{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "YG4fX89xbPeQ" }, "source": [ "### EDA" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "id": "I0-UozYqPKzD" }, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 678 }, "id": "likh-JLqO50o", "outputId": "82a49b50-f756-4cbf-c440-700d20b384dd" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shape of the Dataset: (12194, 18)\n" ] }, { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"df_anime\",\n \"rows\": 12194,\n \"fields\": [\n {\n \"column\": \"anime_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 11437,\n \"min\": 1,\n \"max\": 34527,\n \"num_unique_values\": 12194,\n \"samples\": [\n 3834,\n 32936,\n 8792\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"genres\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3256,\n \"samples\": [\n \"Mystery, Sci-Fi, Space\",\n \"Music, Sci-Fi\",\n \"Magic, Romance\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 12194,\n \"samples\": [\n \"Hoshi no Ko Chobin\",\n \"Gin no Guardian\",\n \"Madobe Nanami no Windows 7 de PC Jisaku Ouen Commercial!!\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"average_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 552,\n \"samples\": [\n \"4.06\",\n \"8.03\",\n \"7.32\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"overview\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 11689,\n \"samples\": [\n \"Serebii, a Legendary Pok\\u00e9mon known for its ability to traverse time, is hunted by an unnamed Pok\\u00e9mon poacher seeking to capture it. Yukinari, a young Pok\\u00e9mon trainer who enjoys drawing portraits of Pok\\u00e9mon, tries to protect Serebii after it stumbles upon him; but in the middle of its escape, both vanish without a trace.\\n\\nForty years later, ambitious Pok\\u00e9mon trainer Satoshi hopes to sight rare Pok\\u00e9mon around his local area. White, a boat driver, takes Satoshi to his village. Satoshi is accompanied by his three closest friends: Takeshi, a former gym leader training to be a great Pok\\u00e9mon breeder; Kasumi, a young girl wanting to become a skilled Water-type trainer; and Pikachu, Satoshi's Pok\\u00e9mon partner and first comrade.\\n\\nMeanwhile, Yukinari and Serebii reappear in Satoshi's present time and run into his group. Masked Lord Vicious, the strongest executive staff member of Team Rocket, desires to capture Serebii. Using the Dark Ball, a variant of Monster Ball that corrupts the Pok\\u00e9mon caught within and draws out their maximum power, Vicious can transform innocent Pok\\u00e9mon into powerful and frightening obstacles\\u2014including Serebii itself! Placed in a tough position, Satoshi, Yukinari, and their friends must work together to defeat Vicious and save Serebii and themselves.\",\n \"East Force meets West Force and all Hell breaks loose. The Solonoids, that lovable race of female warriors, are at it again, fighting amongst themselves. During a heated battle, however, it looks like the leaders of the two factions have hung their warriors out to dry. In the middle of all this fighting and chaos, East Force detects a transmission from an unidentified planet. The Gall Force gals leave their posts to go. Lufy, the West Force's Ace Pilot, who is after Rabby, follows them to the planet.\\n\\n(Source: AnimeNfo)\",\n \"Netto's father Yuuichirou Hikari has made a scientific breakthrough by introducing the \\\"synchro chips\\\". If an operator and his or her navi are in a special enviroment known as a \\\"dimensional area\\\", they can fuse together in the real world via a technique called \\\"cross fusion\\\"! Yuuichirou's first test subject, Misaki Gorou, attempts the process and sadly fails. Netto offers to try with Rockman, but his father forbids it. Cross Fusion puts enormous strain on the operator's health, and battling in the real world could mean death. \\n\\n(Source: Official Site)\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"type\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"TV\",\n \"Movie\",\n \"ONA\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"episodes\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 200,\n \"samples\": [\n \"93\",\n \"27\",\n \"110\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"producers\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3070,\n \"samples\": [\n \"AIC, Lantis, Media Factory, Pony Canyon, Rakuonsha, AT-X, KlockWorx, Ryukyu Asahi Broadcasting\",\n \"Tama Production, Tokyo MX\",\n \"KAGAYA Studio\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"licensors\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 253,\n \"samples\": [\n \"ADV Films, Media Blasters\",\n \"Funimation\",\n \"Central Park Media, Maiden Japan\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"studios\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 941,\n \"samples\": [\n \"Group TAC, Ginga Ya\",\n \"J.C.Staff, Production I.G\",\n \"drop\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"source\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 16,\n \"samples\": [\n \"Visual novel\",\n \"Manga\",\n \"Novel\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"anime_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"PG-13 - Teens 13 or older\",\n \"R - 17+ (violence & profanity)\",\n \"UNKNOWN\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rank\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8846,\n \"samples\": [\n \"12549\",\n \"7778\",\n \"7761\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"popularity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5211,\n \"min\": 1,\n \"max\": 19844,\n \"num_unique_values\": 10304,\n \"samples\": [\n 5880,\n 7617,\n 17889\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"favorites\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5821,\n \"min\": 0,\n \"max\": 217606,\n \"num_unique_values\": 1357,\n \"samples\": [\n 46556,\n 1136,\n 1570\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"scored by\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 6497,\n \"samples\": [\n \"19239\",\n \"1926\",\n \"300\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"members\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 191364,\n \"min\": 142,\n \"max\": 3744541,\n \"num_unique_values\": 8192,\n \"samples\": [\n 9390,\n 8376,\n 1998\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"image url\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 12135,\n \"samples\": [\n \"https://cdn.myanimelist.net/images/anime/1195/111544.jpg\",\n \"https://cdn.myanimelist.net/images/anime/6/26448.jpg\",\n \"https://cdn.myanimelist.net/images/anime/1405/112400.jpg\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe", "variable_name": "df_anime" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
anime_idgenresnameaverage_ratingoverviewtypeepisodesproducerslicensorsstudiossourceanime_ratingrankpopularityfavoritesscored bymembersimage url
04181Drama, Fantasy, Romance, Slice of Life, Supern...Clannad: After Story8.93Clannad: After Story, the sequel to the critic...TV24Pony Canyon, TBS, Rakuonsha, Animation DoSentai FilmworksKyoto AnimationVisual novelPG-13 - Teens 13 or older19114689496397291149886https://cdn.myanimelist.net/images/anime/1299/...
128735Drama, Historical, JoseiShouwa Genroku Rakugo Shinjuu8.57Yotarou is a former yakuza member fresh out of...TV13Starchild Records, Mainichi Broadcasting Syste...UNKNOWNStudio DeenMangaPG-13 - Teens 13 or older93804571191359281445https://cdn.myanimelist.net/images/anime/1354/...
25205Action, Mystery, Romance, Supernatural, ThrillerKara no Kyoukai Movie 7: Satsujin Kousatsu (Go)8.39In February 1999, a string of murders has Shik...Movie1NotesAniplex of AmericaufotableLight novelR - 17+ (violence & profanity)18211152261108703200492https://cdn.myanimelist.net/images/anime/9/566...
3170Comedy, Drama, School, Shounen, SportsSlam Dunk8.54Hanamichi Sakuragi, infamous for his temper, m...TV101TV Asahi, AnimaxFlatiron Film Company, Geneon Entertainment USAToei AnimationMangaPG-13 - Teens 13 or older1087976879128920283226https://cdn.myanimelist.net/images/anime/12/86...
410162Josei, Slice of LifeUsagi Drop8.36Daikichi Kawachi is a 30-year-old bachelor wor...TV11Dentsu, Fuji TV, Toho, Tohokushinsha Film Corp...NIS America, Inc.Production I.GMangaPG-13 - Teens 13 or older2024255975237156479967https://cdn.myanimelist.net/images/anime/2/296...
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " anime_id genres \\\n", "0 4181 Drama, Fantasy, Romance, Slice of Life, Supern... \n", "1 28735 Drama, Historical, Josei \n", "2 5205 Action, Mystery, Romance, Supernatural, Thriller \n", "3 170 Comedy, Drama, School, Shounen, Sports \n", "4 10162 Josei, Slice of Life \n", "\n", " name average_rating \\\n", "0 Clannad: After Story 8.93 \n", "1 Shouwa Genroku Rakugo Shinjuu 8.57 \n", "2 Kara no Kyoukai Movie 7: Satsujin Kousatsu (Go) 8.39 \n", "3 Slam Dunk 8.54 \n", "4 Usagi Drop 8.36 \n", "\n", " overview type episodes \\\n", "0 Clannad: After Story, the sequel to the critic... TV 24 \n", "1 Yotarou is a former yakuza member fresh out of... TV 13 \n", "2 In February 1999, a string of murders has Shik... Movie 1 \n", "3 Hanamichi Sakuragi, infamous for his temper, m... TV 101 \n", "4 Daikichi Kawachi is a 30-year-old bachelor wor... TV 11 \n", "\n", " producers \\\n", "0 Pony Canyon, TBS, Rakuonsha, Animation Do \n", "1 Starchild Records, Mainichi Broadcasting Syste... \n", "2 Notes \n", "3 TV Asahi, Animax \n", "4 Dentsu, Fuji TV, Toho, Tohokushinsha Film Corp... \n", "\n", " licensors studios \\\n", "0 Sentai Filmworks Kyoto Animation \n", "1 UNKNOWN Studio Deen \n", "2 Aniplex of America ufotable \n", "3 Flatiron Film Company, Geneon Entertainment USA Toei Animation \n", "4 NIS America, Inc. Production I.G \n", "\n", " source anime_rating rank popularity favorites \\\n", "0 Visual novel PG-13 - Teens 13 or older 19 114 68949 \n", "1 Manga PG-13 - Teens 13 or older 93 804 5711 \n", "2 Light novel R - 17+ (violence & profanity) 182 1115 2261 \n", "3 Manga PG-13 - Teens 13 or older 108 797 6879 \n", "4 Manga PG-13 - Teens 13 or older 202 425 5975 \n", "\n", " scored by members image url \n", "0 639729 1149886 https://cdn.myanimelist.net/images/anime/1299/... \n", "1 91359 281445 https://cdn.myanimelist.net/images/anime/1354/... \n", "2 108703 200492 https://cdn.myanimelist.net/images/anime/9/566... \n", "3 128920 283226 https://cdn.myanimelist.net/images/anime/12/86... \n", "4 237156 479967 https://cdn.myanimelist.net/images/anime/2/296... " ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime=pd.read_csv(\"/content/Animes.csv\")\n", "print(\"Shape of the Dataset:\", df_anime.shape)\n", "df_anime.head()" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 488 }, "id": "Wuxs0dgVI1UL", "outputId": "fc2c7b6e-8214-4ff5-8ded-54451869e89b" }, "outputs": [ { "data": { "text/html": [ "" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from IPython.display import Image, display\n", "anime_name = 'One Piece' # Replace with the desired anime name\n", "anime_row = df_anime[df_anime['name'] == anime_name].iloc[0]\n", "\n", "image_url = anime_row['image url']\n", "display(Image(url=image_url, width=300))" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 224 }, "id": "shOuDaGzNfIl", "outputId": "c2d140e4-955f-42d0-8c48-c1d361ed7abf" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shape of the dataset: (1112830, 4)\n" ] }, { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df_rating" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_idusernameanime_idrating
0357zhambi329499
1357zhambi117597
2357zhambi63476
3357zhambi16828
4357zhambi364567
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " user_id username anime_id rating\n", "0 357 zhambi 32949 9\n", "1 357 zhambi 11759 7\n", "2 357 zhambi 6347 6\n", "3 357 zhambi 1682 8\n", "4 357 zhambi 36456 7" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_rating = pd.read_csv('/content/UserRatings.csv')\n", "print(\"Shape of the dataset:\",df_rating.shape)\n", "df_rating.head()" ] }, { "cell_type": "markdown", "metadata": { "id": "kNQcwHOARyLr" }, "source": [ "### Data Preprocessing" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 489 }, "id": "-2x7u42nj9OG", "outputId": "22e36279-398f-4f43-a70c-9138c859f245" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
genres
Hentai821
Comedy517
Music298
Kids197
Comedy, Slice of Life176
......
Action, Comedy, Fantasy, Hentai, Horror, Mecha, Romance, Supernatural1
Action, Hentai, Martial Arts1
Action, Hentai, Historical, Mecha, Military1
Drama, Music, Psychological, Romance, Slice of Life, Yaoi1
Adventure, Fantasy, Hentai1
\n", "

3256 rows × 1 columns

\n", "

" ], "text/plain": [ "genres\n", "Hentai 821\n", "Comedy 517\n", "Music 298\n", "Kids 197\n", "Comedy, Slice of Life 176\n", " ... \n", "Action, Comedy, Fantasy, Hentai, Horror, Mecha, Romance, Supernatural 1\n", "Action, Hentai, Martial Arts 1\n", "Action, Hentai, Historical, Mecha, Military 1\n", "Drama, Music, Psychological, Romance, Slice of Life, Yaoi 1\n", "Adventure, Fantasy, Hentai 1\n", "Name: count, Length: 3256, dtype: int64" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['genres'].value_counts()" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 335 }, "id": "VMPKGQMgj9K_", "outputId": "f88e6945-b1d5-41cf-f0d3-9dd2f07bf6b6" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
anime_rating
PG-13 - Teens 13 or older4520
G - All Ages3406
Rx - Hentai1122
PG - Children1105
R+ - Mild Nudity952
R - 17+ (violence & profanity)896
UNKNOWN193
\n", "

" ], "text/plain": [ "anime_rating\n", "PG-13 - Teens 13 or older 4520\n", "G - All Ages 3406\n", "Rx - Hentai 1122\n", "PG - Children 1105\n", "R+ - Mild Nudity 952\n", "R - 17+ (violence & profanity) 896\n", "UNKNOWN 193\n", "Name: count, dtype: int64" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['anime_rating'].value_counts()" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 489 }, "id": "7BvcGc4-j9C2", "outputId": "211e2609-17f3-41c5-b741-1e0caed1a590" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
producers
UNKNOWN4621
NHK260
Pink Pineapple189
Sanrio142
Bandai Visual118
......
Lantis, AT-X, Barnum Studio, Magic Capsule, Warner Bros. Japan, KlockWorx, Shueisha, Bulls Eye1
Dentsu, Sotsu Music Publishing1
TV Tokyo, Kadokawa Shoten, Kadokawa Pictures Japan, KlockWorx, NTT Docomo, Dwango1
Kadokawa Shoten, Rondo Robe1
Frontier Works, Media Factory, Mainichi Broadcasting System, Movic, Gentosha Comics1
\n", "

3070 rows × 1 columns

\n", "

" ], "text/plain": [ "producers\n", "UNKNOWN 4621\n", "NHK 260\n", "Pink Pineapple 189\n", "Sanrio 142\n", "Bandai Visual 118\n", " ... \n", "Lantis, AT-X, Barnum Studio, Magic Capsule, Warner Bros. Japan, KlockWorx, Shueisha, Bulls Eye 1\n", "Dentsu, Sotsu Music Publishing 1\n", "TV Tokyo, Kadokawa Shoten, Kadokawa Pictures Japan, KlockWorx, NTT Docomo, Dwango 1\n", "Kadokawa Shoten, Rondo Robe 1\n", "Frontier Works, Media Factory, Mainichi Broadcasting System, Movic, Gentosha Comics 1\n", "Name: count, Length: 3070, dtype: int64" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['producers'].value_counts()" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 489 }, "id": "j9f0t6G3j8_V", "outputId": "6b766f50-dc27-4dcb-bebe-02fffa189519" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
studios
UNKNOWN2977
Toei Animation726
Sunrise442
J.C.Staff296
Madhouse294
......
KKC Animation Production1
Echo1
Project Team Muu1
Studio March1
IKK Room1
\n", "

941 rows × 1 columns

\n", "

" ], "text/plain": [ "studios\n", "UNKNOWN 2977\n", "Toei Animation 726\n", "Sunrise 442\n", "J.C.Staff 296\n", "Madhouse 294\n", " ... \n", "KKC Animation Production 1\n", "Echo 1\n", "Project Team Muu 1\n", "Studio March 1\n", "IKK Room 1\n", "Name: count, Length: 941, dtype: int64" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['studios'].value_counts()" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 335 }, "id": "b3yvb-1qkWXn", "outputId": "a5e03a1b-419f-4c2f-88e5-22ba2fd54420" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
type
TV3791
OVA3280
Movie2340
Special1626
ONA674
Music482
UNKNOWN1
\n", "

" ], "text/plain": [ "type\n", "TV 3791\n", "OVA 3280\n", "Movie 2340\n", "Special 1626\n", "ONA 674\n", "Music 482\n", "UNKNOWN 1\n", "Name: count, dtype: int64" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['type'].value_counts()" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 617 }, "id": "LSku7YhHkXR_", "outputId": "80f3ab22-f673-4d66-a133-93dbb47710a9" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
source
Manga3131
Original2983
Unknown2660
Visual novel937
Light novel541
Game529
Other355
Novel324
4-koma manga207
Music163
Book98
Web manga97
Picture book88
Card game40
Mixed media31
Radio10
\n", "

" ], "text/plain": [ "source\n", "Manga 3131\n", "Original 2983\n", "Unknown 2660\n", "Visual novel 937\n", "Light novel 541\n", "Game 529\n", "Other 355\n", "Novel 324\n", "4-koma manga 207\n", "Music 163\n", "Book 98\n", "Web manga 97\n", "Picture book 88\n", "Card game 40\n", "Mixed media 31\n", "Radio 10\n", "Name: count, dtype: int64" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['source'].value_counts()" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 1000 }, "id": "4moz0gfuaDfF", "outputId": "6bdca5b7-114b-4090-8e7c-224ddd6b5d12" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"df_anime[df_anime['average_rating']=='UNKNOWN']\",\n \"rows\": 1617,\n \"fields\": [\n {\n \"column\": \"anime_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 8248,\n \"min\": 1739,\n \"max\": 34527,\n \"num_unique_values\": 1617,\n \"samples\": [\n 32237,\n 23973,\n 27497\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"genres\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 375,\n \"samples\": [\n \"Action, Adventure, Shounen\",\n \"Demons, Horror, Kids\",\n \"Drama, Historical, Kids, Samurai\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 1617,\n \"samples\": [\n \"Burutabu-chan\",\n \"Fuku-chan\",\n \"Sore Ike! Anpanman: Kieta Jam Ojisan\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"average_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 1,\n \"samples\": [\n \"UNKNOWN\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"overview\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 1409,\n \"samples\": [\n \"Depiction of Japan's economy, based on Shoutarou Ishinomori's epynomous manga.\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"type\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 6,\n \"samples\": [\n \"OVA\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"episodes\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 101,\n \"samples\": [\n \"110\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"producers\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 264,\n \"samples\": [\n \"SPO Entertainment\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"licensors\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 5,\n \"samples\": [\n \"Saban Entertainment\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"studios\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 134,\n \"samples\": [\n \"Studio Junio\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"source\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 15,\n \"samples\": [\n \"Web manga\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"anime_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"G - All Ages\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rank\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 1390,\n \"samples\": [\n \"18249\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"popularity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 1617,\n \"min\": 6986,\n \"max\": 19844,\n \"num_unique_values\": 1453,\n \"samples\": [\n 14878\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"favorites\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0,\n \"min\": 0,\n \"max\": 7,\n \"num_unique_values\": 8,\n \"samples\": [\n 1\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"scored by\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 1,\n \"samples\": [\n \"UNKNOWN\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"members\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 246,\n \"min\": 142,\n \"max\": 6989,\n \"num_unique_values\": 498,\n \"samples\": [\n 699\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"image url\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 1563,\n \"samples\": [\n \"https://cdn.myanimelist.net/images/anime/1939/96650.jpg\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
anime_idgenresnameaverage_ratingoverviewtypeepisodesproducerslicensorsstudiossourceanime_ratingrankpopularityfavoritesscored bymembersimage url
56395693HistoricalShinran-sama: Negai, Soshite HikariUNKNOWNThe life story of Shinran, founder of Joudo Sh...OVA1UNKNOWNUNKNOWNUNKNOWNUnknownG - All Ages19745155690UNKNOWN432https://cdn.myanimelist.net/images/anime/9/292...
663928247Comedy, MusicPankis! 2-jigenUNKNOWNSt. Muse Academy: the place for the \"sound of ...TV24UNKNOWNUNKNOWNDLEOriginalG - All Ages18802138111UNKNOWN698https://cdn.myanimelist.net/images/anime/11/74...
767229750ComedyAAA de Ikou!!: Yuuna & AkikoUNKNOWNNo description available for this anime.OVA1Opera HouseUNKNOWNUNKNOWNUnknownG - All Ages14448159140UNKNOWN397https://cdn.myanimelist.net/images/anime/5/715...
767625063Comedy, Drama, HistoricalAnime Roukyoku Kikou Shimizu no JirochoudenUNKNOWNA story based on Shimizu no Jirochou, an early...TV30UNKNOWNUNKNOWNUNKNOWNUnknownG - All Ages14562164280UNKNOWN359https://cdn.myanimelist.net/images/anime/4/641...
776230914Kids, SportsAnimax TaisouUNKNOWNNo description available for this anime.Special30UNKNOWNUNKNOWNUNKNOWNUnknownG - All Ages14557179130UNKNOWN254https://cdn.myanimelist.net/images/anime/6/743...
.........................................................
1102932596Adventure, Comedy, Fantasy, KidsThe Snack World (TV)UNKNOWNIn \"a certain era\" on \"a certain continent,\" b...TV50TOHO animation, Level-5UNKNOWNOLM, OLM DigitalMixed mediaG - All Ages12922137521UNKNOWN713https://cdn.myanimelist.net/images/anime/3/833...
1103534488Comedy, Kids, Sci-FiGan Gan Ganko-chanUNKNOWNThe anime follows a boy named Gen, who time tr...TV10UNKNOWNUNKNOWN10GaugeOtherG - All Ages16001159320UNKNOWN397https://cdn.myanimelist.net/images/anime/8/831...
1104034456ComedySentai Hero Sukiyaki Force: Gunma no Heiwa wo ...UNKNOWNThe anime will be a \"surreal comedy\" that emph...TV24UNKNOWNUNKNOWNStudio 4°COriginalG - All Ages19473143260UNKNOWN593https://cdn.myanimelist.net/images/anime/10/83...
119999304HentaiHappy DayUNKNOWNA 2000 film originally banned in South Korea, ...OVA1UNKNOWNUNKNOWNUNKNOWNUnknownRx - HentaiUNKNOWN139031UNKNOWN675https://cdn.myanimelist.net/images/anime/1385/...
1206420007Action, HentaiHi Gekiga Ukiyoe Senya IchiyaUNKNOWNAn adult movie about a painter whose pictures ...Movie1Toei VideoUNKNOWNUNKNOWNUnknownRx - HentaiUNKNOWN137071UNKNOWN722https://cdn.myanimelist.net/images/anime/6/533...
\n", "

1617 rows × 18 columns

\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " anime_id genres \\\n", "5639 5693 Historical \n", "6639 28247 Comedy, Music \n", "7672 29750 Comedy \n", "7676 25063 Comedy, Drama, Historical \n", "7762 30914 Kids, Sports \n", "... ... ... \n", "11029 32596 Adventure, Comedy, Fantasy, Kids \n", "11035 34488 Comedy, Kids, Sci-Fi \n", "11040 34456 Comedy \n", "11999 9304 Hentai \n", "12064 20007 Action, Hentai \n", "\n", " name average_rating \\\n", "5639 Shinran-sama: Negai, Soshite Hikari UNKNOWN \n", "6639 Pankis! 2-jigen UNKNOWN \n", "7672 AAA de Ikou!!: Yuuna & Akiko UNKNOWN \n", "7676 Anime Roukyoku Kikou Shimizu no Jirochouden UNKNOWN \n", "7762 Animax Taisou UNKNOWN \n", "... ... ... \n", "11029 The Snack World (TV) UNKNOWN \n", "11035 Gan Gan Ganko-chan UNKNOWN \n", "11040 Sentai Hero Sukiyaki Force: Gunma no Heiwa wo ... UNKNOWN \n", "11999 Happy Day UNKNOWN \n", "12064 Hi Gekiga Ukiyoe Senya Ichiya UNKNOWN \n", "\n", " overview type episodes \\\n", "5639 The life story of Shinran, founder of Joudo Sh... OVA 1 \n", "6639 St. Muse Academy: the place for the \"sound of ... TV 24 \n", "7672 No description available for this anime. OVA 1 \n", "7676 A story based on Shimizu no Jirochou, an early... TV 30 \n", "7762 No description available for this anime. Special 30 \n", "... ... ... ... \n", "11029 In \"a certain era\" on \"a certain continent,\" b... TV 50 \n", "11035 The anime follows a boy named Gen, who time tr... TV 10 \n", "11040 The anime will be a \"surreal comedy\" that emph... TV 24 \n", "11999 A 2000 film originally banned in South Korea, ... OVA 1 \n", "12064 An adult movie about a painter whose pictures ... Movie 1 \n", "\n", " producers licensors studios source \\\n", "5639 UNKNOWN UNKNOWN UNKNOWN Unknown \n", "6639 UNKNOWN UNKNOWN DLE Original \n", "7672 Opera House UNKNOWN UNKNOWN Unknown \n", "7676 UNKNOWN UNKNOWN UNKNOWN Unknown \n", "7762 UNKNOWN UNKNOWN UNKNOWN Unknown \n", "... ... ... ... ... \n", "11029 TOHO animation, Level-5 UNKNOWN OLM, OLM Digital Mixed media \n", "11035 UNKNOWN UNKNOWN 10Gauge Other \n", "11040 UNKNOWN UNKNOWN Studio 4°C Original \n", "11999 UNKNOWN UNKNOWN UNKNOWN Unknown \n", "12064 Toei Video UNKNOWN UNKNOWN Unknown \n", "\n", " anime_rating rank popularity favorites scored by members \\\n", "5639 G - All Ages 19745 15569 0 UNKNOWN 432 \n", "6639 G - All Ages 18802 13811 1 UNKNOWN 698 \n", "7672 G - All Ages 14448 15914 0 UNKNOWN 397 \n", "7676 G - All Ages 14562 16428 0 UNKNOWN 359 \n", "7762 G - All Ages 14557 17913 0 UNKNOWN 254 \n", "... ... ... ... ... ... ... \n", "11029 G - All Ages 12922 13752 1 UNKNOWN 713 \n", "11035 G - All Ages 16001 15932 0 UNKNOWN 397 \n", "11040 G - All Ages 19473 14326 0 UNKNOWN 593 \n", "11999 Rx - Hentai UNKNOWN 13903 1 UNKNOWN 675 \n", "12064 Rx - Hentai UNKNOWN 13707 1 UNKNOWN 722 \n", "\n", " image url \n", "5639 https://cdn.myanimelist.net/images/anime/9/292... \n", "6639 https://cdn.myanimelist.net/images/anime/11/74... \n", "7672 https://cdn.myanimelist.net/images/anime/5/715... \n", "7676 https://cdn.myanimelist.net/images/anime/4/641... \n", "7762 https://cdn.myanimelist.net/images/anime/6/743... \n", "... ... \n", "11029 https://cdn.myanimelist.net/images/anime/3/833... \n", "11035 https://cdn.myanimelist.net/images/anime/8/831... \n", "11040 https://cdn.myanimelist.net/images/anime/10/83... \n", "11999 https://cdn.myanimelist.net/images/anime/1385/... \n", "12064 https://cdn.myanimelist.net/images/anime/6/533... \n", "\n", "[1617 rows x 18 columns]" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime[df_anime['average_rating']=='UNKNOWN']" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 458 }, "id": "P6TkDp_RZsKA", "outputId": "d3e49166-1cf0-481c-d56f-2bb7284d3f3e" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
average_rating
08.93
18.57
28.39
38.54
48.36
......
121895.13
121905.22
121915.29
121924.87
121935.27
\n", "

12194 rows × 1 columns

\n", "

" ], "text/plain": [ "0 8.93\n", "1 8.57\n", "2 8.39\n", "3 8.54\n", "4 8.36\n", " ... \n", "12189 5.13\n", "12190 5.22\n", "12191 5.29\n", "12192 4.87\n", "12193 5.27\n", "Name: average_rating, Length: 12194, dtype: float64" ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime['average_rating'].replace('UNKNOWN', np.nan)\n", "df_anime['average_rating'] = pd.to_numeric(df_anime['average_rating'], errors='coerce')\n", "df_anime['average_rating'].fillna(df_anime['average_rating'].median())" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "i0ypOHSolUZA", "outputId": "7702b9cb-2267-4e8b-854e-e49a8ddd0417" }, "outputs": [ { "data": { "text/plain": [ "(12194, 18)" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_anime.shape" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 655 }, "id": "JT7QEJlCOgTO", "outputId": "9f2405ed-2bb6-4254-e203-f3ea5e3ad990" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shape of the dataset: (597114, 21)\n" ] }, { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df_merged" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_idusernameanime_idratinggenresnameaverage_ratingoverviewtypeepisodes...licensorsstudiossourceanime_ratingrankpopularityfavoritesscored bymembersimage url
0357zhambi329499Drama, Romance, School, SeinenKuzu no Honkai7.14To the outside world, Hanabi Yasuraoka and Mug...TV12...Sentai FilmworksLercheMangaR+ - Mild Nudity33452866262320010657725https://cdn.myanimelist.net/images/anime/5/839...
1357zhambi117597Action, Game, Romance, School, Sci-FiAccel World7.23Haruyuki Arita is an overweight, bullied middl...TV24...VIZ MediaSunriseLight novelPG-13 - Teens 13 or older28552633882380124697894https://cdn.myanimelist.net/images/anime/1002/...
2357zhambi63476Comedy, Romance, School, Super PowerBaka to Test to Shoukanjuu7.52Fumizuki Academy isn't a typical Japanese high...TV13...FunimationSILVER LINK.Light novelPG-13 - Teens 13 or older16453145586339916620729https://cdn.myanimelist.net/images/anime/3/503...
7357zhambi289996Drama, School, Super PowerCharlotte7.75If not for his ability to take over people's m...TV13...Aniplex of AmericaP.A. WorksOriginalPG-13 - Teens 13 or older102566221809390701536653https://cdn.myanimelist.net/images/anime/12/74...
9357zhambi18187Action, Adventure, Demons, Fantasy, Shounen, S...Claymore7.74When a shapeshifting demon with a thirst for h...TV26...FunimationMadhouseMangaR+ - Mild Nudity10362928252317263650639https://cdn.myanimelist.net/images/anime/3/218...
\n", "

5 rows × 21 columns

\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " user_id username anime_id rating \\\n", "0 357 zhambi 32949 9 \n", "1 357 zhambi 11759 7 \n", "2 357 zhambi 6347 6 \n", "7 357 zhambi 28999 6 \n", "9 357 zhambi 1818 7 \n", "\n", " genres \\\n", "0 Drama, Romance, School, Seinen \n", "1 Action, Game, Romance, School, Sci-Fi \n", "2 Comedy, Romance, School, Super Power \n", "7 Drama, School, Super Power \n", "9 Action, Adventure, Demons, Fantasy, Shounen, S... \n", "\n", " name average_rating \\\n", "0 Kuzu no Honkai 7.14 \n", "1 Accel World 7.23 \n", "2 Baka to Test to Shoukanjuu 7.52 \n", "7 Charlotte 7.75 \n", "9 Claymore 7.74 \n", "\n", " overview type episodes ... \\\n", "0 To the outside world, Hanabi Yasuraoka and Mug... TV 12 ... \n", "1 Haruyuki Arita is an overweight, bullied middl... TV 24 ... \n", "2 Fumizuki Academy isn't a typical Japanese high... TV 13 ... \n", "7 If not for his ability to take over people's m... TV 13 ... \n", "9 When a shapeshifting demon with a thirst for h... TV 26 ... \n", "\n", " licensors studios source anime_rating \\\n", "0 Sentai Filmworks Lerche Manga R+ - Mild Nudity \n", "1 VIZ Media Sunrise Light novel PG-13 - Teens 13 or older \n", "2 Funimation SILVER LINK. Light novel PG-13 - Teens 13 or older \n", "7 Aniplex of America P.A. Works Original PG-13 - Teens 13 or older \n", "9 Funimation Madhouse Manga R+ - Mild Nudity \n", "\n", " rank popularity favorites scored by members \\\n", "0 3345 286 6262 320010 657725 \n", "1 2855 263 3882 380124 697894 \n", "2 1645 314 5586 339916 620729 \n", "7 1025 66 22180 939070 1536653 \n", "9 1036 292 8252 317263 650639 \n", "\n", " image url \n", "0 https://cdn.myanimelist.net/images/anime/5/839... \n", "1 https://cdn.myanimelist.net/images/anime/1002/... \n", "2 https://cdn.myanimelist.net/images/anime/3/503... \n", "7 https://cdn.myanimelist.net/images/anime/12/74... \n", "9 https://cdn.myanimelist.net/images/anime/3/218... \n", "\n", "[5 rows x 21 columns]" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_merged = pd.merge(df_rating, df_anime, on='anime_id', how='inner')\n", "df_merged = df_merged[df_merged['average_rating'] >7]\n", "print(\"Shape of the dataset:\",df_merged.shape)\n", "df_merged.head()" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 489 }, "id": "5kfjePvSql7r", "outputId": "91f5eea0-ccaa-4871-a2b1-f49f62aa83f2" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
count
average_rating
7.2910210
7.288907
7.038265
7.417549
7.197312
......
8.40285
9.06283
8.98235
9.02213
8.68163
\n", "

188 rows × 1 columns

\n", "

" ], "text/plain": [ "average_rating\n", "7.29 10210\n", "7.28 8907\n", "7.03 8265\n", "7.41 7549\n", "7.19 7312\n", " ... \n", "8.40 285\n", "9.06 283\n", "8.98 235\n", "9.02 213\n", "8.68 163\n", "Name: count, Length: 188, dtype: int64" ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_merged['average_rating'].value_counts()" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "jAOJO2udOjiM", "outputId": "05a08e2f-1dc3-48bf-ee2c-221db6973f6c" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "Index: 597114 entries, 0 to 916415\n", "Data columns (total 21 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 user_id 597114 non-null int64 \n", " 1 username 597114 non-null object \n", " 2 anime_id 597114 non-null int64 \n", " 3 rating 597114 non-null int64 \n", " 4 genres 597036 non-null object \n", " 5 name 597114 non-null object \n", " 6 average_rating 597114 non-null float64\n", " 7 overview 597114 non-null object \n", " 8 type 597114 non-null object \n", " 9 episodes 597114 non-null object \n", " 10 producers 597114 non-null object \n", " 11 licensors 597114 non-null object \n", " 12 studios 597114 non-null object \n", " 13 source 597114 non-null object \n", " 14 anime_rating 597114 non-null object \n", " 15 rank 597114 non-null object \n", " 16 popularity 597114 non-null int64 \n", " 17 favorites 597114 non-null int64 \n", " 18 scored by 597114 non-null object \n", " 19 members 597114 non-null int64 \n", " 20 image url 597114 non-null object \n", "dtypes: float64(1), int64(6), object(14)\n", "memory usage: 100.2+ MB\n" ] } ], "source": [ "df_merged.info()" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 300 }, "id": "QtgsHwiFOjfZ", "outputId": "1d844e91-c764-443e-9cc7-eb99c6c3e373" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"df_merged\",\n \"rows\": 8,\n \"fields\": [\n {\n \"column\": \"user_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 398882.69546289934,\n \"min\": 357.0,\n \"max\": 1289699.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 356239.8046302716,\n 333068.0,\n 597114.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"anime_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 207053.72350569145,\n \"min\": 1.0,\n \"max\": 597114.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 11478.22319523575,\n 8726.0,\n 597114.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rating\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 211109.3100937837,\n \"min\": 1.2842744331903606,\n \"max\": 597114.0,\n \"num_unique_values\": 7,\n \"samples\": [\n 597114.0,\n 7.623353329514967,\n 8.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"average_rating\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 211109.3078458958,\n \"min\": 0.44481283135607746,\n \"max\": 597114.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 7.637327545493827,\n 7.55,\n 597114.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"popularity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 210119.99997902242,\n \"min\": 1.0,\n \"max\": 597114.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 1710.9576596763768,\n 1266.0,\n 597114.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"favorites\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 212144.85938762734,\n \"min\": 0.0,\n \"max\": 597114.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 6485.020051447463,\n 775.0,\n 597114.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"members\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 1232709.0091458065,\n \"min\": 928.0,\n \"max\": 3744541.0,\n \"num_unique_values\": 8,\n \"samples\": [\n 377879.0538825082,\n 174702.0,\n 597114.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_idanime_idratingaverage_ratingpopularityfavoritesmembers
count5.971140e+05597114.000000597114.000000597114.000000597114.000000597114.0000005.971140e+05
mean3.562398e+0511478.2231957.6233537.6373281710.9576606485.0200513.778791e+05
std3.420692e+0510619.1915511.2842740.4448131586.61333919296.7735245.285431e+05
min3.570000e+021.0000005.0000007.0100001.0000000.0000009.280000e+02
25%5.949300e+041858.0000007.0000007.280000477.000000161.0000006.512800e+04
50%3.330680e+058726.0000008.0000007.5500001266.000000775.0000001.747020e+05
75%4.783660e+0518689.0000008.0000007.9300002527.0000003729.0000004.363000e+05
max1.289699e+0634514.00000010.0000009.10000012871.000000217606.0000003.744541e+06
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " user_id anime_id rating average_rating \\\n", "count 5.971140e+05 597114.000000 597114.000000 597114.000000 \n", "mean 3.562398e+05 11478.223195 7.623353 7.637328 \n", "std 3.420692e+05 10619.191551 1.284274 0.444813 \n", "min 3.570000e+02 1.000000 5.000000 7.010000 \n", "25% 5.949300e+04 1858.000000 7.000000 7.280000 \n", "50% 3.330680e+05 8726.000000 8.000000 7.550000 \n", "75% 4.783660e+05 18689.000000 8.000000 7.930000 \n", "max 1.289699e+06 34514.000000 10.000000 9.100000 \n", "\n", " popularity favorites members \n", "count 597114.000000 597114.000000 5.971140e+05 \n", "mean 1710.957660 6485.020051 3.778791e+05 \n", "std 1586.613339 19296.773524 5.285431e+05 \n", "min 1.000000 0.000000 9.280000e+02 \n", "25% 477.000000 161.000000 6.512800e+04 \n", "50% 1266.000000 775.000000 1.747020e+05 \n", "75% 2527.000000 3729.000000 4.363000e+05 \n", "max 12871.000000 217606.000000 3.744541e+06 " ] }, "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_merged.describe()" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 299 }, "id": "SsHywGWbOjc6", "outputId": "ff4738cb-41ab-4a84-e953-8e2c863a3b29" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"df_merged\",\n \"rows\": 4,\n \"fields\": [\n {\n \"column\": \"username\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 813,\n \"1590\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"genres\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 1439,\n \"9469\",\n \"597036\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 2945,\n \"737\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"overview\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 2928,\n \"924\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"type\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 6,\n \"372132\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"episodes\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 137,\n \"157020\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"producers\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 1416,\n \"79628\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"licensors\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 179,\n \"118730\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"studios\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 322,\n \"37217\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"source\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 15,\n \"269970\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"anime_rating\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 7,\n \"376604\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rank\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 2502,\n \"6549\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"scored by\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 2877,\n \"737\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"image url\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4,\n \"samples\": [\n 2945,\n \"737\",\n \"597114\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
usernamegenresnameoverviewtypeepisodesproducerslicensorsstudiossourceanime_ratingrankscored byimage url
count597114597036597114597114597114597114597114597114597114597114597114597114597114597114
unique81314392945292861371416179322157250228772945
topMunrowjinBuuComedy, School, Slice of LifeToradora!The story follows a boy named Quon and others ...TV1UNKNOWNFunimationJ.C.StaffMangaPG-13 - Teens 13 or olderUNKNOWN1330877https://cdn.myanimelist.net/images/anime/13/22...
freq1590946973792437213215702079628118730372172699703766046549737737
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " username genres name \\\n", "count 597114 597036 597114 \n", "unique 813 1439 2945 \n", "top MunrowjinBuu Comedy, School, Slice of Life Toradora! \n", "freq 1590 9469 737 \n", "\n", " overview type episodes \\\n", "count 597114 597114 597114 \n", "unique 2928 6 137 \n", "top The story follows a boy named Quon and others ... TV 1 \n", "freq 924 372132 157020 \n", "\n", " producers licensors studios source anime_rating \\\n", "count 597114 597114 597114 597114 597114 \n", "unique 1416 179 322 15 7 \n", "top UNKNOWN Funimation J.C.Staff Manga PG-13 - Teens 13 or older \n", "freq 79628 118730 37217 269970 376604 \n", "\n", " rank scored by image url \n", "count 597114 597114 597114 \n", "unique 2502 2877 2945 \n", "top UNKNOWN 1330877 https://cdn.myanimelist.net/images/anime/13/22... \n", "freq 6549 737 737 " ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_merged.describe(include='object')" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 742 }, "id": "nAlo4dVEftL6", "outputId": "a83c31be-d0d2-4bce-d4cb-314bc99cd90e" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
0
user_id0
username0
anime_id0
rating0
genres78
name0
average_rating0
overview0
type0
episodes0
producers0
licensors0
studios0
source0
anime_rating0
rank0
popularity0
favorites0
scored by0
members0
image url0
\n", "

" ], "text/plain": [ "user_id 0\n", "username 0\n", "anime_id 0\n", "rating 0\n", "genres 78\n", "name 0\n", "average_rating 0\n", "overview 0\n", "type 0\n", "episodes 0\n", "producers 0\n", "licensors 0\n", "studios 0\n", "source 0\n", "anime_rating 0\n", "rank 0\n", "popularity 0\n", "favorites 0\n", "scored by 0\n", "members 0\n", "image url 0\n", "dtype: int64" ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_merged.isnull().sum()" ] }, { "cell_type": "markdown", "metadata": { "id": "AMU3CALHM_-e" }, "source": [ "# Recommendation Systems" ] }, { "cell_type": "markdown", "metadata": { "id": "l3FU-uaODae7" }, "source": [ "### Collaborative Filtering" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Collaborative filtering is a widely used recommendation technique that makes predictions about a user's interests by analyzing the preferences of similar users. Unlike content-based filtering, which relies on item features, collaborative filtering focuses on user interactions, such as ratings or watch history, to generate recommendations." ] }, { "cell_type": "markdown", "metadata": { "id": "BnpwD5O30YV2" }, "source": [ "#### 1) Collaborative filtering using tensorflow" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We will implement a collaborative filtering model using TensorFlow to provide personalized anime recommendations. We will use user-anime interaction data to learn latent patterns and predict user preferences for unseen anime. TensorFlow’s deep learning capabilities will help optimize the recommendation model, improving accuracy and scalability." ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 224 }, "id": "7JX5DVAgid0B", "outputId": "168a0632-01c4-410b-9b9a-fc552162421f" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shape of the dataset: (1112830, 4)\n" ] }, { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_idusernameanime_idrating
0357zhambi329499
1357zhambi117597
2357zhambi63476
3357zhambi16828
4357zhambi364567
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " user_id username anime_id rating\n", "0 357 zhambi 32949 9\n", "1 357 zhambi 11759 7\n", "2 357 zhambi 6347 6\n", "3 357 zhambi 1682 8\n", "4 357 zhambi 36456 7" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv('/content/UserRatings.csv')\n", "print(\"Shape of the dataset:\",df.shape)\n", "df.head()" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "id": "avygejj1-kWH" }, "outputs": [], "source": [ "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "%matplotlib inline\n", "import pickle\n", "import warnings\n", "warnings.filterwarnings(action='ignore')\n", "\n", "# Data Preprocessing\n", "from sklearn.preprocessing import MinMaxScaler\n", "from sklearn.preprocessing import LabelEncoder\n", "\n", "# Model Training\n", "from sklearn.utils import shuffle\n", "from sklearn.model_selection import train_test_split\n", "import tensorflow as tf\n", "\n", "## Import necessary modules for collaborative filtering\n", "from tensorflow.keras.layers import Input, Embedding, Dot, Flatten, Dense\n", "from tensorflow.keras.models import Model\n", "from tensorflow.keras.optimizers import Adam\n", "from collections import defaultdict\n", "from collections import Counter\n", "\n", "## Import necessary modules for content-based filtering\n", "from sklearn.feature_extraction.text import TfidfVectorizer\n", "from sklearn.metrics.pairwise import linear_kernel" ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "E5W9cLbu-MvR", "outputId": "b5f4c9d0-7acb-42ab-b3dd-b4a7ada88f0c" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Average Score: 7.307548322744714\n" ] } ], "source": [ "avg_score = np.mean(df['rating'])\n", "print('Average Score:', avg_score)" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "id": "iVVi9_A9-Mrl" }, "outputs": [], "source": [ "scaler = MinMaxScaler(feature_range=(0, 1))\n", "df['scaled_score'] = scaler.fit_transform(df[['rating']])" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "xSfeiwl8-MoV", "outputId": "3765d804-278e-4530-a50f-bfe52a72c098" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Number of unique users: 813, Number of unique anime: 14579\n", "Minimum rating: 5, Maximum rating: 10\n" ] } ], "source": [ "## Encoding user IDs\n", "user_encoder = LabelEncoder()\n", "df[\"user_encoded\"] = user_encoder.fit_transform(df[\"user_id\"])\n", "num_users = len(user_encoder.classes_)\n", "\n", "## Encoding anime IDs\n", "anime_encoder = LabelEncoder()\n", "df[\"anime_encoded\"] = anime_encoder.fit_transform(df[\"anime_id\"])\n", "num_animes = len(anime_encoder.classes_)\n", "\n", "# Printing dataset information\n", "print(\"Number of unique users: {}, Number of unique anime: {}\".format(num_users, num_animes))\n", "print(\"Minimum rating: {}, Maximum rating: {}\".format(min(df['rating']), max(df['rating'])))" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "qBu74dfr-Mkk", "outputId": "62179123-5ea0-49f4-b618-8aad5977e25e" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shape of X: (1112830, 2)\n", "Shape of y: (1112830,)\n" ] } ], "source": [ "# Shuffle the dataset\n", "df = shuffle(df, random_state=100)\n", "\n", "# Create feature matrix X and target variable y\n", "X = df[['user_encoded', 'anime_encoded']].values\n", "y = df[\"scaled_score\"].values\n", "\n", "# Printing dataset information\n", "print(\"Shape of X:\", X.shape)\n", "print(\"Shape of y:\", y.shape)" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "iP2HuHv0-Mhd", "outputId": "33ad6be8-7a74-4ec0-b8e5-e576f4946f63" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Number of samples in the training set: 1102830\n", "Number of samples in the test set: 10000\n" ] } ], "source": [ "test_set_size = 10000 # Number of samples to include in the test set\n", "\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=test_set_size, random_state=73)\n", "\n", "print(\"Number of samples in the training set:\", len(y_train))\n", "print(\"Number of samples in the test set:\", len(y_test))" ] }, { "cell_type": "code", "execution_count": 29, "metadata": { "id": "xue2NDXI-Md8" }, "outputs": [], "source": [ "# Prepare input data for model training and evaluation\n", "X_train_array = [X_train[:, 0], X_train[:, 1]]\n", "X_test_array = [X_test[:, 0], X_test[:, 1]]" ] }, { "cell_type": "code", "execution_count": 30, "metadata": { "id": "UsBsSr3s-May" }, "outputs": [], "source": [ "def RecommenderNet(num_users, num_animes, embedding_size=128):\n", " # User input layer and embedding layer\n", " user = Input(name='user_encoded', shape=[1])\n", " user_embedding = Embedding(name='user_embedding', input_dim=num_users, output_dim=embedding_size)(user)\n", "\n", " # Anime input layer and embedding layer\n", " anime = Input(name='anime_encoded', shape=[1])\n", " anime_embedding = Embedding(name='anime_embedding', input_dim=num_animes, output_dim=embedding_size)(anime)\n", "\n", " # Dot product of user and anime embeddings\n", " dot_product = Dot(name='dot_product', normalize=True, axes=2)([user_embedding, anime_embedding])\n", " flattened = Flatten()(dot_product)\n", "\n", " # Dense layers for prediction\n", " dense = Dense(64, activation='relu')(flattened)\n", " output = Dense(1, activation='sigmoid')(dense)\n", "\n", " # Create and compile the model\n", " model = Model(inputs=[user, anime], outputs=output)\n", " model.compile(loss='binary_crossentropy', optimizer=Adam(learning_rate=0.001), metrics=[\"mae\", \"mse\"])\n", "\n", " return model\n", "model = RecommenderNet(num_users, num_animes)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "cKr7GRrOjcyL" }, "outputs": [], "source": [ "!pip install cloud-tpu-client==0.10 https://storage.googleapis.com/tpu-pytorch/wheels/torch_xla-1.13-cp310-cp310-linux_x86_64.whl\n" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 487 }, "id": "BBRvWziYjbls", "outputId": "149d5bf9-20be-4624-eef8-bef7267f017d" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "All devices: [LogicalDevice(name='/device:TPU:0', device_type='TPU'), LogicalDevice(name='/device:TPU:1', device_type='TPU'), LogicalDevice(name='/device:TPU:2', device_type='TPU'), LogicalDevice(name='/device:TPU:3', device_type='TPU'), LogicalDevice(name='/device:TPU:4', device_type='TPU'), LogicalDevice(name='/device:TPU:5', device_type='TPU'), LogicalDevice(name='/device:TPU:6', device_type='TPU'), LogicalDevice(name='/device:TPU:7', device_type='TPU')]\n" ] }, { "data": { "text/html": [ "
Model: \"functional_1\"\n",
              "
\n" ], "text/plain": [ "\u001b[1mModel: \"functional_1\"\u001b[0m\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┓\n",
              "┃ Layer (type)              ┃ Output Shape           ┃        Param # ┃ Connected to           ┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━┩\n",
              "│ user_encoded (InputLayer) │ (None, 1)              │              0 │ -                      │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ anime_encoded             │ (None, 1)              │              0 │ -                      │\n",
              "│ (InputLayer)              │                        │                │                        │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ user_embedding            │ (None, 1, 128)         │        104,064 │ user_encoded[0][0]     │\n",
              "│ (Embedding)               │                        │                │                        │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ anime_embedding           │ (None, 1, 128)         │      1,866,112 │ anime_encoded[0][0]    │\n",
              "│ (Embedding)               │                        │                │                        │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ dot_product (Dot)         │ (None, 1, 1)           │              0 │ user_embedding[0][0],  │\n",
              "│                           │                        │                │ anime_embedding[0][0]  │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ flatten_1 (Flatten)       │ (None, 1)              │              0 │ dot_product[0][0]      │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ dense_2 (Dense)           │ (None, 64)             │            128 │ flatten_1[0][0]        │\n",
              "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n",
              "│ dense_3 (Dense)           │ (None, 1)              │             65 │ dense_2[0][0]          │\n",
              "└───────────────────────────┴────────────────────────┴────────────────┴────────────────────────┘\n",
              "
\n" ], "text/plain": [ "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┓\n", "┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mConnected to \u001b[0m\u001b[1m \u001b[0m┃\n", "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━┩\n", "│ user_encoded (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ anime_encoded │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n", "│ (\u001b[38;5;33mInputLayer\u001b[0m) │ │ │ │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ user_embedding │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m104,064\u001b[0m │ user_encoded[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", "│ (\u001b[38;5;33mEmbedding\u001b[0m) │ │ │ │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ anime_embedding │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m1,866,112\u001b[0m │ anime_encoded[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", "│ (\u001b[38;5;33mEmbedding\u001b[0m) │ │ │ │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ dot_product (\u001b[38;5;33mDot\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ user_embedding[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m], │\n", "│ │ │ │ anime_embedding[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ flatten_1 (\u001b[38;5;33mFlatten\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ dot_product[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ dense_2 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m128\u001b[0m │ flatten_1[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", "├───────────────────────────┼────────────────────────┼────────────────┼────────────────────────┤\n", "│ dense_3 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m65\u001b[0m │ dense_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n", "└───────────────────────────┴────────────────────────┴────────────────┴────────────────────────┘\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Total params: 1,970,369 (7.52 MB)\n",
              "
\n" ], "text/plain": [ "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m1,970,369\u001b[0m (7.52 MB)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Trainable params: 1,970,369 (7.52 MB)\n",
              "
\n" ], "text/plain": [ "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m1,970,369\u001b[0m (7.52 MB)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Non-trainable params: 0 (0.00 B)\n",
              "
\n" ], "text/plain": [ "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import tensorflow as tf\n", "tpu_strategy = None\n", "try:\n", " resolver = tf.distribute.cluster_resolver.TPUClusterResolver() # TPU detection\n", " tf.config.experimental_connect_to_cluster(resolver)\n", " tf.tpu.experimental.initialize_tpu_system(resolver)\n", " print(\"All devices: \", tf.config.list_logical_devices('TPU'))\n", " tpu_strategy = tf.distribute.TPUStrategy(resolver)\n", "except ValueError:\n", " raise BaseException('ERROR: Not connected to a TPU runtime; please see the previous cell in this notebook for instructions!')\n", "\n", "with tpu_strategy.scope(): # Creating the model in the TPUStrategy scope means we will train the model on the TPU\n", " model = RecommenderNet(num_users, num_animes)\n", "\n", "# Printing my model summary\n", "model.summary()" ] }, { "cell_type": "code", "execution_count": 33, "metadata": { "id": "tAW-DSmn-MXB" }, "outputs": [], "source": [ "## Import necessary callbacks\n", "from tensorflow.keras.callbacks import ModelCheckpoint, LearningRateScheduler, EarlyStopping\n", "\n", "# Define the initial learning rate, minimum learning rate, maximum learning rate, and batch size\n", "start_lr = 0.00001\n", "min_lr = 0.00001\n", "max_lr = 0.00005\n", "batch_size = 10000\n", "\n", "\n", "# Define the number of epochs for ramp-up, sustain, and exponential decay\n", "rampup_epochs = 5\n", "sustain_epochs = 0\n", "exp_decay = .8\n", "\n", "# Learning rate schedule function\n", "def lrfn(epoch):\n", " if epoch < rampup_epochs:\n", " return (max_lr - start_lr) / rampup_epochs * epoch + start_lr\n", " elif epoch < rampup_epochs + sustain_epochs:\n", " return max_lr\n", " else:\n", " return (max_lr - min_lr) * exp_decay**(epoch - rampup_epochs - sustain_epochs) + min_lr\n", "\n", "# Learning rate scheduler callback\n", "lr_callback = LearningRateScheduler(lambda epoch: lrfn(epoch), verbose=0)\n", "\n", "# File path for saving the model weights\n", "checkpoint_filepath = 'myanimeweights.weights.h5'\n", "\n", "# Model checkpoint callback to save the best weights\n", "model_checkpoints = ModelCheckpoint(filepath=checkpoint_filepath,\n", " save_weights_only=True,\n", " monitor='val_loss',\n", " mode='min',\n", " save_best_only=True)\n", "\n", "# Early stopping callback to prevent overfitting\n", "early_stopping = EarlyStopping(patience=3, monitor='val_loss', mode='min', restore_best_weights=True)\n", "\n", "# Define the list of callbacks\n", "my_callbacks = [\n", " model_checkpoints,\n", " lr_callback,\n", " early_stopping\n", "]" ] }, { "cell_type": "code", "execution_count": 34, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "Hf1Hr03I-MTw", "outputId": "b8a5ed33-50b6-458c-ebfe-76f65c67a41f" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m9s\u001b[0m 46ms/step - loss: 0.6932 - mae: 0.2255 - mse: 0.0732 - val_loss: 0.6929 - val_mae: 0.2204 - val_mse: 0.0702 - learning_rate: 1.0000e-05\n", "Epoch 2/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 29ms/step - loss: 0.6928 - mae: 0.2243 - mse: 0.0725 - val_loss: 0.6923 - val_mae: 0.2198 - val_mse: 0.0699 - learning_rate: 1.8000e-05\n", "Epoch 3/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6921 - mae: 0.2236 - mse: 0.0723 - val_loss: 0.6916 - val_mae: 0.2190 - val_mse: 0.0695 - learning_rate: 2.6000e-05\n", "Epoch 4/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6915 - mae: 0.2242 - mse: 0.0727 - val_loss: 0.6909 - val_mae: 0.2179 - val_mse: 0.0691 - learning_rate: 3.4000e-05\n", "Epoch 5/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 29ms/step - loss: 0.6905 - mae: 0.2219 - mse: 0.0717 - val_loss: 0.6903 - val_mae: 0.2168 - val_mse: 0.0689 - learning_rate: 4.2000e-05\n", "Epoch 6/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6898 - mae: 0.2205 - mse: 0.0713 - val_loss: 0.6900 - val_mae: 0.2157 - val_mse: 0.0687 - learning_rate: 5.0000e-05\n", "Epoch 7/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6893 - mae: 0.2181 - mse: 0.0703 - val_loss: 0.6898 - val_mae: 0.2150 - val_mse: 0.0686 - learning_rate: 4.2000e-05\n", "Epoch 8/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6878 - mae: 0.2181 - mse: 0.0705 - val_loss: 0.6897 - val_mae: 0.2145 - val_mse: 0.0685 - learning_rate: 3.5600e-05\n", "Epoch 9/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6870 - mae: 0.2167 - mse: 0.0700 - val_loss: 0.6893 - val_mae: 0.2139 - val_mse: 0.0684 - learning_rate: 3.0480e-05\n", "Epoch 10/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6852 - mae: 0.2154 - mse: 0.0693 - val_loss: 0.6888 - val_mae: 0.2131 - val_mse: 0.0681 - learning_rate: 2.6384e-05\n", "Epoch 11/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6831 - mae: 0.2136 - mse: 0.0683 - val_loss: 0.6880 - val_mae: 0.2121 - val_mse: 0.0677 - learning_rate: 2.3107e-05\n", "Epoch 12/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6815 - mae: 0.2118 - mse: 0.0674 - val_loss: 0.6871 - val_mae: 0.2110 - val_mse: 0.0673 - learning_rate: 2.0486e-05\n", "Epoch 13/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6795 - mae: 0.2091 - mse: 0.0659 - val_loss: 0.6861 - val_mae: 0.2098 - val_mse: 0.0668 - learning_rate: 1.8389e-05\n", "Epoch 14/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6778 - mae: 0.2073 - mse: 0.0651 - val_loss: 0.6850 - val_mae: 0.2086 - val_mse: 0.0662 - learning_rate: 1.6711e-05\n", "Epoch 15/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6760 - mae: 0.2057 - mse: 0.0644 - val_loss: 0.6837 - val_mae: 0.2074 - val_mse: 0.0656 - learning_rate: 1.5369e-05\n", "Epoch 16/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6740 - mae: 0.2035 - mse: 0.0633 - val_loss: 0.6825 - val_mae: 0.2062 - val_mse: 0.0650 - learning_rate: 1.4295e-05\n", "Epoch 17/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6719 - mae: 0.2030 - mse: 0.0628 - val_loss: 0.6812 - val_mae: 0.2051 - val_mse: 0.0644 - learning_rate: 1.3436e-05\n", "Epoch 18/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6703 - mae: 0.2011 - mse: 0.0618 - val_loss: 0.6799 - val_mae: 0.2040 - val_mse: 0.0637 - learning_rate: 1.2749e-05\n", "Epoch 19/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 27ms/step - loss: 0.6684 - mae: 0.1981 - mse: 0.0604 - val_loss: 0.6786 - val_mae: 0.2028 - val_mse: 0.0631 - learning_rate: 1.2199e-05\n", "Epoch 20/20\n", "\u001b[1m135/135\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 30ms/step - loss: 0.6667 - mae: 0.1968 - mse: 0.0595 - val_loss: 0.6773 - val_mae: 0.2017 - val_mse: 0.0625 - learning_rate: 1.1759e-05\n" ] } ], "source": [ "# Model training\n", "batch_size = 1024 * tpu_strategy.num_replicas_in_sync\n", "history = model.fit(\n", " x=X_train_array,\n", " y=y_train,\n", " batch_size=batch_size,\n", " epochs=20,\n", " verbose=1,\n", " validation_data=(X_test_array, y_test),\n", " callbacks=my_callbacks\n", ")\n", "\n", "model.load_weights(checkpoint_filepath)" ] }, { "cell_type": "code", "execution_count": 35, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 472 }, "id": "xibONJ2W-MQh", "outputId": "2bb48a74-fd57-4298-9768-bb9fffdd308f" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Training results visualization\n", "plt.plot(history.history[\"loss\"][0:-2])\n", "plt.plot(history.history[\"val_loss\"][0:-2])\n", "plt.title(\"Training Loss\")\n", "plt.ylabel(\"Loss\")\n", "plt.xlabel(\"Epoch\")\n", "plt.legend([\"Train\", \"Validation\"], loc=\"upper left\")\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "s-p1X7eZRcRU", "outputId": "8119c518-69ab-49e3-c78b-c553f06c212f" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 18ms/step - loss: 0.6745 - mae: 0.2032 - mse: 0.0639\n", "Test Loss: 0.6732999682426453, Test MAE: 0.2060622125864029, Test MSE: 0.06531194597482681\n" ] } ], "source": [ "# Evaluate the model on the test set\n", "evaluation = model.evaluate(X_test_array, y_test)\n", "print(f\"Test Loss: {evaluation[0]}, Test MAE: {evaluation[1]}, Test MSE: {evaluation[2]}\")" ] }, { "cell_type": "code", "execution_count": 37, "metadata": { "id": "5eZltYvrnadX" }, "outputs": [], "source": [ "def get_unrated_anime(user_id, df, num_animes):\n", " watched_anime_ids = df[df['user_id'] == user_id]['anime_id'].tolist()\n", " all_anime_ids = list(range(num_animes))\n", " unrated_anime_ids = [anime_id for anime_id in all_anime_ids if anime_id not in watched_anime_ids]\n", " return unrated_anime_ids" ] }, { "cell_type": "code", "execution_count": 38, "metadata": { "id": "QS0Zg4DwnYIF" }, "outputs": [], "source": [ "def recommend_anime(user_id, model, num_animes, n_recommendations=10):\n", " unrated_anime_ids = get_unrated_anime(user_id, df, num_animes)\n", " user_array = np.array([user_id] * len(unrated_anime_ids))\n", " anime_array = np.array(unrated_anime_ids)\n", "\n", " predictions = model.predict([user_array, anime_array])\n", " predicted_ratings = predictions.flatten()\n", "\n", " top_indices = predicted_ratings.argsort()[-n_recommendations:][::-1]\n", " recommended_anime_ids = [unrated_anime_ids[i] for i in top_indices]\n", "\n", " return recommended_anime_ids" ] }, { "cell_type": "code", "execution_count": 39, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "sJ4RJNCPnelV", "outputId": "b66b14dd-a19b-4fef-cb8f-f4eb42679a0c" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m456/456\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m20s\u001b[0m 43ms/step\n", "Original Anime IDs for User 0: [37208 41339 3927 329 8142 36129 5091 16706 35843 28891]\n", "Recommended Anime Names for User 0: ['Planetes' 'Haikyuu!! Second Season'\n", " 'Kami nomi zo Shiru Sekai: Megami-hen'\n", " 'Kidou Senshi Gundam 00 Second Season' 'Colorful (Movie)'\n", " 'Manga Sarutobi Sasuke']\n" ] } ], "source": [ "user_id = 10\n", "recommended_anime_ids = recommend_anime(user_id, model, num_animes, n_recommendations=10)\n", "\n", "# Map the recommended anime IDs back to their names\n", "original_anime_ids = anime_encoder.inverse_transform(recommended_anime_ids)\n", "print(\"Original Anime IDs for User 0:\", original_anime_ids)\n", "recommended_anime_names = df_anime[df_anime['anime_id'].isin(original_anime_ids)]['name'].values\n", "print(\"Recommended Anime Names for User 0:\", recommended_anime_names)" ] }, { "cell_type": "markdown", "metadata": { "id": "3f4p_zFMDdXd" }, "source": [ "#### 2) Singular Value Decomposition - SVD\n", "\n", "Singular Value Decomposition (SVD) is a powerful **matrix factorization technique** widely used in collaborative filtering for recommendation systems. It helps reduce the dimensionality of user-item interaction data while preserving essential patterns, allowing us to make accurate recommendations based on user preferences.\n", " \n", "SVD helps uncover **latent relationships** between users and items, making it effective for personalized recommendations. By approximating missing values (unrated items), we can predict user preferences and recommend anime based on their past interactions. \n", "\n", "In this notebook, we will use SVD to build a recommendation model and evaluate its performance in predicting user preferences." ] }, { "cell_type": "code", "execution_count": 40, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "RBHd69NvDgyc", "outputId": "5b274a78-e948-4ab8-977f-13210ce84813" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Collecting surprise\n", " Downloading surprise-0.1-py2.py3-none-any.whl.metadata (327 bytes)\n", "Collecting scikit-surprise (from surprise)\n", " Downloading scikit_surprise-1.1.4.tar.gz (154 kB)\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m154.4/154.4 kB\u001b[0m \u001b[31m3.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[?25h Installing build dependencies ... \u001b[?25l\u001b[?25hdone\n", " Getting requirements to build wheel ... \u001b[?25l\u001b[?25hdone\n", " Preparing metadata (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n", "Requirement already satisfied: joblib>=1.2.0 in /usr/local/lib/python3.11/dist-packages (from scikit-surprise->surprise) (1.4.2)\n", "Requirement already satisfied: numpy>=1.19.5 in /usr/local/lib/python3.11/dist-packages (from scikit-surprise->surprise) (1.26.4)\n", "Requirement already satisfied: scipy>=1.6.0 in /usr/local/lib/python3.11/dist-packages (from scikit-surprise->surprise) (1.13.1)\n", "Downloading surprise-0.1-py2.py3-none-any.whl (1.8 kB)\n", "Building wheels for collected packages: scikit-surprise\n", " Building wheel for scikit-surprise (pyproject.toml) ... \u001b[?25l\u001b[?25hdone\n", " Created wheel for scikit-surprise: filename=scikit_surprise-1.1.4-cp311-cp311-linux_x86_64.whl size=2505176 sha256=6cd9480d9f79089b96685f62df64f3247d2e3c8c6e28a8ebb4ac50c35806bdaf\n", " Stored in directory: /root/.cache/pip/wheels/2a/8f/6e/7e2899163e2d85d8266daab4aa1cdabec7a6c56f83c015b5af\n", "Successfully built scikit-surprise\n", "Installing collected packages: scikit-surprise, surprise\n", "Successfully installed scikit-surprise-1.1.4 surprise-0.1\n" ] } ], "source": [ "!pip install surprise" ] }, { "cell_type": "code", "execution_count": 41, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "VbKyIBLnTUFL", "outputId": "4dd5a2a4-4590-4d0d-ca0f-bac4ef8eb4ad" }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from surprise import SVD, Dataset, Reader\n", "from surprise.model_selection import cross_validate\n", "\n", "# Prepare the data for Surprise\n", "reader = Reader(rating_scale=(1, 10))\n", "data = Dataset.load_from_df(df_merged[['user_id', 'anime_id', 'rating']], reader)\n", "\n", "# Train the SVD model\n", "svd = SVD()\n", "cross_validate(svd, data, cv=5)\n", "\n", "# Fit the model on the entire dataset\n", "trainset = data.build_full_trainset()\n", "svd.fit(trainset)" ] }, { "cell_type": "code", "execution_count": 42, "metadata": { "id": "c0qPhkSzKvdd" }, "outputs": [], "source": [ "# Function to get top recommendations for a user\n", "def get_collab_recommendations(user_id, n=10):\n", " anime_ids = df_merged['anime_id'].unique()\n", " predictions = [(anime_id, svd.predict(user_id, anime_id).est) for anime_id in anime_ids]\n", " predictions.sort(key=lambda x: x[1], reverse=True)\n", " anime_ids = [pred[0] for pred in predictions[:n]]\n", " anime_names = set(df_merged[df_merged['anime_id'].isin(anime_ids)]['name'].tolist())\n", " return anime_names" ] }, { "cell_type": "code", "execution_count": 43, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "kQ4htwfYTUBH", "outputId": "94964b78-3e39-469a-d909-e42e70fd5075" }, "outputs": [ { "data": { "text/plain": [ "{'Clannad: After Story',\n", " 'Code Geass: Hangyaku no Lelouch',\n", " 'Code Geass: Hangyaku no Lelouch R2',\n", " 'Fullmetal Alchemist: Brotherhood',\n", " 'Ginga Eiyuu Densetsu',\n", " 'Gintama°',\n", " 'Hajime no Ippo',\n", " 'Kimi no Na wa.',\n", " 'Steins;Gate',\n", " 'Suzumiya Haruhi no Shoushitsu'}" ] }, "execution_count": 43, "metadata": {}, "output_type": "execute_result" } ], "source": [ "get_collab_recommendations(34, n=10)" ] }, { "cell_type": "markdown", "metadata": { "id": "tIdBRcwnM0pc" }, "source": [ "#### 3) Item-Based Collaborative Filtering using KNN- Anime recommendations based on user ratings\n", "\n", "Item-Based Collaborative Filtering (IBCF) is a recommendation technique that suggests items similar to those a user has already rated or interacted with. Unlike **user-based collaborative filtering**, which finds similar users, **item-based filtering** focuses on item-to-item similarities. \n", "\n", "- How It Works:\n", " 1. **Compute Item Similarity**: Using **K-Nearest Neighbors (KNN)**, we measure similarity between anime based on user rating patterns. \n", " 2. **Find Nearest Neighbors**: Identify anime that are most similar to the ones a user has rated. \n", " 3. **Generate Recommendations**: Suggest anime that are highly rated by users who liked similar items. " ] }, { "cell_type": "code", "execution_count": 44, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "FDSs7u4aM0Ep", "outputId": "8c976525-3338-4ea4-aa11-56a0eeda2fbb" }, "outputs": [ { "data": { "text/plain": [ "(597114, 21)" ] }, "execution_count": 44, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = df_merged.copy()\n", "df.shape" ] }, { "cell_type": "code", "execution_count": 45, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 637 }, "id": "YE9XO3V7fjmj", "outputId": "772fd4dc-9fcc-42a7-9b2c-21b9d595e391" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_idusernameanime_idratinggenresnameaverage_ratingoverviewtypeepisodes...licensorsstudiossourceanime_ratingrankpopularityfavoritesscored bymembersimage url
0357zhambi329499Drama, Romance, School, SeinenKuzu no Honkai7.14To the outside world, Hanabi Yasuraoka and Mug...TV12...Sentai FilmworksLercheMangaR+ - Mild Nudity33452866262320010657725https://cdn.myanimelist.net/images/anime/5/839...
1357zhambi117597Action, Game, Romance, School, Sci-FiAccel World7.23Haruyuki Arita is an overweight, bullied middl...TV24...VIZ MediaSunriseLight novelPG-13 - Teens 13 or older28552633882380124697894https://cdn.myanimelist.net/images/anime/1002/...
2357zhambi63476Comedy, Romance, School, Super PowerBaka to Test to Shoukanjuu7.52Fumizuki Academy isn't a typical Japanese high...TV13...FunimationSILVER LINK.Light novelPG-13 - Teens 13 or older16453145586339916620729https://cdn.myanimelist.net/images/anime/3/503...
7357zhambi289996Drama, School, Super PowerCharlotte7.75If not for his ability to take over people's m...TV13...Aniplex of AmericaP.A. WorksOriginalPG-13 - Teens 13 or older102566221809390701536653https://cdn.myanimelist.net/images/anime/12/74...
9357zhambi18187Action, Adventure, Demons, Fantasy, Shounen, S...Claymore7.74When a shapeshifting demon with a thirst for h...TV26...FunimationMadhouseMangaR+ - Mild Nudity10362928252317263650639https://cdn.myanimelist.net/images/anime/3/218...
\n", "

5 rows × 21 columns

\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " user_id username anime_id rating \\\n", "0 357 zhambi 32949 9 \n", "1 357 zhambi 11759 7 \n", "2 357 zhambi 6347 6 \n", "7 357 zhambi 28999 6 \n", "9 357 zhambi 1818 7 \n", "\n", " genres \\\n", "0 Drama, Romance, School, Seinen \n", "1 Action, Game, Romance, School, Sci-Fi \n", "2 Comedy, Romance, School, Super Power \n", "7 Drama, School, Super Power \n", "9 Action, Adventure, Demons, Fantasy, Shounen, S... \n", "\n", " name average_rating \\\n", "0 Kuzu no Honkai 7.14 \n", "1 Accel World 7.23 \n", "2 Baka to Test to Shoukanjuu 7.52 \n", "7 Charlotte 7.75 \n", "9 Claymore 7.74 \n", "\n", " overview type episodes ... \\\n", "0 To the outside world, Hanabi Yasuraoka and Mug... TV 12 ... \n", "1 Haruyuki Arita is an overweight, bullied middl... TV 24 ... \n", "2 Fumizuki Academy isn't a typical Japanese high... TV 13 ... \n", "7 If not for his ability to take over people's m... TV 13 ... \n", "9 When a shapeshifting demon with a thirst for h... TV 26 ... \n", "\n", " licensors studios source anime_rating \\\n", "0 Sentai Filmworks Lerche Manga R+ - Mild Nudity \n", "1 VIZ Media Sunrise Light novel PG-13 - Teens 13 or older \n", "2 Funimation SILVER LINK. Light novel PG-13 - Teens 13 or older \n", "7 Aniplex of America P.A. Works Original PG-13 - Teens 13 or older \n", "9 Funimation Madhouse Manga R+ - Mild Nudity \n", "\n", " rank popularity favorites scored by members \\\n", "0 3345 286 6262 320010 657725 \n", "1 2855 263 3882 380124 697894 \n", "2 1645 314 5586 339916 620729 \n", "7 1025 66 22180 939070 1536653 \n", "9 1036 292 8252 317263 650639 \n", "\n", " image url \n", "0 https://cdn.myanimelist.net/images/anime/5/839... \n", "1 https://cdn.myanimelist.net/images/anime/1002/... \n", "2 https://cdn.myanimelist.net/images/anime/3/503... \n", "7 https://cdn.myanimelist.net/images/anime/12/74... \n", "9 https://cdn.myanimelist.net/images/anime/3/218... \n", "\n", "[5 rows x 21 columns]" ] }, "execution_count": 45, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head()" ] }, { "cell_type": "code", "execution_count": 46, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 371 }, "id": "yd75KmrcM3gA", "outputId": "2d07b032-c11b-42fb-cfc7-3fde6c42f104" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "anime_pivot" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
user_id357436467609817840949103111051251...1266443126684512686631268745127841912810591283219128817512892171289699
name
\"Bungaku Shoujo\" Memoire0.00.00.00.00.00.05.00.00.00.0...7.00.07.00.00.00.00.09.00.08.0
\"Bungaku Shoujo\" Movie0.08.00.00.00.00.00.08.09.00.0...7.00.07.00.00.00.00.09.00.07.0
.hack//G.U. Trilogy0.00.00.05.00.00.00.00.07.00.0...0.00.08.00.00.00.00.00.00.00.0
.hack//Quantum0.00.00.06.00.00.00.00.06.00.0...0.00.08.00.00.00.00.00.00.00.0
.hack//The Movie: Sekai no Mukou ni0.00.00.06.00.00.00.00.05.00.0...0.00.08.00.00.010.00.00.00.00.0
\n", "

5 rows × 813 columns

\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ "user_id 357 436 467 609 \\\n", "name \n", "\"Bungaku Shoujo\" Memoire 0.0 0.0 0.0 0.0 \n", "\"Bungaku Shoujo\" Movie 0.0 8.0 0.0 0.0 \n", ".hack//G.U. Trilogy 0.0 0.0 0.0 5.0 \n", ".hack//Quantum 0.0 0.0 0.0 6.0 \n", ".hack//The Movie: Sekai no Mukou ni 0.0 0.0 0.0 6.0 \n", "\n", "user_id 817 840 949 1031 \\\n", "name \n", "\"Bungaku Shoujo\" Memoire 0.0 0.0 5.0 0.0 \n", "\"Bungaku Shoujo\" Movie 0.0 0.0 0.0 8.0 \n", ".hack//G.U. Trilogy 0.0 0.0 0.0 0.0 \n", ".hack//Quantum 0.0 0.0 0.0 0.0 \n", ".hack//The Movie: Sekai no Mukou ni 0.0 0.0 0.0 0.0 \n", "\n", "user_id 1105 1251 ... 1266443 1266845 \\\n", "name ... \n", "\"Bungaku Shoujo\" Memoire 0.0 0.0 ... 7.0 0.0 \n", "\"Bungaku Shoujo\" Movie 9.0 0.0 ... 7.0 0.0 \n", ".hack//G.U. Trilogy 7.0 0.0 ... 0.0 0.0 \n", ".hack//Quantum 6.0 0.0 ... 0.0 0.0 \n", ".hack//The Movie: Sekai no Mukou ni 5.0 0.0 ... 0.0 0.0 \n", "\n", "user_id 1268663 1268745 1278419 1281059 \\\n", "name \n", "\"Bungaku Shoujo\" Memoire 7.0 0.0 0.0 0.0 \n", "\"Bungaku Shoujo\" Movie 7.0 0.0 0.0 0.0 \n", ".hack//G.U. Trilogy 8.0 0.0 0.0 0.0 \n", ".hack//Quantum 8.0 0.0 0.0 0.0 \n", ".hack//The Movie: Sekai no Mukou ni 8.0 0.0 0.0 10.0 \n", "\n", "user_id 1283219 1288175 1289217 1289699 \n", "name \n", "\"Bungaku Shoujo\" Memoire 0.0 9.0 0.0 8.0 \n", "\"Bungaku Shoujo\" Movie 0.0 9.0 0.0 7.0 \n", ".hack//G.U. Trilogy 0.0 0.0 0.0 0.0 \n", ".hack//Quantum 0.0 0.0 0.0 0.0 \n", ".hack//The Movie: Sekai no Mukou ni 0.0 0.0 0.0 0.0 \n", "\n", "[5 rows x 813 columns]" ] }, "execution_count": 46, "metadata": {}, "output_type": "execute_result" } ], "source": [ "anime_pivot = df.pivot_table(index='name',columns='user_id',values='rating').fillna(0)\n", "anime_pivot.head()" ] }, { "cell_type": "code", "execution_count": 47, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 80 }, "id": "z-Bunc6-XT0r", "outputId": "781cb610-004e-4613-9944-47fc02d9ed14" }, "outputs": [ { "data": { "text/html": [ "
NearestNeighbors(algorithm='brute', metric='cosine')
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "NearestNeighbors(algorithm='brute', metric='cosine')" ] }, "execution_count": 47, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from scipy.sparse import csr_matrix\n", "from sklearn.neighbors import NearestNeighbors\n", "\n", "item_user_matrix = csr_matrix(anime_pivot.values)\n", "knn_item_based = NearestNeighbors(metric = 'cosine', algorithm = 'brute')\n", "knn_item_based.fit(item_user_matrix)" ] }, { "cell_type": "code", "execution_count": 48, "metadata": { "id": "dQkd4zIdXTxK" }, "outputs": [], "source": [ "def get_item_based_recommendations(anime_name, n_recommendations=5):\n", " # Find the index of the anime title\n", " if anime_name not in anime_pivot.index:\n", " return f\"Anime title '{anime_name}' not found in the dataset.\"\n", "\n", " query_index = anime_pivot.index.get_loc(anime_name)\n", "\n", " # Use the KNN model to find the nearest neighbors\n", " distances, indices = knn_item_based.kneighbors(anime_pivot.iloc[query_index,:].values.reshape(1, -1), n_neighbors=n_recommendations + 1)\n", "\n", " recommendations = []\n", " for i in range(1, len(distances.flatten())):\n", " anime_title = anime_pivot.index[indices.flatten()[i]]\n", " distance = distances.flatten()[i]\n", " recommendations.append((anime_title, distance))\n", "\n", " return recommendations" ] }, { "cell_type": "code", "execution_count": 49, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "K7wvMvI0M3cw", "outputId": "9e9604d5-a806-41f6-990e-9176f024b785" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Based on user rating, If you like 'One Piece' you will definetly like below recommendations....\n", "1: Bleach, with distance of 0.2537\n", "2: Naruto, with distance of 0.2768\n", "3: One Piece Film: Strong World, with distance of 0.2836\n", "4: Fairy Tail, with distance of 0.2905\n", "5: Soul Eater, with distance of 0.2945\n", "6: Death Note, with distance of 0.2964\n", "7: Code Geass: Hangyaku no Lelouch, with distance of 0.3043\n", "8: Naruto: Shippuuden, with distance of 0.3101\n", "9: Code Geass: Hangyaku no Lelouch R2, with distance of 0.3106\n", "10: Fullmetal Alchemist: Brotherhood, with distance of 0.3122\n" ] } ], "source": [ "anime_name = \"One Piece\"\n", "item_based_recommendations = get_item_based_recommendations(anime_name,10)\n", "\n", "if isinstance(item_based_recommendations, str):\n", " print(item_based_recommendations)\n", "else:\n", " print(f\"Based on user rating, If you like '{anime_name}' you will definetly like below recommendations....\")\n", " for i, (title, distance) in enumerate(item_based_recommendations, 1):\n", " print(f\"{i}: {title}, with distance of {distance:.4f}\")" ] }, { "cell_type": "markdown", "metadata": { "id": "XXKBq-x9qALF" }, "source": [ "#### 4) User-Based Collaborative Filtering using KNN\n", "\n", "\n", "User-Based Collaborative Filtering (UBCF) is a recommendation approach that suggests items to users based on the preferences of similar users. Unlike **item-based filtering**, which focuses on item similarities, **user-based filtering** identifies users with similar rating patterns and recommends items they have liked. \n", "\n", "- How It Works:\n", " 1. **Compute User Similarity**: Using **K-Nearest Neighbors (KNN)**, we measure similarity between users based on their anime rating history. \n", " 2. **Find Nearest Neighbors**: Identify users who have rated anime similarly. \n", " 3. **Generate Recommendations**: Recommend anime that similar users have highly rated but the target user has not watched yet. \n" ] }, { "cell_type": "code", "execution_count": 50, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 391 }, "id": "btDeSD4ap384", "outputId": "683cc21a-fc4e-4324-aa73-a37eb9430b5d" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "user_pivot" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
name\"Bungaku Shoujo\" Memoire\"Bungaku Shoujo\" Movie.hack//G.U. Trilogy.hack//Quantum.hack//The Movie: Sekai no Mukou ni07-Ghost11-nin Iru!3-gatsu no Lion3-gatsu no Lion meets Bump of Chicken30-pun de Wakaru! Kore made no Love Live!...ef: A Tale of Melodies. - Prologueef: A Tale of Memories.ef: A Tale of Memories. - Prologueef: A Tale of Memories. - Recollectionss.CRY.edxxxHOLiCxxxHOLiC Movie: Manatsu no Yoru no YumexxxHOLiC RouxxxHOLiC ShunmukixxxHOLiC◆Kei
user_id
3570.00.00.00.00.00.00.08.00.00.0...0.07.00.00.00.07.00.00.00.00.0
4360.08.00.00.00.00.00.08.00.00.0...7.010.07.00.07.07.07.00.00.00.0
4670.00.00.00.00.00.00.08.00.00.0...0.00.00.00.07.09.08.08.07.08.0
6090.00.05.06.06.08.00.00.00.00.0...0.00.00.00.06.00.06.00.00.00.0
8170.00.00.00.00.00.00.00.00.00.0...0.08.00.00.08.08.07.09.08.08.0
\n", "

5 rows × 2945 columns

\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ "name \"Bungaku Shoujo\" Memoire \"Bungaku Shoujo\" Movie \\\n", "user_id \n", "357 0.0 0.0 \n", "436 0.0 8.0 \n", "467 0.0 0.0 \n", "609 0.0 0.0 \n", "817 0.0 0.0 \n", "\n", "name .hack//G.U. Trilogy .hack//Quantum \\\n", "user_id \n", "357 0.0 0.0 \n", "436 0.0 0.0 \n", "467 0.0 0.0 \n", "609 5.0 6.0 \n", "817 0.0 0.0 \n", "\n", "name .hack//The Movie: Sekai no Mukou ni 07-Ghost 11-nin Iru! \\\n", "user_id \n", "357 0.0 0.0 0.0 \n", "436 0.0 0.0 0.0 \n", "467 0.0 0.0 0.0 \n", "609 6.0 8.0 0.0 \n", "817 0.0 0.0 0.0 \n", "\n", "name 3-gatsu no Lion 3-gatsu no Lion meets Bump of Chicken \\\n", "user_id \n", "357 8.0 0.0 \n", "436 8.0 0.0 \n", "467 8.0 0.0 \n", "609 0.0 0.0 \n", "817 0.0 0.0 \n", "\n", "name 30-pun de Wakaru! Kore made no Love Live! ... \\\n", "user_id ... \n", "357 0.0 ... \n", "436 0.0 ... \n", "467 0.0 ... \n", "609 0.0 ... \n", "817 0.0 ... \n", "\n", "name ef: A Tale of Melodies. - Prologue ef: A Tale of Memories. \\\n", "user_id \n", "357 0.0 7.0 \n", "436 7.0 10.0 \n", "467 0.0 0.0 \n", "609 0.0 0.0 \n", "817 0.0 8.0 \n", "\n", "name ef: A Tale of Memories. - Prologue \\\n", "user_id \n", "357 0.0 \n", "436 7.0 \n", "467 0.0 \n", "609 0.0 \n", "817 0.0 \n", "\n", "name ef: A Tale of Memories. - Recollections s.CRY.ed xxxHOLiC \\\n", "user_id \n", "357 0.0 0.0 7.0 \n", "436 0.0 7.0 7.0 \n", "467 0.0 7.0 9.0 \n", "609 0.0 6.0 0.0 \n", "817 0.0 8.0 8.0 \n", "\n", "name xxxHOLiC Movie: Manatsu no Yoru no Yume xxxHOLiC Rou \\\n", "user_id \n", "357 0.0 0.0 \n", "436 7.0 0.0 \n", "467 8.0 8.0 \n", "609 6.0 0.0 \n", "817 7.0 9.0 \n", "\n", "name xxxHOLiC Shunmuki xxxHOLiC◆Kei \n", "user_id \n", "357 0.0 0.0 \n", "436 0.0 0.0 \n", "467 7.0 8.0 \n", "609 0.0 0.0 \n", "817 8.0 8.0 \n", "\n", "[5 rows x 2945 columns]" ] }, "execution_count": 50, "metadata": {}, "output_type": "execute_result" } ], "source": [ "user_pivot = df.pivot_table(index='user_id',columns='name',values='rating').fillna(0)\n", "user_pivot.head()" ] }, { "cell_type": "code", "execution_count": 51, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 80 }, "id": "QunI9SV3p35d", "outputId": "920d4970-3e54-4349-bfee-b7ddd7b23a81" }, "outputs": [ { "data": { "text/html": [ "
NearestNeighbors(algorithm='brute', metric='cosine')
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
" ], "text/plain": [ "NearestNeighbors(algorithm='brute', metric='cosine')" ] }, "execution_count": 51, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from scipy.sparse import csr_matrix\n", "from sklearn.neighbors import NearestNeighbors\n", "\n", "user_item_matrix = csr_matrix(user_pivot.values)\n", "knn_user_based = NearestNeighbors(metric = 'cosine', algorithm = 'brute')\n", "knn_user_based.fit(user_item_matrix)" ] }, { "cell_type": "code", "execution_count": 52, "metadata": { "id": "W_lzzcUr2Qoq" }, "outputs": [], "source": [ "def get_user_based_recommendations(user_id, n_recommendations=5):\n", " # Convert user_id to the same type as user_pivot index\n", " user_id = float(user_id)\n", "\n", " # Check if the user exists in the matrix\n", " if user_id not in user_pivot.index:\n", " return f\"User '{user_id}' not found in the dataset.\"\n", "\n", " # Find the user index\n", " user_idx = user_pivot.index.get_loc(user_id)\n", "\n", " # Find the nearest neighbors (most similar users)\n", " distances, indices = knn_user_based.kneighbors(user_pivot.iloc[user_idx, :].values.reshape(1, -1), n_neighbors=n_recommendations + 1)\n", "\n", " # Get the list of anime this user has already rated\n", " user_rated_anime = set(user_pivot.columns[user_pivot.iloc[user_idx, :] > 100])\n", "\n", " # Gather all anime rated by nearest neighbors\n", " all_neighbor_ratings = []\n", " for i in range(1, len(distances.flatten())):\n", " neighbor_idx = indices.flatten()[i]\n", " neighbor_rated_anime = user_pivot.iloc[neighbor_idx, :]\n", " neighbor_ratings = neighbor_rated_anime[neighbor_rated_anime > 0]\n", " all_neighbor_ratings.extend(neighbor_ratings.index)\n", "\n", " # Count the frequency of each anime rated by the neighbors\n", " from collections import Counter\n", " anime_counter = Counter(all_neighbor_ratings)\n", "\n", " # Recommend the most common anime among the neighbors that the user hasn't rated yet\n", " recommendations = [(anime, count) for anime, count in anime_counter.items() if anime not in user_rated_anime]\n", " recommendations.sort(key=lambda x: x[1], reverse=True)\n", "\n", " # Return the top N recommendations\n", " return recommendations[:n_recommendations]" ] }, { "cell_type": "code", "execution_count": 53, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "Pwenx_vZ2TSz", "outputId": "b98596e5-cab9-4cdf-d59f-4be10d4f23c9" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "User-based recommendations for user ID 817:\n", "Accel World, Recommended by 10 similar users\n", "Ai Yori Aoshi, Recommended by 10 similar users\n", "Amagami SS, Recommended by 10 similar users\n", "Angel Beats!, Recommended by 10 similar users\n", "Ano Hi Mita Hana no Namae wo Bokutachi wa Mada Shiranai., Recommended by 10 similar users\n", "Ano Natsu de Matteru, Recommended by 10 similar users\n", "Another, Recommended by 10 similar users\n", "Ao no Exorcist, Recommended by 10 similar users\n", "Arakawa Under the Bridge, Recommended by 10 similar users\n", "Asura Cryin' 2, Recommended by 10 similar users\n" ] } ], "source": [ "user_id = 817\n", "user_recommendations = get_user_based_recommendations(user_id,10)\n", "\n", "# Check if recommendations is a string (indicating an error)\n", "if isinstance(user_recommendations, str):\n", " print(user_recommendations)\n", "else:\n", " print(f\"User-based recommendations for user ID {user_id}:\")\n", " for anime, count in user_recommendations:\n", " print(f\"{anime}, Recommended by {count} similar users\")" ] }, { "cell_type": "markdown", "metadata": { "id": "lTIbtd3uL4Tm" }, "source": [ "## Content based Filtering" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Content-Based Filtering (CBF) also referred to as cognitive filtering, recommends anime based on their content attributes, such as **genre, synopsis, or other metadata**. Unlike collaborative filtering, which relies on user interactions, CBF focuses on **similarities between anime** using textual data and machine learning techniques. \n", "\n", "1. **TF-IDF with Sigmoid Kernel** \n", " - **TF-IDF (Term Frequency-Inverse Document Frequency)** transforms text into numerical features. \n", " - The **Sigmoid Kernel** measures similarity between anime based on their textual descriptions. \n", "\n", "2. **CountVectorizer with Linear Kernel** \n", " - **CountVectorizer** converts text into a bag-of-words representation. \n", " - The **Linear Kernel** computes the similarity between anime using these feature vectors. \n", "\n", "- Why Use These Methods?\n", " - **TF-IDF + Sigmoid Kernel** captures the importance of words while smoothing similarity scores. \n", " - **CountVectorizer + Linear Kernel** is efficient for comparing anime descriptions at a high level. \n" ] }, { "cell_type": "markdown", "metadata": { "id": "zzorVg94k4HF" }, "source": [ "#### 1) Using TFIDF" ] }, { "cell_type": "markdown", "metadata": { "id": "PShKOncdVfbJ" }, "source": [ "##### a. Sigmoid Kernel" ] }, { "cell_type": "code", "execution_count": 54, "metadata": { "id": "rVvjDyI9MKSr" }, "outputs": [], "source": [ "import pandas as pd" ] }, { "cell_type": "code", "execution_count": 55, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 678 }, "id": "cO7tLpcxjLh0", "outputId": "69147217-a12c-484f-9573-c63a3bf4175c" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Shape of the Dataset: (12194, 18)\n" ] }, { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"df\",\n \"rows\": 12194,\n \"fields\": [\n {\n \"column\": \"anime_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 11437,\n \"min\": 1,\n \"max\": 34527,\n \"num_unique_values\": 12194,\n \"samples\": [\n 3834,\n 32936,\n 8792\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"genres\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3256,\n \"samples\": [\n \"Mystery, Sci-Fi, Space\",\n \"Music, Sci-Fi\",\n \"Magic, Romance\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 12194,\n \"samples\": [\n \"Hoshi no Ko Chobin\",\n \"Gin no Guardian\",\n \"Madobe Nanami no Windows 7 de PC Jisaku Ouen Commercial!!\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"average_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 552,\n \"samples\": [\n \"4.06\",\n \"8.03\",\n \"7.32\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"overview\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 11689,\n \"samples\": [\n \"Serebii, a Legendary Pok\\u00e9mon known for its ability to traverse time, is hunted by an unnamed Pok\\u00e9mon poacher seeking to capture it. Yukinari, a young Pok\\u00e9mon trainer who enjoys drawing portraits of Pok\\u00e9mon, tries to protect Serebii after it stumbles upon him; but in the middle of its escape, both vanish without a trace.\\n\\nForty years later, ambitious Pok\\u00e9mon trainer Satoshi hopes to sight rare Pok\\u00e9mon around his local area. White, a boat driver, takes Satoshi to his village. Satoshi is accompanied by his three closest friends: Takeshi, a former gym leader training to be a great Pok\\u00e9mon breeder; Kasumi, a young girl wanting to become a skilled Water-type trainer; and Pikachu, Satoshi's Pok\\u00e9mon partner and first comrade.\\n\\nMeanwhile, Yukinari and Serebii reappear in Satoshi's present time and run into his group. Masked Lord Vicious, the strongest executive staff member of Team Rocket, desires to capture Serebii. Using the Dark Ball, a variant of Monster Ball that corrupts the Pok\\u00e9mon caught within and draws out their maximum power, Vicious can transform innocent Pok\\u00e9mon into powerful and frightening obstacles\\u2014including Serebii itself! Placed in a tough position, Satoshi, Yukinari, and their friends must work together to defeat Vicious and save Serebii and themselves.\",\n \"East Force meets West Force and all Hell breaks loose. The Solonoids, that lovable race of female warriors, are at it again, fighting amongst themselves. During a heated battle, however, it looks like the leaders of the two factions have hung their warriors out to dry. In the middle of all this fighting and chaos, East Force detects a transmission from an unidentified planet. The Gall Force gals leave their posts to go. Lufy, the West Force's Ace Pilot, who is after Rabby, follows them to the planet.\\n\\n(Source: AnimeNfo)\",\n \"Netto's father Yuuichirou Hikari has made a scientific breakthrough by introducing the \\\"synchro chips\\\". If an operator and his or her navi are in a special enviroment known as a \\\"dimensional area\\\", they can fuse together in the real world via a technique called \\\"cross fusion\\\"! Yuuichirou's first test subject, Misaki Gorou, attempts the process and sadly fails. Netto offers to try with Rockman, but his father forbids it. Cross Fusion puts enormous strain on the operator's health, and battling in the real world could mean death. \\n\\n(Source: Official Site)\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"type\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"TV\",\n \"Movie\",\n \"ONA\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"episodes\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 200,\n \"samples\": [\n \"93\",\n \"27\",\n \"110\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"producers\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3070,\n \"samples\": [\n \"AIC, Lantis, Media Factory, Pony Canyon, Rakuonsha, AT-X, KlockWorx, Ryukyu Asahi Broadcasting\",\n \"Tama Production, Tokyo MX\",\n \"KAGAYA Studio\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"licensors\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 253,\n \"samples\": [\n \"ADV Films, Media Blasters\",\n \"Funimation\",\n \"Central Park Media, Maiden Japan\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"studios\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 941,\n \"samples\": [\n \"Group TAC, Ginga Ya\",\n \"J.C.Staff, Production I.G\",\n \"drop\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"source\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 16,\n \"samples\": [\n \"Visual novel\",\n \"Manga\",\n \"Novel\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"anime_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"PG-13 - Teens 13 or older\",\n \"R - 17+ (violence & profanity)\",\n \"UNKNOWN\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rank\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8846,\n \"samples\": [\n \"12549\",\n \"7778\",\n \"7761\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"popularity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5211,\n \"min\": 1,\n \"max\": 19844,\n \"num_unique_values\": 10304,\n \"samples\": [\n 5880,\n 7617,\n 17889\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"favorites\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5821,\n \"min\": 0,\n \"max\": 217606,\n \"num_unique_values\": 1357,\n \"samples\": [\n 46556,\n 1136,\n 1570\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"scored by\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 6497,\n \"samples\": [\n \"19239\",\n \"1926\",\n \"300\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"members\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 191364,\n \"min\": 142,\n \"max\": 3744541,\n \"num_unique_values\": 8192,\n \"samples\": [\n 9390,\n 8376,\n 1998\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"image url\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 12135,\n \"samples\": [\n \"https://cdn.myanimelist.net/images/anime/1195/111544.jpg\",\n \"https://cdn.myanimelist.net/images/anime/6/26448.jpg\",\n \"https://cdn.myanimelist.net/images/anime/1405/112400.jpg\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe", "variable_name": "df" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
anime_idgenresnameaverage_ratingoverviewtypeepisodesproducerslicensorsstudiossourceanime_ratingrankpopularityfavoritesscored bymembersimage url
04181Drama, Fantasy, Romance, Slice of Life, Supern...Clannad: After Story8.93Clannad: After Story, the sequel to the critic...TV24Pony Canyon, TBS, Rakuonsha, Animation DoSentai FilmworksKyoto AnimationVisual novelPG-13 - Teens 13 or older19114689496397291149886https://cdn.myanimelist.net/images/anime/1299/...
128735Drama, Historical, JoseiShouwa Genroku Rakugo Shinjuu8.57Yotarou is a former yakuza member fresh out of...TV13Starchild Records, Mainichi Broadcasting Syste...UNKNOWNStudio DeenMangaPG-13 - Teens 13 or older93804571191359281445https://cdn.myanimelist.net/images/anime/1354/...
25205Action, Mystery, Romance, Supernatural, ThrillerKara no Kyoukai Movie 7: Satsujin Kousatsu (Go)8.39In February 1999, a string of murders has Shik...Movie1NotesAniplex of AmericaufotableLight novelR - 17+ (violence & profanity)18211152261108703200492https://cdn.myanimelist.net/images/anime/9/566...
3170Comedy, Drama, School, Shounen, SportsSlam Dunk8.54Hanamichi Sakuragi, infamous for his temper, m...TV101TV Asahi, AnimaxFlatiron Film Company, Geneon Entertainment USAToei AnimationMangaPG-13 - Teens 13 or older1087976879128920283226https://cdn.myanimelist.net/images/anime/12/86...
410162Josei, Slice of LifeUsagi Drop8.36Daikichi Kawachi is a 30-year-old bachelor wor...TV11Dentsu, Fuji TV, Toho, Tohokushinsha Film Corp...NIS America, Inc.Production I.GMangaPG-13 - Teens 13 or older2024255975237156479967https://cdn.myanimelist.net/images/anime/2/296...
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " anime_id genres \\\n", "0 4181 Drama, Fantasy, Romance, Slice of Life, Supern... \n", "1 28735 Drama, Historical, Josei \n", "2 5205 Action, Mystery, Romance, Supernatural, Thriller \n", "3 170 Comedy, Drama, School, Shounen, Sports \n", "4 10162 Josei, Slice of Life \n", "\n", " name average_rating \\\n", "0 Clannad: After Story 8.93 \n", "1 Shouwa Genroku Rakugo Shinjuu 8.57 \n", "2 Kara no Kyoukai Movie 7: Satsujin Kousatsu (Go) 8.39 \n", "3 Slam Dunk 8.54 \n", "4 Usagi Drop 8.36 \n", "\n", " overview type episodes \\\n", "0 Clannad: After Story, the sequel to the critic... TV 24 \n", "1 Yotarou is a former yakuza member fresh out of... TV 13 \n", "2 In February 1999, a string of murders has Shik... Movie 1 \n", "3 Hanamichi Sakuragi, infamous for his temper, m... TV 101 \n", "4 Daikichi Kawachi is a 30-year-old bachelor wor... TV 11 \n", "\n", " producers \\\n", "0 Pony Canyon, TBS, Rakuonsha, Animation Do \n", "1 Starchild Records, Mainichi Broadcasting Syste... \n", "2 Notes \n", "3 TV Asahi, Animax \n", "4 Dentsu, Fuji TV, Toho, Tohokushinsha Film Corp... \n", "\n", " licensors studios \\\n", "0 Sentai Filmworks Kyoto Animation \n", "1 UNKNOWN Studio Deen \n", "2 Aniplex of America ufotable \n", "3 Flatiron Film Company, Geneon Entertainment USA Toei Animation \n", "4 NIS America, Inc. Production I.G \n", "\n", " source anime_rating rank popularity favorites \\\n", "0 Visual novel PG-13 - Teens 13 or older 19 114 68949 \n", "1 Manga PG-13 - Teens 13 or older 93 804 5711 \n", "2 Light novel R - 17+ (violence & profanity) 182 1115 2261 \n", "3 Manga PG-13 - Teens 13 or older 108 797 6879 \n", "4 Manga PG-13 - Teens 13 or older 202 425 5975 \n", "\n", " scored by members image url \n", "0 639729 1149886 https://cdn.myanimelist.net/images/anime/1299/... \n", "1 91359 281445 https://cdn.myanimelist.net/images/anime/1354/... \n", "2 108703 200492 https://cdn.myanimelist.net/images/anime/9/566... \n", "3 128920 283226 https://cdn.myanimelist.net/images/anime/12/86... \n", "4 237156 479967 https://cdn.myanimelist.net/images/anime/2/296... " ] }, "execution_count": 55, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df = pd.read_csv(\"/content/Animes.csv\")\n", "print(\"Shape of the Dataset:\", df.shape)\n", "df.head()" ] }, { "cell_type": "code", "execution_count": 57, "metadata": { "id": "8-vUA38AMMCS" }, "outputs": [], "source": [ "df.dropna(inplace = True)" ] }, { "cell_type": "code", "execution_count": 58, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "9GvI7884MMCS", "outputId": "96c483b6-b506-489d-f0ef-f20cdf7fcfa2" }, "outputs": [ { "data": { "text/plain": [ "Index(['anime_id', 'genres', 'name', 'average_rating', 'overview', 'type',\n", " 'episodes', 'producers', 'licensors', 'studios', 'source',\n", " 'anime_rating', 'rank', 'popularity', 'favorites', 'scored by',\n", " 'members', 'image url'],\n", " dtype='object')" ] }, "execution_count": 58, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.columns" ] }, { "cell_type": "code", "execution_count": 59, "metadata": { "id": "syXBYlBqL9O_" }, "outputs": [], "source": [ "from sklearn.feature_extraction.text import TfidfVectorizer\n", "\n", "tfv = TfidfVectorizer(min_df=3,\n", " strip_accents='unicode', analyzer='word',token_pattern=r'\\w{1,}',\n", " ngram_range=(1, 3),\n", " stop_words = 'english')\n", "\n", "tfv_matrix = tfv.fit_transform(df['genres'])" ] }, { "cell_type": "code", "execution_count": 60, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "EDUM2vu7L99g", "outputId": "69816148-37e1-48f4-b6af-d69337b8bfeb" }, "outputs": [ { "data": { "text/plain": [ "(12133, 1552)" ] }, "execution_count": 60, "metadata": {}, "output_type": "execute_result" } ], "source": [ "tfv_matrix.shape" ] }, { "cell_type": "code", "execution_count": 61, "metadata": { "id": "0jVvemqYL95s" }, "outputs": [], "source": [ "from sklearn.metrics.pairwise import sigmoid_kernel\n", "sig = sigmoid_kernel(tfv_matrix, tfv_matrix)" ] }, { "cell_type": "code", "execution_count": 62, "metadata": { "id": "Qy4uip8AL92D" }, "outputs": [], "source": [ "indices = pd.Series(df.index, index=df['name']).drop_duplicates()" ] }, { "cell_type": "code", "execution_count": 63, "metadata": { "id": "ezJ6vl80L9xo" }, "outputs": [], "source": [ "def get_rec_sig(title, sig=sig,n_recommendations = 10):\n", " if title not in indices.index:\n", " print(f\"Anime title '{title}' not found in the dataset.\")\n", " # Get the index corresponding to original_title\n", " idx = indices[title]\n", "\n", " # Get the pairwsie similarity scores\n", " sig_scores = list(enumerate(sig[idx]))\n", "\n", " # Sort the movies\n", " sig_scores = sorted(sig_scores, key=lambda x: x[1], reverse=True)\n", "\n", " sig_scores = sig_scores[1:n_recommendations+1]\n", "\n", " # Movie indices\n", " anime_indices = [i[0] for i in sig_scores]\n", " return pd.DataFrame({'Anime name': df['name'].iloc[anime_indices].values,\n", " 'Rating': df['average_rating'].iloc[anime_indices].values})" ] }, { "cell_type": "code", "execution_count": 64, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 519 }, "id": "1sktZ0lWL9uO", "outputId": "a09176a6-dc66-48ec-9b63-a46f2afdec48" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"get_rec_sig('Naruto', n_recommendations=15)\",\n \"rows\": 15,\n \"fields\": [\n {\n \"column\": \"Anime name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 15,\n \"samples\": [\n \"Kyutai Panic Adventure!\",\n \"Ben-To\",\n \"Boruto: Naruto the Movie\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Rating\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 14,\n \"samples\": [\n \"7.49\",\n \"7.68\",\n \"7.4\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Anime nameRating
0Boruto: Naruto the Movie7.4
1Naruto7.99
2Naruto x UT7.37
3Boruto: Naruto the Movie - Naruto ga Hokage ni...7.33
4Naruto: Shippuuden Movie 4 - The Lost Tower7.42
5Naruto: Shippuuden Movie 3 - Hi no Ishi wo Tsu...7.33
6Naruto: Shippuuden - Sunny Side Battle7.56
7Naruto Soyokazeden Movie: Naruto to Mashin to ...6.96
8Battle Spirits: Ryuuko no Ken4.78
9Kyutai Panic Adventure!4.67
10Ranma ½: Akumu! Shunmin Kou7.49
11Ben-To7.2
12Naruto: Shippuuden Movie 6 - Road to Ninja7.68
13Rekka no Honoo7.34
14Naruto: Honoo no Chuunin Shiken! Naruto vs. Ko...7.17
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " Anime name Rating\n", "0 Boruto: Naruto the Movie 7.4\n", "1 Naruto 7.99\n", "2 Naruto x UT 7.37\n", "3 Boruto: Naruto the Movie - Naruto ga Hokage ni... 7.33\n", "4 Naruto: Shippuuden Movie 4 - The Lost Tower 7.42\n", "5 Naruto: Shippuuden Movie 3 - Hi no Ishi wo Tsu... 7.33\n", "6 Naruto: Shippuuden - Sunny Side Battle 7.56\n", "7 Naruto Soyokazeden Movie: Naruto to Mashin to ... 6.96\n", "8 Battle Spirits: Ryuuko no Ken 4.78\n", "9 Kyutai Panic Adventure! 4.67\n", "10 Ranma ½: Akumu! Shunmin Kou 7.49\n", "11 Ben-To 7.2\n", "12 Naruto: Shippuuden Movie 6 - Road to Ninja 7.68\n", "13 Rekka no Honoo 7.34\n", "14 Naruto: Honoo no Chuunin Shiken! Naruto vs. Ko... 7.17" ] }, "execution_count": 64, "metadata": {}, "output_type": "execute_result" } ], "source": [ "get_rec_sig('Naruto', n_recommendations=15)" ] }, { "cell_type": "markdown", "metadata": { "id": "9JTNMQdTWDQV" }, "source": [ "##### b. Linear Kernel" ] }, { "cell_type": "code", "execution_count": 65, "metadata": { "id": "6D9pKHkBWDQX" }, "outputs": [], "source": [ "from sklearn.metrics.pairwise import linear_kernel\n", "\n", "lin = linear_kernel(tfv_matrix, tfv_matrix)" ] }, { "cell_type": "code", "execution_count": 66, "metadata": { "id": "xFlCrJOeWDQX" }, "outputs": [], "source": [ "indices = pd.Series(df.index, index=df['name']).drop_duplicates()" ] }, { "cell_type": "code", "execution_count": 67, "metadata": { "id": "nikXuR1fWDQY" }, "outputs": [], "source": [ "def get_rec_lin(title, lin=lin,n_recommendations = 10):\n", " if title not in indices.index:\n", " print(f\"Anime title '{title}' not found in the dataset.\")\n", " # Get the index corresponding to original_title\n", " idx = indices[title]\n", "\n", " # Get the pairwsie similarity scores\n", " lin_scores = list(enumerate(lin[idx]))\n", "\n", " # Sort the movies\n", " lin_scores = sorted(lin_scores, key=lambda x: x[1], reverse=True)\n", "\n", " lin_scores = lin_scores[1:n_recommendations+1]\n", "\n", " # Movie indices\n", " anime_indices = [i[0] for i in lin_scores]\n", " return pd.DataFrame({'Anime name': df['name'].iloc[anime_indices].values,\n", " 'Rating': df['average_rating'].iloc[anime_indices].values})" ] }, { "cell_type": "code", "execution_count": 68, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 519 }, "id": "sRJM0gdqWDQY", "outputId": "8feb0c8b-9193-4be8-f731-6e81c346e9da" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"get_rec_lin('Naruto', n_recommendations=15)\",\n \"rows\": 15,\n \"fields\": [\n {\n \"column\": \"Anime name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 15,\n \"samples\": [\n \"Kyutai Panic Adventure!\",\n \"Ben-To\",\n \"Boruto: Naruto the Movie\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Rating\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 14,\n \"samples\": [\n \"7.49\",\n \"7.68\",\n \"7.4\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Anime nameRating
0Boruto: Naruto the Movie7.4
1Naruto7.99
2Naruto x UT7.37
3Boruto: Naruto the Movie - Naruto ga Hokage ni...7.33
4Naruto: Shippuuden Movie 4 - The Lost Tower7.42
5Naruto: Shippuuden Movie 3 - Hi no Ishi wo Tsu...7.33
6Naruto: Shippuuden - Sunny Side Battle7.56
7Naruto Soyokazeden Movie: Naruto to Mashin to ...6.96
8Battle Spirits: Ryuuko no Ken4.78
9Kyutai Panic Adventure!4.67
10Ranma ½: Akumu! Shunmin Kou7.49
11Ben-To7.2
12Naruto: Shippuuden Movie 6 - Road to Ninja7.68
13Rekka no Honoo7.34
14Naruto: Honoo no Chuunin Shiken! Naruto vs. Ko...7.17
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " Anime name Rating\n", "0 Boruto: Naruto the Movie 7.4\n", "1 Naruto 7.99\n", "2 Naruto x UT 7.37\n", "3 Boruto: Naruto the Movie - Naruto ga Hokage ni... 7.33\n", "4 Naruto: Shippuuden Movie 4 - The Lost Tower 7.42\n", "5 Naruto: Shippuuden Movie 3 - Hi no Ishi wo Tsu... 7.33\n", "6 Naruto: Shippuuden - Sunny Side Battle 7.56\n", "7 Naruto Soyokazeden Movie: Naruto to Mashin to ... 6.96\n", "8 Battle Spirits: Ryuuko no Ken 4.78\n", "9 Kyutai Panic Adventure! 4.67\n", "10 Ranma ½: Akumu! Shunmin Kou 7.49\n", "11 Ben-To 7.2\n", "12 Naruto: Shippuuden Movie 6 - Road to Ninja 7.68\n", "13 Rekka no Honoo 7.34\n", "14 Naruto: Honoo no Chuunin Shiken! Naruto vs. Ko... 7.17" ] }, "execution_count": 68, "metadata": {}, "output_type": "execute_result" } ], "source": [ "get_rec_lin('Naruto', n_recommendations=15)" ] }, { "cell_type": "markdown", "metadata": { "id": "rtPCq8UMkyE-" }, "source": [ "#### 2) Using count vectorizer" ] }, { "cell_type": "code", "execution_count": 69, "metadata": { "id": "DY4p2FYCk1PT" }, "outputs": [], "source": [ "from sklearn.feature_extraction.text import CountVectorizer\n", "from sklearn.metrics.pairwise import cosine_similarity\n", "# Initialize CountVectorizer\n", "count_vectorizer = CountVectorizer(analyzer='word', token_pattern=r'\\w{1,}', ngram_range=(1, 3), stop_words='english',max_features=5000)\n", "\n", "# Fit and transform the genre data\n", "count_matrix = count_vectorizer.fit_transform(df['genres'])\n", "\n", "# Compute the cosine similarity matrix\n", "cosine_sim = cosine_similarity(count_matrix, count_matrix)\n", "\n", "# Create a reverse mapping of anime titles to indices\n", "indices = pd.Series(df.index, index=df['name']).drop_duplicates()\n", "\n", "# Function to get recommendations\n", "def get_rec(title, cosine_sim=cosine_sim, n_recommendations=5):\n", " if title not in indices.index:\n", " return f\"Anime title '{title}' not found in the dataset.\"\n", "\n", " # Get the index of the anime that matches the title\n", " idx = indices[title]\n", "\n", " # Get the pairwise similarity scores of all anime with that anime\n", " sim_scores = list(enumerate(cosine_sim[idx]))\n", "\n", " # Sort the anime based on the similarity scores\n", " sim_scores = sorted(sim_scores, key=lambda x: x[1], reverse=True)\n", "\n", " # Get the scores of the most similar anime (excluding the input anime itself)\n", " sim_scores = sim_scores[1:n_recommendations+1]\n", "\n", " # Get the anime indices\n", " anime_indices = [i[0] for i in sim_scores]\n", "\n", " # Return the top n most similar anime\n", " return pd.DataFrame({'Anime name': df['name'].iloc[anime_indices].values,\n", " 'Rating': df['average_rating'].iloc[anime_indices].values})" ] }, { "cell_type": "code", "execution_count": 70, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 519 }, "id": "PXxkJk2qleU0", "outputId": "fab7b7ec-d72b-4d22-e714-b4e14be2ce00" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"get_rec( \\\"Naruto\\\", n_recommendations=15)\",\n \"rows\": 15,\n \"fields\": [\n {\n \"column\": \"Anime name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 15,\n \"samples\": [\n \"Kyutai Panic Adventure!\",\n \"Naruto: Shippuuden Movie 6 - Road to Ninja\",\n \"Boruto: Naruto the Movie\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"Rating\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 14,\n \"samples\": [\n \"7.49\",\n \"7.34\",\n \"7.4\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Anime nameRating
0Boruto: Naruto the Movie7.4
1Naruto7.99
2Naruto x UT7.37
3Boruto: Naruto the Movie - Naruto ga Hokage ni...7.33
4Naruto: Shippuuden Movie 4 - The Lost Tower7.42
5Naruto: Shippuuden Movie 3 - Hi no Ishi wo Tsu...7.33
6Naruto: Shippuuden - Sunny Side Battle7.56
7Naruto Soyokazeden Movie: Naruto to Mashin to ...6.96
8Battle Spirits: Ryuuko no Ken4.78
9Kyutai Panic Adventure!4.67
10Ranma ½: Akumu! Shunmin Kou7.49
11Naruto: Shippuuden Movie 6 - Road to Ninja7.68
12Rekka no Honoo7.34
13Naruto: Honoo no Chuunin Shiken! Naruto vs. Ko...7.17
14Street Fighter Zero The Animation6.51
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " Anime name Rating\n", "0 Boruto: Naruto the Movie 7.4\n", "1 Naruto 7.99\n", "2 Naruto x UT 7.37\n", "3 Boruto: Naruto the Movie - Naruto ga Hokage ni... 7.33\n", "4 Naruto: Shippuuden Movie 4 - The Lost Tower 7.42\n", "5 Naruto: Shippuuden Movie 3 - Hi no Ishi wo Tsu... 7.33\n", "6 Naruto: Shippuuden - Sunny Side Battle 7.56\n", "7 Naruto Soyokazeden Movie: Naruto to Mashin to ... 6.96\n", "8 Battle Spirits: Ryuuko no Ken 4.78\n", "9 Kyutai Panic Adventure! 4.67\n", "10 Ranma ½: Akumu! Shunmin Kou 7.49\n", "11 Naruto: Shippuuden Movie 6 - Road to Ninja 7.68\n", "12 Rekka no Honoo 7.34\n", "13 Naruto: Honoo no Chuunin Shiken! Naruto vs. Ko... 7.17\n", "14 Street Fighter Zero The Animation 6.51" ] }, "execution_count": 70, "metadata": {}, "output_type": "execute_result" } ], "source": [ "get_rec( \"Naruto\", n_recommendations=15)" ] }, { "cell_type": "markdown", "metadata": { "id": "ny0Ub9tJmZk6" }, "source": [ "As you can see both Count Vectorizer and TF-IDF are giving same recommendations. We will lean towards TfidfVectorizer for more accurate and meaningful recommendations, especially in content-based systems that handle large or complex textual data. CountVectorizer might be used in scenarios where quick and simple solutions are sufficient or for smaller datasets." ] }, { "cell_type": "markdown", "metadata": { "id": "QSPSoR1RH9ZE" }, "source": [ "## Popularity-Based Filtering" ] }, { "cell_type": "code", "execution_count": 71, "metadata": { "id": "5r8qPk9PzL_S" }, "outputs": [], "source": [ "import seaborn as sns" ] }, { "cell_type": "code", "execution_count": 72, "metadata": { "id": "ppgmlYpA20D3" }, "outputs": [], "source": [ "df = df_anime.copy()" ] }, { "cell_type": "code", "execution_count": 73, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 340 }, "id": "GK_FQCX53Hth", "outputId": "9a55ecdc-302e-409f-d68d-f39ba7437943" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"df\",\n \"rows\": 12194,\n \"fields\": [\n {\n \"column\": \"anime_id\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 11437,\n \"min\": 1,\n \"max\": 34527,\n \"num_unique_values\": 12194,\n \"samples\": [\n 3834,\n 32936,\n 8792\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"genres\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3256,\n \"samples\": [\n \"Mystery, Sci-Fi, Space\",\n \"Music, Sci-Fi\",\n \"Magic, Romance\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 12194,\n \"samples\": [\n \"Hoshi no Ko Chobin\",\n \"Gin no Guardian\",\n \"Madobe Nanami no Windows 7 de PC Jisaku Ouen Commercial!!\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"average_rating\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.9422757691869167,\n \"min\": 1.85,\n \"max\": 9.1,\n \"num_unique_values\": 551,\n \"samples\": [\n 4.06,\n 8.03,\n 7.32\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"overview\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 11689,\n \"samples\": [\n \"Serebii, a Legendary Pok\\u00e9mon known for its ability to traverse time, is hunted by an unnamed Pok\\u00e9mon poacher seeking to capture it. Yukinari, a young Pok\\u00e9mon trainer who enjoys drawing portraits of Pok\\u00e9mon, tries to protect Serebii after it stumbles upon him; but in the middle of its escape, both vanish without a trace.\\n\\nForty years later, ambitious Pok\\u00e9mon trainer Satoshi hopes to sight rare Pok\\u00e9mon around his local area. White, a boat driver, takes Satoshi to his village. Satoshi is accompanied by his three closest friends: Takeshi, a former gym leader training to be a great Pok\\u00e9mon breeder; Kasumi, a young girl wanting to become a skilled Water-type trainer; and Pikachu, Satoshi's Pok\\u00e9mon partner and first comrade.\\n\\nMeanwhile, Yukinari and Serebii reappear in Satoshi's present time and run into his group. Masked Lord Vicious, the strongest executive staff member of Team Rocket, desires to capture Serebii. Using the Dark Ball, a variant of Monster Ball that corrupts the Pok\\u00e9mon caught within and draws out their maximum power, Vicious can transform innocent Pok\\u00e9mon into powerful and frightening obstacles\\u2014including Serebii itself! Placed in a tough position, Satoshi, Yukinari, and their friends must work together to defeat Vicious and save Serebii and themselves.\",\n \"East Force meets West Force and all Hell breaks loose. The Solonoids, that lovable race of female warriors, are at it again, fighting amongst themselves. During a heated battle, however, it looks like the leaders of the two factions have hung their warriors out to dry. In the middle of all this fighting and chaos, East Force detects a transmission from an unidentified planet. The Gall Force gals leave their posts to go. Lufy, the West Force's Ace Pilot, who is after Rabby, follows them to the planet.\\n\\n(Source: AnimeNfo)\",\n \"Netto's father Yuuichirou Hikari has made a scientific breakthrough by introducing the \\\"synchro chips\\\". If an operator and his or her navi are in a special enviroment known as a \\\"dimensional area\\\", they can fuse together in the real world via a technique called \\\"cross fusion\\\"! Yuuichirou's first test subject, Misaki Gorou, attempts the process and sadly fails. Netto offers to try with Rockman, but his father forbids it. Cross Fusion puts enormous strain on the operator's health, and battling in the real world could mean death. \\n\\n(Source: Official Site)\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"type\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"TV\",\n \"Movie\",\n \"ONA\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"episodes\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 200,\n \"samples\": [\n \"93\",\n \"27\",\n \"110\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"producers\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3070,\n \"samples\": [\n \"AIC, Lantis, Media Factory, Pony Canyon, Rakuonsha, AT-X, KlockWorx, Ryukyu Asahi Broadcasting\",\n \"Tama Production, Tokyo MX\",\n \"KAGAYA Studio\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"licensors\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 253,\n \"samples\": [\n \"ADV Films, Media Blasters\",\n \"Funimation\",\n \"Central Park Media, Maiden Japan\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"studios\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 941,\n \"samples\": [\n \"Group TAC, Ginga Ya\",\n \"J.C.Staff, Production I.G\",\n \"drop\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"source\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 16,\n \"samples\": [\n \"Visual novel\",\n \"Manga\",\n \"Novel\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"anime_rating\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 7,\n \"samples\": [\n \"PG-13 - Teens 13 or older\",\n \"R - 17+ (violence & profanity)\",\n \"UNKNOWN\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rank\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8846,\n \"samples\": [\n \"12549\",\n \"7778\",\n \"7761\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"popularity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5211,\n \"min\": 1,\n \"max\": 19844,\n \"num_unique_values\": 10304,\n \"samples\": [\n 5880,\n 7617,\n 17889\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"favorites\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5821,\n \"min\": 0,\n \"max\": 217606,\n \"num_unique_values\": 1357,\n \"samples\": [\n 46556,\n 1136,\n 1570\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"scored by\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 6497,\n \"samples\": [\n \"19239\",\n \"1926\",\n \"300\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"members\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 191364,\n \"min\": 142,\n \"max\": 3744541,\n \"num_unique_values\": 8192,\n \"samples\": [\n 9390,\n 8376,\n 1998\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"image url\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 12135,\n \"samples\": [\n \"https://cdn.myanimelist.net/images/anime/1195/111544.jpg\",\n \"https://cdn.myanimelist.net/images/anime/6/26448.jpg\",\n \"https://cdn.myanimelist.net/images/anime/1405/112400.jpg\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe", "variable_name": "df" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
anime_idgenresnameaverage_ratingoverviewtypeepisodesproducerslicensorsstudiossourceanime_ratingrankpopularityfavoritesscored bymembersimage url
04181Drama, Fantasy, Romance, Slice of Life, Supern...Clannad: After Story8.93Clannad: After Story, the sequel to the critic...TV24Pony Canyon, TBS, Rakuonsha, Animation DoSentai FilmworksKyoto AnimationVisual novelPG-13 - Teens 13 or older19114689496397291149886https://cdn.myanimelist.net/images/anime/1299/...
128735Drama, Historical, JoseiShouwa Genroku Rakugo Shinjuu8.57Yotarou is a former yakuza member fresh out of...TV13Starchild Records, Mainichi Broadcasting Syste...UNKNOWNStudio DeenMangaPG-13 - Teens 13 or older93804571191359281445https://cdn.myanimelist.net/images/anime/1354/...
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " anime_id genres \\\n", "0 4181 Drama, Fantasy, Romance, Slice of Life, Supern... \n", "1 28735 Drama, Historical, Josei \n", "\n", " name average_rating \\\n", "0 Clannad: After Story 8.93 \n", "1 Shouwa Genroku Rakugo Shinjuu 8.57 \n", "\n", " overview type episodes \\\n", "0 Clannad: After Story, the sequel to the critic... TV 24 \n", "1 Yotarou is a former yakuza member fresh out of... TV 13 \n", "\n", " producers licensors \\\n", "0 Pony Canyon, TBS, Rakuonsha, Animation Do Sentai Filmworks \n", "1 Starchild Records, Mainichi Broadcasting Syste... UNKNOWN \n", "\n", " studios source anime_rating rank popularity \\\n", "0 Kyoto Animation Visual novel PG-13 - Teens 13 or older 19 114 \n", "1 Studio Deen Manga PG-13 - Teens 13 or older 93 804 \n", "\n", " favorites scored by members \\\n", "0 68949 639729 1149886 \n", "1 5711 91359 281445 \n", "\n", " image url \n", "0 https://cdn.myanimelist.net/images/anime/1299/... \n", "1 https://cdn.myanimelist.net/images/anime/1354/... " ] }, "execution_count": 73, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df.head(2)" ] }, { "cell_type": "markdown", "metadata": { "id": "CYqncb_s3t5n" }, "source": [ "#### 1) Top n animes based on Popularity" ] }, { "cell_type": "code", "execution_count": 74, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 363 }, "id": "6f9X0QLq3NhC", "outputId": "f9ec7fbe-4fa6-4188-8906-cbeae62e48f2" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"popular_animes(n=10)\",\n \"rows\": 10,\n \"fields\": [\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 10,\n \"samples\": [\n \"Hunter x Hunter (2011)\",\n \"Death Note\",\n \"Boku no Hero Academia\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"popularity\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 3,\n \"min\": 1,\n \"max\": 11,\n \"num_unique_values\": 10,\n \"samples\": [\n 10,\n 2,\n 6\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namepopularity
29Shingeki no Kyojin1
176Death Note2
66Fullmetal Alchemist: Brotherhood3
39One Punch Man4
508Sword Art Online5
57Boku no Hero Academia6
726Naruto8
855Tokyo Ghoul9
38Hunter x Hunter (2011)10
102Kimi no Na wa.11
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " name popularity\n", "29 Shingeki no Kyojin 1\n", "176 Death Note 2\n", "66 Fullmetal Alchemist: Brotherhood 3\n", "39 One Punch Man 4\n", "508 Sword Art Online 5\n", "57 Boku no Hero Academia 6\n", "726 Naruto 8\n", "855 Tokyo Ghoul 9\n", "38 Hunter x Hunter (2011) 10\n", "102 Kimi no Na wa. 11" ] }, "execution_count": 74, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def popular_animes(n=10):\n", " sorted_df = df.sort_values(by=['popularity'], ascending=[ True])\n", " top_n_anime = sorted_df.head(n)\n", " return top_n_anime[['name', 'popularity']]\n", "\n", "popular_animes(n=10)" ] }, { "cell_type": "markdown", "metadata": { "id": "OCh8zH5O3-SG" }, "source": [ "#### 2) Top n animes based on Rank" ] }, { "cell_type": "code", "execution_count": 75, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 363 }, "id": "O1sVeB6R5oWY", "outputId": "ab8c1927-073c-4ba9-f856-58fe805b69b4" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"top_ranked_animes(n=10)\",\n \"rows\": 10,\n \"fields\": [\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 10,\n \"samples\": [\n \"Koe no Katachi\",\n \"Gintama\\u00b0\",\n \"Ginga Eiyuu Densetsu\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"rank\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 5.400617248673217,\n \"min\": 3.0,\n \"max\": 19.0,\n \"num_unique_values\": 10,\n \"samples\": [\n 17.0,\n 4.0,\n 12.0\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namerank
138Steins;Gate3.0
103Gintama°4.0
49Gintama'8.0
38Hunter x Hunter (2011)10.0
58Gintama': Enchousen11.0
90Ginga Eiyuu Densetsu12.0
10580Gintama.15.0
12Gintama16.0
165Koe no Katachi17.0
0Clannad: After Story19.0
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " name rank\n", "138 Steins;Gate 3.0\n", "103 Gintama° 4.0\n", "49 Gintama' 8.0\n", "38 Hunter x Hunter (2011) 10.0\n", "58 Gintama': Enchousen 11.0\n", "90 Ginga Eiyuu Densetsu 12.0\n", "10580 Gintama. 15.0\n", "12 Gintama 16.0\n", "165 Koe no Katachi 17.0\n", "0 Clannad: After Story 19.0" ] }, "execution_count": 75, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def top_ranked_animes(n=10):\n", " df['rank'] = df['rank'].replace('UNKNOWN', np.nan).astype(float)\n", " df_filtered = df[df['rank'] > 1]\n", " sorted_df = df_filtered.sort_values(by=['rank'], ascending=True)\n", " top_n_anime = sorted_df.head(n)\n", " return top_n_anime[['name', 'rank']]\n", "top_ranked_animes(n=10)" ] }, { "cell_type": "markdown", "metadata": { "id": "rwVX3Htf6AIn" }, "source": [ "#### 3) Top n animes based on average rated" ] }, { "cell_type": "code", "execution_count": 76, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 363 }, "id": "O33iCVxQ5_iW", "outputId": "a95d0e04-15bf-43b2-e679-8d49247b8a30" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"overall_top_rated_animes(n=10)\",\n \"rows\": 10,\n \"fields\": [\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 10,\n \"samples\": [\n \"Gintama\",\n \"Steins;Gate\",\n \"Gintama': Enchousen\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"average_rating\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.053499740393970416,\n \"min\": 8.94,\n \"max\": 9.1,\n \"num_unique_values\": 8,\n \"samples\": [\n 9.07,\n 9.02,\n 9.1\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
nameaverage_rating
66Fullmetal Alchemist: Brotherhood9.10
138Steins;Gate9.07
103Gintama°9.06
49Gintama'9.04
38Hunter x Hunter (2011)9.04
58Gintama': Enchousen9.03
90Ginga Eiyuu Densetsu9.02
10580Gintama.8.98
12Gintama8.94
165Koe no Katachi8.94
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " name average_rating\n", "66 Fullmetal Alchemist: Brotherhood 9.10\n", "138 Steins;Gate 9.07\n", "103 Gintama° 9.06\n", "49 Gintama' 9.04\n", "38 Hunter x Hunter (2011) 9.04\n", "58 Gintama': Enchousen 9.03\n", "90 Ginga Eiyuu Densetsu 9.02\n", "10580 Gintama. 8.98\n", "12 Gintama 8.94\n", "165 Koe no Katachi 8.94" ] }, "execution_count": 76, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def overall_top_rated_animes(n = 10):\n", " sorted_df = df.sort_values(by=['average_rating'], ascending=False)\n", " top_n_anime = sorted_df.head(n)\n", " return top_n_anime[['name', 'average_rating']]\n", "overall_top_rated_animes(n=10)" ] }, { "cell_type": "markdown", "metadata": { "id": "e9gaqzXQ7VrD" }, "source": [ "#### 4) Top n Favorite animes" ] }, { "cell_type": "code", "execution_count": 77, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 363 }, "id": "BHeyp1aw7W0Y", "outputId": "741d2e14-25c1-4267-db57-cf2a0e8da590" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"favorite_animes(n=10)\",\n \"rows\": 10,\n \"fields\": [\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 10,\n \"samples\": [\n \"Neon Genesis Evangelion\",\n \"Hunter x Hunter (2011)\",\n \"Shingeki no Kyojin\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"favorites\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 48453,\n \"min\": 87268,\n \"max\": 217606,\n \"num_unique_values\": 10,\n \"samples\": [\n 100638,\n 200265,\n 163844\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namefavorites
66Fullmetal Alchemist: Brotherhood217606
38Hunter x Hunter (2011)200265
145One Piece198986
138Steins;Gate182964
176Death Note167586
29Shingeki no Kyojin163844
470Naruto: Shippuuden107735
141Code Geass: Hangyaku no Lelouch105379
325Neon Genesis Evangelion100638
102Kimi no Na wa.87268
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " name favorites\n", "66 Fullmetal Alchemist: Brotherhood 217606\n", "38 Hunter x Hunter (2011) 200265\n", "145 One Piece 198986\n", "138 Steins;Gate 182964\n", "176 Death Note 167586\n", "29 Shingeki no Kyojin 163844\n", "470 Naruto: Shippuuden 107735\n", "141 Code Geass: Hangyaku no Lelouch 105379\n", "325 Neon Genesis Evangelion 100638\n", "102 Kimi no Na wa. 87268" ] }, "execution_count": 77, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def favorite_animes(n=10):\n", " sorted_df = df.sort_values(by=['favorites'], ascending=False)\n", " top_n_anime = sorted_df.head(n)\n", " return top_n_anime[['name', 'favorites']]\n", "\n", "favorite_animes(n=10)" ] }, { "cell_type": "markdown", "metadata": { "id": "6UKaSPMs79pd" }, "source": [ "#### 5) Top n anime based on Members" ] }, { "cell_type": "code", "execution_count": 78, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 363 }, "id": "xbkce_I677v2", "outputId": "413398d3-1199-42c3-afcd-c0acc1349ef2" }, "outputs": [ { "data": { "application/vnd.google.colaboratory.intrinsic+json": { "summary": "{\n \"name\": \"top_animes_members(10)\",\n \"rows\": 10,\n \"fields\": [\n {\n \"column\": \"name\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 10,\n \"samples\": [\n \"Hunter x Hunter (2011)\",\n \"Death Note\",\n \"Boku no Hero Academia\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"members\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 416349,\n \"min\": 2597495,\n \"max\": 3744541,\n \"num_unique_values\": 10,\n \"samples\": [\n 2656870,\n 3713315,\n 2882333\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}", "type": "dataframe" }, "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
namemembers
29Shingeki no Kyojin3744541
176Death Note3713315
66Fullmetal Alchemist: Brotherhood3176556
39One Punch Man3058666
508Sword Art Online2951821
57Boku no Hero Academia2882333
726Naruto2717330
855Tokyo Ghoul2699241
38Hunter x Hunter (2011)2656870
102Kimi no Na wa.2597495
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "text/plain": [ " name members\n", "29 Shingeki no Kyojin 3744541\n", "176 Death Note 3713315\n", "66 Fullmetal Alchemist: Brotherhood 3176556\n", "39 One Punch Man 3058666\n", "508 Sword Art Online 2951821\n", "57 Boku no Hero Academia 2882333\n", "726 Naruto 2717330\n", "855 Tokyo Ghoul 2699241\n", "38 Hunter x Hunter (2011) 2656870\n", "102 Kimi no Na wa. 2597495" ] }, "execution_count": 78, "metadata": {}, "output_type": "execute_result" } ], "source": [ "def top_animes_members(n=10):\n", " sorted_df = df.sort_values(by=['members'], ascending=False)\n", " top_n_anime = sorted_df.head(n)\n", " return top_n_anime[['name', 'members']]\n", "top_animes_members(10)" ] }, { "cell_type": "markdown", "metadata": { "id": "FYHmuNupISw_" }, "source": [ "#### 6) Most popular anime based on members." ] }, { "cell_type": "code", "execution_count": 79, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "ydCt6MsWmmBO", "outputId": "ab9bbbe6-49f1-4e5f-bbc4-9f4105991c81" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Top 10 unique popular animes based on members:\n", "1. Shingeki no Kyojin\n", "2. Death Note\n", "3. Fullmetal Alchemist: Brotherhood\n", "4. One Punch Man\n", "5. Sword Art Online\n", "6. Boku no Hero Academia\n", "7. Naruto\n", "8. Tokyo Ghoul\n", "9. Hunter x Hunter (2011)\n", "10. Kimi no Na wa.\n" ] } ], "source": [ "def popular_anime_among_members(n=10):\n", " popular_animes = df.sort_values(by=['members', 'average_rating'], ascending=[False, False]).drop_duplicates(subset='name')['name'].head(n).tolist()\n", "\n", " # Print the top 'n' unique popular animes\n", " print(f\"Top {n} unique popular animes based on members:\")\n", " for i, anime in enumerate(popular_animes, 1):\n", " print(f\"{i}. {anime}\")\n", " return None\n", "\n", "popular_anime_among_members(n=10)" ] }, { "cell_type": "code", "execution_count": 80, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "fKAM4gRTIcQ4", "outputId": "f3f8c7de-e5b4-4856-84dc-5884f8cdd9d7" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Top 10 animes based on average rating:\n", "1. Fullmetal Alchemist: Brotherhood - 9.1\n", "2. Steins;Gate - 9.07\n", "3. Gintama° - 9.06\n", "4. Hunter x Hunter (2011) - 9.04\n", "5. Gintama' - 9.04\n", "6. Gintama': Enchousen - 9.03\n", "7. Ginga Eiyuu Densetsu - 9.02\n", "8. Gintama. - 8.98\n", "9. Koe no Katachi - 8.94\n", "10. Gintama - 8.94\n" ] } ], "source": [ "def top_avg_rated(n=10):\n", "\n", " # Get the top 'n' animes based on average rating, ensuring unique names\n", " top_animes = (\n", " df_merged.drop_duplicates(subset='name')\n", " .nlargest(n, 'average_rating')[['name', 'average_rating']]\n", " )\n", "\n", " # Print the top 'n' animes\n", " print(f\"Top {n} animes based on average rating:\")\n", " for i, row in enumerate(top_animes.itertuples(index=False), 1):\n", " print(f\"{i}. {row.name} - {row.average_rating}\")\n", " return None\n", "\n", "# Example usage\n", "top_avg_rated(n=10)\n" ] } ], "metadata": { "accelerator": "TPU", "colab": { "gpuType": "V28", "provenance": [], "toc_visible": true }, "kernelspec": { "display_name": "Python 3", "name": "python3" }, "language_info": { "name": "python" } }, "nbformat": 4, "nbformat_minor": 0 }