   "source": [
    "A lot of the ratings are in the range of 6 or above. We also see some record as -1 as rating value. This means the user watched the anime but didn't provide a rating."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 554,
   "id": "3f22ba23-a223-4bfe-b0e5-cf07261c21fc",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1, 10,  8,  6,  9,  7,  3,  5,  4,  1,  2])"
      ]
     },
     "execution_count": 554,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rating_df['rating'].unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29b80572-2c8c-463f-8520-1649e1288744",
   "metadata": {},
   "source": [
    "## Add features to anime_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "094f7f3a-e296-42c3-9b69-d1b5e91cec9e",
   "metadata": {},
   "source": [
    "Adding the follwing features to the anime_df for the next phase of the analysis:\n",
    "- number of reviews\n",
    "- number of views without reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 555,
   "id": "3ffe1f4c-37b5-417b-90f6-4b7a0f17bcd8",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id\n",
       "1    2060\n",
       "5    1137\n",
       "6    1692\n",
       "7     460\n",
       "8     105\n",
       "Name: views_without_review, dtype: int64"
      ]
     },
     "execution_count": 555,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "view_count_per_anime = rating_df[rating_df.rating == -1].groupby('anime_id')['user_id'].count()\n",
    "view_count_per_anime = view_count_per_anime.rename('views_without_review')\n",
    "view_count_per_anime.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 556,
   "id": "a8ab79fc-24b9-4667-b1d5-18857ff658f3",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "anime_id\n",
       "1    13449\n",
       "5     5790\n",
       "6     9385\n",
       "7     2169\n",
       "8      308\n",
       "Name: rating_counts, dtype: int64"
      ]
     },
     "execution_count": 556,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rating_count_per_anime = rating_df[rating_df.rating >=0].groupby('anime_id')['user_id'].count()\n",
    "rating_count_per_anime = review_count_per_anime.rename('rating_counts')\n",
    "rating_count_per_anime.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 557,
   "id": "76f86eeb-8a48-452e-ae10-a5dd1cb3206b",
   "metadata": {
    "tags": []
   },
   "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>anime_id</th>\n",
       "      <th>name</th>\n",
       "      <th>type</th>\n",
       "      <th>episodes</th>\n",
       "      <th>rating</th>\n",
       "      <th>members</th>\n",
       "      <th>Action</th>\n",
       "      <th>Adventure</th>\n",
       "      <th>Cars</th>\n",
       "      <th>Comedy</th>\n",
       "      <th>...</th>\n",
       "      <th>Space</th>\n",
       "      <th>Sports</th>\n",
       "      <th>Super Power</th>\n",
       "      <th>Supernatural</th>\n",
       "      <th>Thriller</th>\n",
       "      <th>Vampire</th>\n",
       "      <th>Yaoi</th>\n",
       "      <th>Yuri</th>\n",
       "      <th>views_without_review</th>\n",
       "      <th>rating_counts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32281</td>\n",
       "      <td>Kimi no Na wa.</td>\n",
       "      <td>Movie</td>\n",
       "      <td>1</td>\n",
       "      <td>9.37</td>\n",
       "      <td>200630</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>238.0</td>\n",
       "      <td>1961.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5114</td>\n",
       "      <td>Fullmetal Alchemist: Brotherhood</td>\n",
       "      <td>TV</td>\n",
       "      <td>64</td>\n",
       "      <td>9.26</td>\n",
       "      <td>793665</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3080.0</td>\n",
       "      <td>21494.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>28977</td>\n",
       "      <td>Gintama°</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.25</td>\n",
       "      <td>114262</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>198.0</td>\n",
       "      <td>1188.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>9253</td>\n",
       "      <td>Steins;Gate</td>\n",
       "      <td>TV</td>\n",
       "      <td>24</td>\n",
       "      <td>9.17</td>\n",
       "      <td>673572</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2132.0</td>\n",
       "      <td>17151.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>9969</td>\n",
       "      <td>Gintama&amp;#039;</td>\n",
       "      <td>TV</td>\n",
       "      <td>51</td>\n",
       "      <td>9.16</td>\n",
       "      <td>151266</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>558.0</td>\n",
       "      <td>3115.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 51 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   anime_id                              name   type  episodes  rating  \\\n",
       "0     32281                    Kimi no Na wa.  Movie         1    9.37   \n",
       "1      5114  Fullmetal Alchemist: Brotherhood     TV        64    9.26   \n",
       "2     28977                          Gintama°     TV        51    9.25   \n",
       "3      9253                       Steins;Gate     TV        24    9.17   \n",
       "4      9969                     Gintama&#039;     TV        51    9.16   \n",
       "\n",
       "   members  Action  Adventure  Cars  Comedy  ...  Space  Sports  Super Power  \\\n",
       "0   200630       0          0     0       0  ...      0       0            0   \n",
       "1   793665       1          1     0       0  ...      0       0            0   \n",
       "2   114262       1          0     0       1  ...      0       0            0   \n",
       "3   673572       0          0     0       0  ...      0       0            0   \n",
       "4   151266       1          0     0       1  ...      0       0            0   \n",
       "\n",
       "   Supernatural  Thriller  Vampire  Yaoi  Yuri  views_without_review  \\\n",
       "0             1         0        0     0     0                 238.0   \n",
       "1             0         0        0     0     0                3080.0   \n",
       "2             0         0        0     0     0                 198.0   \n",
       "3             0         1        0     0     0                2132.0   \n",
       "4             0         0        0     0     0                 558.0   \n",
       "\n",
       "   rating_counts  \n",
       "0         1961.0  \n",
       "1        21494.0  \n",
       "2         1188.0  \n",
       "3        17151.0  \n",
       "4         3115.0  \n",
       "\n",
       "[5 rows x 51 columns]"
      ]
     },
     "execution_count": 557,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anime_df = anime_df.join(view_count_per_anime, on='anime_id', how='left')\n",
    "anime_df = anime_df.join(rating_count_per_anime, on='anime_id', how='left')\n",
    "anime_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 558,
   "id": "eb173a6b-2990-4433-992a-d7596306e305",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "anime_df.loc[:, 'views_without_review'] = anime_df.loc[:, 'views_without_review'].fillna(value=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 559,
   "id": "c49a2a94-5766-49c0-a450-8c25e0d0b085",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "rating\n",
       "2      327.750000\n",
       "3       11.282609\n",
       "4       15.436823\n",
       "5       25.164444\n",
       "6       90.466711\n",
       "7      504.675872\n",
       "8     2288.455345\n",
       "9     6089.489362\n",
       "10            NaN\n",
       "Name: rating_counts, dtype: float64"
      ]
     },
     "execution_count": 559,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean_rating_count_per_rating = anime_df.groupby(anime_df['rating'].apply(lambda x: round(x)))['rating_counts'].mean()\n",
    "mean_rating_count_per_rating"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 560,
   "id": "280c98a1-0327-4adf-bebc-8b1c06baedb6",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "for i in range(2,10):\n",
    "    anime_df.loc[round(anime_df.rating) == i, 'rating_counts'] = anime_df.loc[round(anime_df.rating) == i, 'rating_counts'].fillna(value=round(mean_rating_count_per_rating[i]))\n",
    "\n",
    "anime_df.loc[round(anime_df.rating) == 10, 'rating_counts'] = anime_df.loc[round(anime_df.rating) == 10, 'rating_counts'].fillna(value=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 562,
   "id": "560e9b0b-d221-4002-aaa4-11ac5239269c",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 11911 entries, 0 to 12293\n",
      "Data columns (total 51 columns):\n",
      " #   Column                Non-Null Count  Dtype   \n",
      "---  ------                --------------  -----   \n",
      " 0   anime_id              11911 non-null  int64   \n",
      " 1   name                  11911 non-null  string  \n",
      " 2   type                  11911 non-null  category\n",
      " 3   episodes              11911 non-null  int64   \n",
      " 4   rating                11911 non-null  float64 \n",
      " 5   members               11911 non-null  int64   \n",
      " 6   Action                11911 non-null  int64   \n",
      " 7   Adventure             11911 non-null  int64   \n",
      " 8   Cars                  11911 non-null  int64   \n",
      " 9   Comedy                11911 non-null  int64   \n",
      " 10  Dementia              11911 non-null  int64   \n",
      " 11  Demons                11911 non-null  int64   \n",
      " 12  Drama                 11911 non-null  int64   \n",
      " 13  Ecchi                 11911 non-null  int64   \n",
      " 14  Fantasy               11911 non-null  int64   \n",
      " 15  Game                  11911 non-null  int64   \n",
      " 16  Harem                 11911 non-null  int64   \n",
      " 17  Hentai                11911 non-null  int64   \n",
      " 18  Historical            11911 non-null  int64   \n",
      " 19  Horror                11911 non-null  int64   \n",
      " 20  Josei                 11911 non-null  int64   \n",
      " 21  Kids                  11911 non-null  int64   \n",
      " 22  Magic                 11911 non-null  int64   \n",
      " 23  Martial Arts          11911 non-null  int64   \n",
      " 24  Mecha                 11911 non-null  int64   \n",
      " 25  Military              11911 non-null  int64   \n",
      " 26  Music                 11911 non-null  int64   \n",
      " 27  Mystery               11911 non-null  int64   \n",
      " 28  Parody                11911 non-null  int64   \n",
      " 29  Police                11911 non-null  int64   \n",
      " 30  Psychological         11911 non-null  int64   \n",
      " 31  Romance               11911 non-null  int64   \n",
      " 32  Samurai               11911 non-null  int64   \n",
      " 33  School                11911 non-null  int64   \n",
      " 34  Sci-Fi                11911 non-null  int64   \n",
      " 35  Seinen                11911 non-null  int64   \n",
      " 36  Shoujo                11911 non-null  int64   \n",
      " 37  Shoujo Ai             11911 non-null  int64   \n",
      " 38  Shounen               11911 non-null  int64   \n",
      " 39  Shounen Ai            11911 non-null  int64   \n",
      " 40  Slice of Life         11911 non-null  int64   \n",
      " 41  Space                 11911 non-null  int64   \n",
      " 42  Sports                11911 non-null  int64   \n",
      " 43  Super Power           11911 non-null  int64   \n",
      " 44  Supernatural          11911 non-null  int64   \n",
      " 45  Thriller              11911 non-null  int64   \n",
      " 46  Vampire               11911 non-null  int64   \n",
      " 47  Yaoi                  11911 non-null  int64   \n",
      " 48  Yuri                  11911 non-null  int64   \n",
      " 49  views_without_review  11911 non-null  int64   \n",
      " 50  rating_counts         11911 non-null  int64   \n",
      "dtypes: category(1), float64(1), int64(48), string(1)\n",
      "memory usage: 4.6 MB\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "anime_df['views_without_review'] = anime_df['views_without_review'].astype('int')\n",
    "anime_df['rating_counts'] = anime_df['rating_counts'].astype('int')\n",
    "\n",
    "print(anime_df.info())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8fd4e550-bac6-4adc-b928-0114687c15e4",
   "metadata": {},
   "source": [
    "## Data Definition<a id='data_definition'></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 563,
   "id": "3959a8a6-20fb-4c4f-bba7-12512297cd6d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# save data\n",
    "anime_df.to_csv('data/anime_df.csv', sep=',')\n",
    "rating_df.to_csv('data/rating_df.csv', sep=',')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a4792d62-6062-4db3-89d9-948ed14af2a5",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}