10/lekce.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Lekce 10\n",
    "\n",
    "## Systémy na doporučování obsahu\n",
    "\n",
    "Systémy na doporučování obsahu (*recommendation engine*) se snaží z dostupných možností vybrat tu nejzajímavější pro konkrétního uživatele. Systémy na doporučování obsahu jsou běžnou součástí řady služeb, například e-shopů, streamovacích služeb, sociálních sítí, pracovních portálů. Pomocí těchto systémů se poskytovatelé snaží udržet uživatele co nejdéle na jejich platformě (aby jim mohli zobrazit co nejvíce reklamy), nabídnout produkt, který si uživatel s největší pravděpodobností koupí atd.\n",
    "\n",
    "Systémy na doporučování obsahu mohou fungovat na dvou principech.\n",
    "\n",
    "Filtrování na základě obsahu (*content based filtering*) využívá data o jednotlivých položkách (např. filmech, knihách atd.). Jestliže se uživateli (uživatelce) líbí nějaký obsah, systém mu nabídne obsah, který je nejvíce podobný. Při vývoji tohoto systému tedy řešíme podobnost jednotlivých položek (např. na základě popisu).\n",
    "\n",
    "![content-based.png](content-based.png)\n",
    "\n",
    "K vývoji nám stačí data o obsahu, tento přístup je tedy možné použít pro začínající službu. V některých případech může ale vést k \"očividným\" výsledkům, například pokud se někomu líbí první díl Star Wars, je mu nabídnutý druhý díl atd.\n",
    "\n",
    "Kolaborativní filtrování (*collaborative filtering*) využívá data z chování uživatelů. Systém nějakému uživateli (uživatelce) vybírá takový obsah, který se líbil jemu podobným uživatelům (uživatelkám). Při vývoji tohoto systému tedy řešíme podobnost preferencí uživatelů (uživatelek) dané služby.\n",
    "\n",
    "Pro vývoj je potřeba mít k dispozici dostatek dat o chování uživatelů. To může být problém například u nově vytvořených služeb, nových příspěvků na sociálních sítích, nových nebo nepřihlášených uživatelů/uživatelek atd. Pokud ale data máme k dispozici, může tento přístup generovat poměrně zajímavé výsledky.\n",
    "\n",
    "![collaborative-filtering](collaborative-filtering.png)\n",
    "\n",
    "Data mohou být založená na tzv. explicitním nebo implicitním feedbacku (*explicit feedback*, *implicit feedback*). Příkladem explicitního feedbacku je použití tlačítka \"To se mi líbí\", napsání pozitivní recenze atd. Příkladem implicitní vazby je dokoukání (nebo naopak nedokokání) filmu do konce, opětovné zhlédnutí atd.\n",
    "\n",
    "### Filtrování na základě obsahu\n",
    "\n",
    "V této části si vyzkoušíme filtrování na základě popisu. Využijeme dataset s 5000 filmy ze serveru [imdb.com](https://imdb.com). Pozor, jedná se o jiný dataset, než jaký jsme používali v minulé lekci. Dataset je uložený v souboru [tmdb_5000_movies.csv](tmdb_5000_movies.csv). Budeme využívat sloupec `overview`, kde jsou uložené slovní popisy filmů. Filmy, které nemají slovní popis, z dat vyřadíme."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>budget</th>\n",
       "      <th>gen_Action</th>\n",
       "      <th>gen_Adventure</th>\n",
       "      <th>gen_Animation</th>\n",
       "      <th>gen_Comedy</th>\n",
       "      <th>gen_Crime</th>\n",
       "      <th>gen_Documentary</th>\n",
       "      <th>gen_Drama</th>\n",
       "      <th>gen_Family</th>\n",
       "      <th>...</th>\n",
       "      <th>production_companies</th>\n",
       "      <th>production_countries</th>\n",
       "      <th>release_date</th>\n",
       "      <th>revenue</th>\n",
       "      <th>runtime</th>\n",
       "      <th>spoken_languages</th>\n",
       "      <th>status</th>\n",
       "      <th>tagline</th>\n",
       "      <th>vote_average</th>\n",
       "      <th>vote_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Avatar</td>\n",
       "      <td>237000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>[{\"name\": \"Ingenious Film Partners\", \"id\": 289...</td>\n",
       "      <td>[{\"iso_3166_1\": \"US\", \"name\": \"United States o...</td>\n",
       "      <td>2009-12-10</td>\n",
       "      <td>2787965087</td>\n",
       "      <td>162.0</td>\n",
       "      <td>[{\"iso_639_1\": \"en\", \"name\": \"English\"}, {\"iso...</td>\n",
       "      <td>Released</td>\n",
       "      <td>Enter the World of Pandora.</td>\n",
       "      <td>7.2</td>\n",
       "      <td>11800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Pirates of the Caribbean: At World's End</td>\n",
       "      <td>300000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>[{\"name\": \"Walt Disney Pictures\", \"id\": 2}, {\"...</td>\n",
       "      <td>[{\"iso_3166_1\": \"US\", \"name\": \"United States o...</td>\n",
       "      <td>2007-05-19</td>\n",
       "      <td>961000000</td>\n",
       "      <td>169.0</td>\n",
       "      <td>[{\"iso_639_1\": \"en\", \"name\": \"English\"}]</td>\n",
       "      <td>Released</td>\n",
       "      <td>At the end of the world, the adventure begins.</td>\n",
       "      <td>6.9</td>\n",
       "      <td>4500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Spectre</td>\n",
       "      <td>245000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>[{\"name\": \"Columbia Pictures\", \"id\": 5}, {\"nam...</td>\n",
       "      <td>[{\"iso_3166_1\": \"GB\", \"name\": \"United Kingdom\"...</td>\n",
       "      <td>2015-10-26</td>\n",
       "      <td>880674609</td>\n",
       "      <td>148.0</td>\n",
       "      <td>[{\"iso_639_1\": \"fr\", \"name\": \"Fran\\u00e7ais\"},...</td>\n",
       "      <td>Released</td>\n",
       "      <td>A Plan No One Escapes</td>\n",
       "      <td>6.3</td>\n",
       "      <td>4466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>The Dark Knight Rises</td>\n",
       "      <td>250000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>[{\"name\": \"Legendary Pictures\", \"id\": 923}, {\"...</td>\n",
       "      <td>[{\"iso_3166_1\": \"US\", \"name\": \"United States o...</td>\n",
       "      <td>2012-07-16</td>\n",
       "      <td>1084939099</td>\n",
       "      <td>165.0</td>\n",
       "      <td>[{\"iso_639_1\": \"en\", \"name\": \"English\"}]</td>\n",
       "      <td>Released</td>\n",
       "      <td>The Legend Ends</td>\n",
       "      <td>7.6</td>\n",
       "      <td>9106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>John Carter</td>\n",
       "      <td>260000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>[{\"name\": \"Walt Disney Pictures\", \"id\": 2}]</td>\n",
       "      <td>[{\"iso_3166_1\": \"US\", \"name\": \"United States o...</td>\n",
       "      <td>2012-03-07</td>\n",
       "      <td>284139100</td>\n",
       "      <td>132.0</td>\n",
       "      <td>[{\"iso_639_1\": \"en\", \"name\": \"English\"}]</td>\n",
       "      <td>Released</td>\n",
       "      <td>Lost in our world, found in another.</td>\n",
       "      <td>6.1</td>\n",
       "      <td>2124</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 40 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      title     budget  gen_Action  \\\n",
       "0                                    Avatar  237000000         1.0   \n",
       "1  Pirates of the Caribbean: At World's End  300000000         1.0   \n",
       "2                                   Spectre  245000000         1.0   \n",
       "3                     The Dark Knight Rises  250000000         1.0   \n",
       "4                               John Carter  260000000         1.0   \n",
       "\n",
       "   gen_Adventure  gen_Animation  gen_Comedy  gen_Crime  gen_Documentary  \\\n",
       "0            1.0            0.0         0.0        0.0              0.0   \n",
       "1            1.0            0.0         0.0        0.0              0.0   \n",
       "2            1.0            0.0         0.0        1.0              0.0   \n",
       "3            0.0            0.0         0.0        1.0              0.0   \n",
       "4            1.0            0.0         0.0        0.0              0.0   \n",
       "\n",
       "   gen_Drama  gen_Family  ...  \\\n",
       "0        0.0         0.0  ...   \n",
       "1        0.0         0.0  ...   \n",
       "2        0.0         0.0  ...   \n",
       "3        1.0         0.0  ...   \n",
       "4        0.0         0.0  ...   \n",
       "\n",
       "                                production_companies  \\\n",
       "0  [{\"name\": \"Ingenious Film Partners\", \"id\": 289...   \n",
       "1  [{\"name\": \"Walt Disney Pictures\", \"id\": 2}, {\"...   \n",
       "2  [{\"name\": \"Columbia Pictures\", \"id\": 5}, {\"nam...   \n",
       "3  [{\"name\": \"Legendary Pictures\", \"id\": 923}, {\"...   \n",
       "4        [{\"name\": \"Walt Disney Pictures\", \"id\": 2}]   \n",
       "\n",
       "                                production_countries  release_date  \\\n",
       "0  [{\"iso_3166_1\": \"US\", \"name\": \"United States o...    2009-12-10   \n",
       "1  [{\"iso_3166_1\": \"US\", \"name\": \"United States o...    2007-05-19   \n",
       "2  [{\"iso_3166_1\": \"GB\", \"name\": \"United Kingdom\"...    2015-10-26   \n",
       "3  [{\"iso_3166_1\": \"US\", \"name\": \"United States o...    2012-07-16   \n",
       "4  [{\"iso_3166_1\": \"US\", \"name\": \"United States o...    2012-03-07   \n",
       "\n",
       "      revenue  runtime                                   spoken_languages  \\\n",
       "0  2787965087    162.0  [{\"iso_639_1\": \"en\", \"name\": \"English\"}, {\"iso...   \n",
       "1   961000000    169.0           [{\"iso_639_1\": \"en\", \"name\": \"English\"}]   \n",
       "2   880674609    148.0  [{\"iso_639_1\": \"fr\", \"name\": \"Fran\\u00e7ais\"},...   \n",
       "3  1084939099    165.0           [{\"iso_639_1\": \"en\", \"name\": \"English\"}]   \n",
       "4   284139100    132.0           [{\"iso_639_1\": \"en\", \"name\": \"English\"}]   \n",
       "\n",
       "     status                                         tagline  vote_average  \\\n",
       "0  Released                     Enter the World of Pandora.           7.2   \n",
       "1  Released  At the end of the world, the adventure begins.           6.9   \n",
       "2  Released                           A Plan No One Escapes           6.3   \n",
       "3  Released                                 The Legend Ends           7.6   \n",
       "4  Released            Lost in our world, found in another.           6.1   \n",
       "\n",
       "   vote_count  \n",
       "0       11800  \n",
       "1        4500  \n",
       "2        4466  \n",
       "3        9106  \n",
       "4        2124  \n",
       "\n",
       "[5 rows x 40 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "\n",
    "data = pd.read_csv(\"tmdb_5000_movies.csv\")\n",
    "data = data.dropna(subset=[\"overview\"])\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Podíváme se, kolik nám zbylo filmů."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4800, 40)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Na sloupec `overview` opět použijeme algoritmus TF-IDF, který jsme již používali v minulé lekci. Tento algoritmus převede jednotlivé popisy (dokumenty) na číselnou matici. Každé slovo (nebo dvojice slov) budou mít svůj sloupeček a číslo v něm bude reflektovat přítomnost slova v dokumentu a přítomnost slova ve všech dokumentech. Nově přidáme dva parametry.\n",
    "\n",
    "- `min_df` je minimální počet výskytů slova v dokumentech, aby nebylo vyřazeno. Potřebujeme slova, která se nachází alespoň ve dvou dokumentech, unikátní slovo nám nijak nepomůže propojit dokument s jiným dokumentem.\n",
    "- `max_df` je maximální počet výskytů slova (zadáváme jako desetinné číslo, tj. 70 %). Slova, která jsou ve většině dokumentů, nám taky nepomůžou vybrat vhodný obsah.\n",
    "\n",
    "Pro oba parametry platí následující pravidla:\n",
    "\n",
    "- Pokud zadáme hodnotu v intervalu 0 až 1, je hodnota braná jako procentuální výskyt.\n",
    "- Pokud zadáme hodnotu vyšší než 1, je hodnota braná jako výskyt v absolutní hodnotě.\n",
    "\n",
    "Parametry `stop_words` a `ngram_range` jsme si již ukázali minule a jich význam je stále stejný. Dále použijeme metodu `fit_transform()` a na výsledná data se podíváme."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = data[\"overview\"]\n",
    "\n",
    "vec = TfidfVectorizer(min_df=2, max_df=0.7, stop_words=\"english\", ngram_range=(1, 2))\n",
    "data_tfidf = vec.fit_transform(X)\n",
    "# Převod na pole\n",
    "data_tfidf.toarray()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Jednotlivá slova, kterým čísla patří, získáme pomocí metody `get_feature_names_out()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['00', '00 agent', '000', ..., 'zookeeper', 'zoologists', 'zorro'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vec.get_feature_names_out()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Abychom se na data mohli podívat v lepším formátu, můžeme si z nich opět sestavit pandas tabulku. Při vytváření tabulky řekneme, že popisky sloupců je možné získat pomocí metody `get_feature_names_out()`. Protože počet ani pořadí řádků se nijak nezměnily, načteme index tabulky ze sloupce `title` z původní tabulky s daty."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>00</th>\n",
       "      <th>00 agent</th>\n",
       "      <th>000</th>\n",
       "      <th>000 000</th>\n",
       "      <th>000 feet</th>\n",
       "      <th>000 foot</th>\n",
       "      <th>000 years</th>\n",
       "      <th>007</th>\n",
       "      <th>10</th>\n",
       "      <th>10 million</th>\n",
       "      <th>...</th>\n",
       "      <th>zombie outbreak</th>\n",
       "      <th>zombies</th>\n",
       "      <th>zone</th>\n",
       "      <th>zoo</th>\n",
       "      <th>zoo need</th>\n",
       "      <th>zooey</th>\n",
       "      <th>zooey deschanel</th>\n",
       "      <th>zookeeper</th>\n",
       "      <th>zoologists</th>\n",
       "      <th>zorro</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Avatar</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pirates of the Caribbean: At World's End</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spectre</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Dark Knight Rises</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>John Carter</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 17468 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           00  00 agent  000  000 000  \\\n",
       "title                                                                   \n",
       "Avatar                                    0.0       0.0  0.0      0.0   \n",
       "Pirates of the Caribbean: At World's End  0.0       0.0  0.0      0.0   \n",
       "Spectre                                   0.0       0.0  0.0      0.0   \n",
       "The Dark Knight Rises                     0.0       0.0  0.0      0.0   \n",
       "John Carter                               0.0       0.0  0.0      0.0   \n",
       "\n",
       "                                          000 feet  000 foot  000 years  007  \\\n",
       "title                                                                          \n",
       "Avatar                                         0.0       0.0        0.0  0.0   \n",
       "Pirates of the Caribbean: At World's End       0.0       0.0        0.0  0.0   \n",
       "Spectre                                        0.0       0.0        0.0  0.0   \n",
       "The Dark Knight Rises                          0.0       0.0        0.0  0.0   \n",
       "John Carter                                    0.0       0.0        0.0  0.0   \n",
       "\n",
       "                                           10  10 million  ...  \\\n",
       "title                                                      ...   \n",
       "Avatar                                    0.0         0.0  ...   \n",
       "Pirates of the Caribbean: At World's End  0.0         0.0  ...   \n",
       "Spectre                                   0.0         0.0  ...   \n",
       "The Dark Knight Rises                     0.0         0.0  ...   \n",
       "John Carter                               0.0         0.0  ...   \n",
       "\n",
       "                                          zombie outbreak  zombies  zone  zoo  \\\n",
       "title                                                                           \n",
       "Avatar                                                0.0      0.0   0.0  0.0   \n",
       "Pirates of the Caribbean: At World's End              0.0      0.0   0.0  0.0   \n",
       "Spectre                                               0.0      0.0   0.0  0.0   \n",
       "The Dark Knight Rises                                 0.0      0.0   0.0  0.0   \n",
       "John Carter                                           0.0      0.0   0.0  0.0   \n",
       "\n",
       "                                          zoo need  zooey  zooey deschanel  \\\n",
       "title                                                                        \n",
       "Avatar                                         0.0    0.0              0.0   \n",
       "Pirates of the Caribbean: At World's End       0.0    0.0              0.0   \n",
       "Spectre                                        0.0    0.0              0.0   \n",
       "The Dark Knight Rises                          0.0    0.0              0.0   \n",
       "John Carter                                    0.0    0.0              0.0   \n",
       "\n",
       "                                          zookeeper  zoologists  zorro  \n",
       "title                                                                   \n",
       "Avatar                                          0.0         0.0    0.0  \n",
       "Pirates of the Caribbean: At World's End        0.0         0.0    0.0  \n",
       "Spectre                                         0.0         0.0    0.0  \n",
       "The Dark Knight Rises                           0.0         0.0    0.0  \n",
       "John Carter                                     0.0         0.0    0.0  \n",
       "\n",
       "[5 rows x 17468 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_tfidf = pd.DataFrame(data_tfidf.toarray(), columns=vec.get_feature_names_out(), index=data[\"title\"])\n",
    "data_tfidf.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nyní si vybereme jeden film, například film [The Fugitive (Uprchlík)](https://www.imdb.com/title/tt0106977/) z roku 1993 s Harrísonem Fordem v hlavní roli. Seřadíme si hodnoty od nejvyšší po nejmenší. Vidíme, že především slovo \"Richard\" může být celkem matoucí, protože jde o jméno hlavní postavy filmu, ale k ději nemá žádnou vazbu. Naopak slova jako \"killer\", \"wrongfully\" nebo \"pursuing team\" můžou být poměrně důležitá a můžou nám pomoci najít dějově podobný film. (Film Uprchlík je o chirurgovi Richardovi Kimblovi, který byl nespravedlivě odsouzen za vraždu své ženy. Při převozu do vězení se mu podaří uprchnout. Poté hledá skutečného vraha a je při tom pronásledován skupinou policitů.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "richard              0.399296\n",
       "killer               0.216403\n",
       "discovers secrets    0.186469\n",
       "wrongfully           0.186469\n",
       "kimble               0.186469\n",
       "death struggles      0.186469\n",
       "pursuing team        0.186469\n",
       "struggles expose     0.186469\n",
       "team                 0.185769\n",
       "gerard               0.180066\n",
       "marshals             0.180066\n",
       "wife                 0.177323\n",
       "accused murdering    0.171041\n",
       "samuel               0.167610\n",
       "intricate            0.167610\n",
       "chases               0.164638\n",
       "deputy               0.162017\n",
       "pursuing             0.159672\n",
       "murdering            0.153832\n",
       "expose               0.149211\n",
       "Name: The Fugitive, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_tfidf.loc[\"The Fugitive\"].sort_values(ascending=False).head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Abychom dokázali změřit podobnost mezi jednotlivými řádky tabulky `data_tfidf`, použijeme kosinovou podobnost.\n",
    "\n",
    "#### Kosinová vzdálenost a podobnost\n",
    "\n",
    "Kosinová vzdálenost (*cosine distance*) a kosinová podobnost (*cosine similarity*) jsou způsoby měření vzdálenosti, se kterým jsme se zatím nesetkali. Ukažme si, jak kosinová vzdálenost a podobnost fungují v dvourozměrném prostoru, tj. jak by změřila vzdálenost dvou hodnot z tabulky se dvěma sloupci. Body vidíme na obrázku níže. Kosinová podobnost je vypočítaná jako kosinus úhlu $\\theta$, tj. úhlu, který svírají vektory, které směrují do obou bodů. Pokud bychom měli dva naprosto totožné body, úhel mezi jejich vektory by měl velikost 0 stupňů. Kosinová vzdálenost těchto bodů by byla 0, kosinová podobnost by byla 1.\n",
    "\n",
    "![cosine.png](cosine.png)\n",
    "\n",
    "Kosinovou vzdálenost a podobnost je možné spočítat i ve vícerozměrném prostoru a i v něm platí, že čím jsou si vektory podobnější, tím větší je hodnota kosinové podobnosti a menší hodnota kosinové vzdálenosti. My budeme využívat kosinovou podobnost. Kosinová podobnost je v modulu `scikit-learn` implementovaná ve funkci `cosine_similarity()`.\n",
    "\n",
    "Funkci můžeme dát dvě hodnoty (v našem případě dva filmy) a ona spočítá kosinovou podobnost, tj. podobnost jejich popisů. Pokud funkci dáváme dva filmy, můžeme je vybrat s využitím indexu (a vlastnosti `loc`). Funkce ale očekává dvourozměrné pole.\n",
    "\n",
    "Jednorozměrné pole můžeme ze série získat s využitím vlastnosti `.values`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., ..., 0., 0., 0.])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_tfidf.loc[\"The Fugitive\"].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Abychom získali dvourozměrnou matici, můžeme použít metodu `reshape()`. Té dáme parametry `1` a `-1`, čímž získáme dvourozměrné pole.\n",
    "\n",
    "Hodnota `1` znamená, že chceme mít pole, které má jeden řádek. A hodnota `-1` udělá pole dvourozměrné."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_tfidf.loc[\"The Fugitive\"].values.reshape(1, -1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Spočítejme nyní podobnost například mezi filmem The Fugitive a Witness"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.01564247]])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cosine_similarity(data_tfidf.loc[\"The Fugitive\"].values.reshape(1, -1), data_tfidf.loc[\"Witness\"].values.reshape(1, -1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Podobnost je sice poměrně nízká (hodnota je blízká 0), absolutní hodnota nemusí být příliš důležitá. Zkusme ještě spočítat podobnost mezi filmy The Fugitive a Titanic."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.00693092]])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cosine_similarity(data_tfidf.loc[\"The Fugitive\"].values.reshape(1, -1), data_tfidf.loc[\"Titanic\"].values.reshape(1, -1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vidíme, že pro tyto fimy je podobnost větší. Náš systém by tedy nabídl uživateli, kterému se líbil film The Fugitive, film Witness a nikoli film Titanic.\n",
    "\n",
    "V celé datové sadě se ale může nacházet ještě vhodnější film. Abychom získali obecná data pro doporučení libovolného filmu, použijeme funkci  `cosine_similarity` na celou tabulku `cosine_similarity()`. Tento příkaz spočítá kosinovou podobnost pro každou dvojici filmů. Získáme tedy čtvercovou matici, kde se počet řádků a počet sloupců rovná počtu filmů v původní tabulce."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
       "        0.        ],\n",
       "       [0.        , 1.        , 0.        , ..., 0.02190503, 0.        ,\n",
       "        0.        ],\n",
       "       [0.        , 0.        , 1.        , ..., 0.01450648, 0.        ,\n",
       "        0.        ],\n",
       "       ...,\n",
       "       [0.        , 0.02190503, 0.01450648, ..., 1.        , 0.01549051,\n",
       "        0.00614697],\n",
       "       [0.        , 0.        , 0.        , ..., 0.01549051, 1.        ,\n",
       "        0.01001404],\n",
       "       [0.        , 0.        , 0.        , ..., 0.00614697, 0.01001404,\n",
       "        1.        ]])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cosine_similarity_array = cosine_similarity(data_tfidf)\n",
    "cosine_similarity_array"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pro snazší práci s daty opět provedeme převod na tabulku. Názvy sloupců i řádků (index) jsou dané jmény filmů."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>title</th>\n",
       "      <th>Avatar</th>\n",
       "      <th>Pirates of the Caribbean: At World's End</th>\n",
       "      <th>Spectre</th>\n",
       "      <th>The Dark Knight Rises</th>\n",
       "      <th>John Carter</th>\n",
       "      <th>Spider-Man 3</th>\n",
       "      <th>Tangled</th>\n",
       "      <th>Avengers: Age of Ultron</th>\n",
       "      <th>Harry Potter and the Half-Blood Prince</th>\n",
       "      <th>Batman v Superman: Dawn of Justice</th>\n",
       "      <th>...</th>\n",
       "      <th>On The Downlow</th>\n",
       "      <th>Sanctuary: Quite a Conundrum</th>\n",
       "      <th>Bang</th>\n",
       "      <th>Primer</th>\n",
       "      <th>Cavite</th>\n",
       "      <th>El Mariachi</th>\n",
       "      <th>Newlyweds</th>\n",
       "      <th>Signed, Sealed, Delivered</th>\n",
       "      <th>Shanghai Calling</th>\n",
       "      <th>My Date with Drew</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Avatar</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.020459</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.024428</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.033653</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.027932</td>\n",
       "      <td>0.040753</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pirates of the Caribbean: At World's End</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.034509</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.021569</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.007011</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.021905</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spectre</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.028307</td>\n",
       "      <td>0.019265</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.025571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.018272</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.014506</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Dark Knight Rises</th>\n",
       "      <td>0.020459</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.009002</td>\n",
       "      <td>0.003669</td>\n",
       "      <td>0.011148</td>\n",
       "      <td>0.021395</td>\n",
       "      <td>0.013906</td>\n",
       "      <td>0.136604</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.028649</td>\n",
       "      <td>0.035299</td>\n",
       "      <td>0.028564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>John Carter</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.034509</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.009002</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.009827</td>\n",
       "      <td>0.035313</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.017015</td>\n",
       "      <td>...</td>\n",
       "      <td>0.012130</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.006163</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 4800 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "title                                       Avatar  \\\n",
       "title                                                \n",
       "Avatar                                    1.000000   \n",
       "Pirates of the Caribbean: At World's End  0.000000   \n",
       "Spectre                                   0.000000   \n",
       "The Dark Knight Rises                     0.020459   \n",
       "John Carter                               0.000000   \n",
       "\n",
       "title                                     Pirates of the Caribbean: At World's End  \\\n",
       "title                                                                                \n",
       "Avatar                                                                    0.000000   \n",
       "Pirates of the Caribbean: At World's End                                  1.000000   \n",
       "Spectre                                                                   0.000000   \n",
       "The Dark Knight Rises                                                     0.000000   \n",
       "John Carter                                                               0.034509   \n",
       "\n",
       "title                                     Spectre  The Dark Knight Rises  \\\n",
       "title                                                                      \n",
       "Avatar                                        0.0               0.020459   \n",
       "Pirates of the Caribbean: At World's End      0.0               0.000000   \n",
       "Spectre                                       1.0               0.000000   \n",
       "The Dark Knight Rises                         0.0               1.000000   \n",
       "John Carter                                   0.0               0.009002   \n",
       "\n",
       "title                                     John Carter  Spider-Man 3   Tangled  \\\n",
       "title                                                                           \n",
       "Avatar                                       0.000000      0.024428  0.000000   \n",
       "Pirates of the Caribbean: At World's End     0.034509      0.000000  0.000000   \n",
       "Spectre                                      0.000000      0.000000  0.000000   \n",
       "The Dark Knight Rises                        0.009002      0.003669  0.011148   \n",
       "John Carter                                  1.000000      0.000000  0.009827   \n",
       "\n",
       "title                                     Avengers: Age of Ultron  \\\n",
       "title                                                               \n",
       "Avatar                                                   0.033653   \n",
       "Pirates of the Caribbean: At World's End                 0.021569   \n",
       "Spectre                                                  0.028307   \n",
       "The Dark Knight Rises                                    0.021395   \n",
       "John Carter                                              0.035313   \n",
       "\n",
       "title                                     Harry Potter and the Half-Blood Prince  \\\n",
       "title                                                                              \n",
       "Avatar                                                                  0.000000   \n",
       "Pirates of the Caribbean: At World's End                                0.000000   \n",
       "Spectre                                                                 0.019265   \n",
       "The Dark Knight Rises                                                   0.013906   \n",
       "John Carter                                                             0.000000   \n",
       "\n",
       "title                                     Batman v Superman: Dawn of Justice  \\\n",
       "title                                                                          \n",
       "Avatar                                                              0.000000   \n",
       "Pirates of the Caribbean: At World's End                            0.000000   \n",
       "Spectre                                                             0.000000   \n",
       "The Dark Knight Rises                                               0.136604   \n",
       "John Carter                                                         0.017015   \n",
       "\n",
       "title                                     ...  On The Downlow  \\\n",
       "title                                     ...                   \n",
       "Avatar                                    ...        0.000000   \n",
       "Pirates of the Caribbean: At World's End  ...        0.000000   \n",
       "Spectre                                   ...        0.025571   \n",
       "The Dark Knight Rises                     ...        0.000000   \n",
       "John Carter                               ...        0.012130   \n",
       "\n",
       "title                                     Sanctuary: Quite a Conundrum  \\\n",
       "title                                                                    \n",
       "Avatar                                                             0.0   \n",
       "Pirates of the Caribbean: At World's End                           0.0   \n",
       "Spectre                                                            0.0   \n",
       "The Dark Knight Rises                                              0.0   \n",
       "John Carter                                                        0.0   \n",
       "\n",
       "title                                         Bang    Primer    Cavite  \\\n",
       "title                                                                    \n",
       "Avatar                                    0.027932  0.040753  0.000000   \n",
       "Pirates of the Caribbean: At World's End  0.007011  0.000000  0.000000   \n",
       "Spectre                                   0.000000  0.000000  0.018272   \n",
       "The Dark Knight Rises                     0.000000  0.000000  0.000000   \n",
       "John Carter                               0.000000  0.000000  0.000000   \n",
       "\n",
       "title                                     El Mariachi  Newlyweds  \\\n",
       "title                                                              \n",
       "Avatar                                            0.0        0.0   \n",
       "Pirates of the Caribbean: At World's End          0.0        0.0   \n",
       "Spectre                                           0.0        0.0   \n",
       "The Dark Knight Rises                             0.0        0.0   \n",
       "John Carter                                       0.0        0.0   \n",
       "\n",
       "title                                     Signed, Sealed, Delivered  \\\n",
       "title                                                                 \n",
       "Avatar                                                     0.000000   \n",
       "Pirates of the Caribbean: At World's End                   0.021905   \n",
       "Spectre                                                    0.014506   \n",
       "The Dark Knight Rises                                      0.028649   \n",
       "John Carter                                                0.006163   \n",
       "\n",
       "title                                     Shanghai Calling  My Date with Drew  \n",
       "title                                                                          \n",
       "Avatar                                            0.000000           0.000000  \n",
       "Pirates of the Caribbean: At World's End          0.000000           0.000000  \n",
       "Spectre                                           0.000000           0.000000  \n",
       "The Dark Knight Rises                             0.035299           0.028564  \n",
       "John Carter                                       0.000000           0.000000  \n",
       "\n",
       "[5 rows x 4800 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cosine_similarity_df = pd.DataFrame(cosine_similarity_array, index=data_tfidf.index, columns=data_tfidf.index)\n",
    "cosine_similarity_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Matice je symetrická. Pokud nás zajímají podobnosti jednotlivých filmů s filmem The Fugitive, můžeme vyjmut řádek (pomocí `.loc`) nebo sloupec (s využitím hranatých závorek). Jednodušší bude práce se sloupcem, protože budeme chtít data seřadit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title\n",
       "Avatar                                      0.000000\n",
       "Pirates of the Caribbean: At World's End    0.000000\n",
       "Spectre                                     0.000000\n",
       "The Dark Knight Rises                       0.018202\n",
       "John Carter                                 0.000000\n",
       "                                              ...   \n",
       "El Mariachi                                 0.048717\n",
       "Newlyweds                                   0.000000\n",
       "Signed, Sealed, Delivered                   0.032423\n",
       "Shanghai Calling                            0.000000\n",
       "My Date with Drew                           0.000000\n",
       "Name: The Fugitive, Length: 4800, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances_the_fugitive = cosine_similarity_df[\"The Fugitive\"]\n",
    "distances_the_fugitive"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Zobrazíme si tedy 10 nejvíce podobných filmů. Nejvíce \"podobný\" je film \"The Rocket: The Legend of Rocket Richard\", který ale dějově nemá s filmem The Fugitive nic společného. Podobnost je daná čistě tím, že hlavní postavy obou filmů se jmenují Richard. To bychom mohli vyřešit například tím, že bychom zohlednili zánr filmu. Naopak následující dva filmy The Iceman a The Life of David Gale lze považovat za dobré tipy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title\n",
       "The Fugitive                                1.000000\n",
       "The Rocket: The Legend of Rocket Richard    0.207967\n",
       "The Iceman                                  0.171206\n",
       "The Life of David Gale                      0.165951\n",
       "Married Life                                0.157624\n",
       "The Work and The Story                      0.153640\n",
       "A Time to Kill                              0.112665\n",
       "Austin Powers in Goldmember                 0.109479\n",
       "Somewhere in Time                           0.108948\n",
       "High Anxiety                                0.100456\n",
       "Name: The Fugitive, dtype: float64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distances_the_fugitive.sort_values(ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Čtení na doma: Výběr na základě uživatelského profilu\n",
    "\n",
    "Pokud uživatel používá nějakou službu delší dobu, budeme určitě vědět o více produktech (např. filmech), které se mu líbily. V takovém případě nebudeme hledat podobnost mezi dvěma produkty, ale mezi uživatelským profilem a produkty, které budou pro uživatele nové (např. filmy, které ještě neviděl). Uvažujme například uživatele (uživatelku), které se líbily firmy Notting Hill, Titanic, The Great Gatsby a The Lovers. To, že se filmy líbily, můžeme poznat podle explicitního (uživatel použil tlačítko \"To se mi líbí\") nebo implicitního (uživatel zhlédl celý film) feedbacku.\n",
    "\n",
    "Vytvoříme tedy proměnnou `data_user_profile`, která bude reprezentovat preference uživatele (uživatelky). Ty získáme tak, že spočítáme průměrné hodnoty jednotlivých slov (a dvojic slov) pro výše uvedené filmy. Abychom měli data ve dvourozměrné matici, oipět je třeba použít metodu `reshape()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "favorite_movies = [\"Notting Hill\", \"Titanic\", \"The Great Gatsby\", \"The Lovers\"]\n",
    "data_user_profile = data_tfidf.loc[favorite_movies]\n",
    "data_user_profile = data_user_profile.mean()\n",
    "data_user_profile = data_user_profile.values.reshape(1, -1)\n",
    "data_user_profile"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nyní zkusme najít filmy, které budou nejvíce podobného uživatelskému profilu. Nejprve z původní tabulky vyřadíme filmy ze seznamu `favorite_movies`. Poté spočítáme kosinové vzdálenosti mezi uživatelským profilem a jednotlivými filmy. Vznikne matice s jedním sloupcem, který pro každý film udává kosinovou vzdálenost mezi popisem filmu a uživatelským profilem."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.02648553],\n",
       "       [0.02318485],\n",
       "       [0.        ],\n",
       "       ...,\n",
       "       [0.01260339],\n",
       "       [0.00843268],\n",
       "       [0.02732532]])"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_tfidf_not_seen = data_tfidf.drop(favorite_movies, axis=0)\n",
    "cosine_similarity_array_not_seen = cosine_similarity(data_tfidf_not_seen, data_user_profile)\n",
    "cosine_similarity_array_not_seen"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data opět převedeme na tabulku, abychom viděli názvy filmů, které náš systém danému uživateli doporučí. Jako indexy použijeme názvy filmů z tabulky `data_tfidf_not_seen`. Nakonec data seřadíme. Vidíme, že náš systém by doporučil filmy Romance & Cigarettes, Four Weddings and a Funeral a Rent."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Romance &amp; Cigarettes</th>\n",
       "      <td>0.130770</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Four Weddings and a Funeral</th>\n",
       "      <td>0.128872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Rent</th>\n",
       "      <td>0.121126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Little Prince</th>\n",
       "      <td>0.112538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dil Jo Bhi Kahey...</th>\n",
       "      <td>0.100109</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                score\n",
       "title                                \n",
       "Romance & Cigarettes         0.130770\n",
       "Four Weddings and a Funeral  0.128872\n",
       "Rent                         0.121126\n",
       "The Little Prince            0.112538\n",
       "Dil Jo Bhi Kahey...          0.100109"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_recommendation = pd.DataFrame(cosine_similarity_array_not_seen, index=data_tfidf_not_seen.index, columns=[\"score\"])\n",
    "user_recommendation.sort_values(by=\"score\", ascending=False).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Čtení na doma: Kolaborativní filtrování\n",
    "\n",
    "U kolaborativního filtrování se řídíme preferencemi jednotlivých uživatelů a uživatelek. Využijeme data ze seouboru [user_ratings.csv](user_ratings.csv), který obsahuje uživatelská hodnocení fimů. U hodnocení víme ID uživatele (`userId`), bodové hodnocení filmu (sloupec `rating`) a název filmu (sloupec `title`)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userId</th>\n",
       "      <th>rating</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Toy Story (1995)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Toy Story (1995)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>4.5</td>\n",
       "      <td>Toy Story (1995)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>15</td>\n",
       "      <td>2.5</td>\n",
       "      <td>Toy Story (1995)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>17</td>\n",
       "      <td>4.5</td>\n",
       "      <td>Toy Story (1995)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   userId  rating             title\n",
       "0       1     4.0  Toy Story (1995)\n",
       "1       5     4.0  Toy Story (1995)\n",
       "2       7     4.5  Toy Story (1995)\n",
       "3      15     2.5  Toy Story (1995)\n",
       "4      17     4.5  Toy Story (1995)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings = pd.read_csv(\"user_ratings.csv\")\n",
    "data_ratings = data_ratings[[\"userId\", \"rating\", \"title\"]].drop_duplicates(subset=[\"userId\", \"title\"], keep=\"last\")\n",
    "data_ratings.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pro další práci potřebujeme mít trochu jinou strukturu tabulky. Měli bychom mít názvy filmů jako sloupce, takže jeden řádek tabulky bude představovat všechna hodnocení jednoho uživatele (uživatelky). K tomu můžeme využít funkce `pivot()`. Pozor, nejedná se o funkci `pivot_table()`! Rozdíl mezi `pivot()` a `pivot_table()` je v tom, že `pivot()` neprovádí žádnou agregaci, ale pouze \"přeskládá\" hodnoty z jednoho sloupce do více sloupců. Z toho důvodu zadáváme funkci `pivot` pouze tři parametry: `index` (názvy řádků - index), `columns` (názvy sloupců) a `values` (hodnoty). Nezadáváme agregační funkci."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>title</th>\n",
       "      <th>'71 (2014)</th>\n",
       "      <th>'Hellboy': The Seeds of Creation (2004)</th>\n",
       "      <th>'Round Midnight (1986)</th>\n",
       "      <th>'Salem's Lot (2004)</th>\n",
       "      <th>'Til There Was You (1997)</th>\n",
       "      <th>'Tis the Season for Love (2015)</th>\n",
       "      <th>'burbs, The (1989)</th>\n",
       "      <th>'night Mother (1986)</th>\n",
       "      <th>(500) Days of Summer (2009)</th>\n",
       "      <th>*batteries not included (1987)</th>\n",
       "      <th>...</th>\n",
       "      <th>Zulu (2013)</th>\n",
       "      <th>[REC] (2007)</th>\n",
       "      <th>[REC]² (2009)</th>\n",
       "      <th>[REC]³ 3 Génesis (2012)</th>\n",
       "      <th>anohana: The Flower We Saw That Day - The Movie (2013)</th>\n",
       "      <th>eXistenZ (1999)</th>\n",
       "      <th>xXx (2002)</th>\n",
       "      <th>xXx: State of the Union (2005)</th>\n",
       "      <th>¡Three Amigos! (1986)</th>\n",
       "      <th>À nous la liberté (Freedom for Us) (1931)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>userId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>606</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>607</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>608</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>609</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>610</th>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.5</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>610 rows × 9719 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "title   '71 (2014)  'Hellboy': The Seeds of Creation (2004)  \\\n",
       "userId                                                        \n",
       "1              NaN                                      NaN   \n",
       "2              NaN                                      NaN   \n",
       "3              NaN                                      NaN   \n",
       "4              NaN                                      NaN   \n",
       "5              NaN                                      NaN   \n",
       "...            ...                                      ...   \n",
       "606            NaN                                      NaN   \n",
       "607            NaN                                      NaN   \n",
       "608            NaN                                      NaN   \n",
       "609            NaN                                      NaN   \n",
       "610            4.0                                      NaN   \n",
       "\n",
       "title   'Round Midnight (1986)  'Salem's Lot (2004)  \\\n",
       "userId                                                \n",
       "1                          NaN                  NaN   \n",
       "2                          NaN                  NaN   \n",
       "3                          NaN                  NaN   \n",
       "4                          NaN                  NaN   \n",
       "5                          NaN                  NaN   \n",
       "...                        ...                  ...   \n",
       "606                        NaN                  NaN   \n",
       "607                        NaN                  NaN   \n",
       "608                        NaN                  NaN   \n",
       "609                        NaN                  NaN   \n",
       "610                        NaN                  NaN   \n",
       "\n",
       "title   'Til There Was You (1997)  'Tis the Season for Love (2015)  \\\n",
       "userId                                                               \n",
       "1                             NaN                              NaN   \n",
       "2                             NaN                              NaN   \n",
       "3                             NaN                              NaN   \n",
       "4                             NaN                              NaN   \n",
       "5                             NaN                              NaN   \n",
       "...                           ...                              ...   \n",
       "606                           NaN                              NaN   \n",
       "607                           NaN                              NaN   \n",
       "608                           NaN                              NaN   \n",
       "609                           NaN                              NaN   \n",
       "610                           NaN                              NaN   \n",
       "\n",
       "title   'burbs, The (1989)  'night Mother (1986)  (500) Days of Summer (2009)  \\\n",
       "userId                                                                          \n",
       "1                      NaN                   NaN                          NaN   \n",
       "2                      NaN                   NaN                          NaN   \n",
       "3                      NaN                   NaN                          NaN   \n",
       "4                      NaN                   NaN                          NaN   \n",
       "5                      NaN                   NaN                          NaN   \n",
       "...                    ...                   ...                          ...   \n",
       "606                    NaN                   NaN                          NaN   \n",
       "607                    NaN                   NaN                          NaN   \n",
       "608                    NaN                   NaN                          NaN   \n",
       "609                    NaN                   NaN                          NaN   \n",
       "610                    NaN                   NaN                          3.5   \n",
       "\n",
       "title   *batteries not included (1987)  ...  Zulu (2013)  [REC] (2007)  \\\n",
       "userId                                  ...                              \n",
       "1                                  NaN  ...          NaN           NaN   \n",
       "2                                  NaN  ...          NaN           NaN   \n",
       "3                                  NaN  ...          NaN           NaN   \n",
       "4                                  NaN  ...          NaN           NaN   \n",
       "5                                  NaN  ...          NaN           NaN   \n",
       "...                                ...  ...          ...           ...   \n",
       "606                                NaN  ...          NaN           NaN   \n",
       "607                                NaN  ...          NaN           NaN   \n",
       "608                                NaN  ...          NaN           NaN   \n",
       "609                                NaN  ...          NaN           NaN   \n",
       "610                                NaN  ...          NaN           4.0   \n",
       "\n",
       "title   [REC]² (2009)  [REC]³ 3 Génesis (2012)  \\\n",
       "userId                                           \n",
       "1                 NaN                      NaN   \n",
       "2                 NaN                      NaN   \n",
       "3                 NaN                      NaN   \n",
       "4                 NaN                      NaN   \n",
       "5                 NaN                      NaN   \n",
       "...               ...                      ...   \n",
       "606               NaN                      NaN   \n",
       "607               NaN                      NaN   \n",
       "608               NaN                      NaN   \n",
       "609               NaN                      NaN   \n",
       "610               3.5                      3.0   \n",
       "\n",
       "title   anohana: The Flower We Saw That Day - The Movie (2013)  \\\n",
       "userId                                                           \n",
       "1                                                     NaN        \n",
       "2                                                     NaN        \n",
       "3                                                     NaN        \n",
       "4                                                     NaN        \n",
       "5                                                     NaN        \n",
       "...                                                   ...        \n",
       "606                                                   NaN        \n",
       "607                                                   NaN        \n",
       "608                                                   NaN        \n",
       "609                                                   NaN        \n",
       "610                                                   NaN        \n",
       "\n",
       "title   eXistenZ (1999)  xXx (2002)  xXx: State of the Union (2005)  \\\n",
       "userId                                                                \n",
       "1                   NaN         NaN                             NaN   \n",
       "2                   NaN         NaN                             NaN   \n",
       "3                   NaN         NaN                             NaN   \n",
       "4                   NaN         NaN                             NaN   \n",
       "5                   NaN         NaN                             NaN   \n",
       "...                 ...         ...                             ...   \n",
       "606                 NaN         NaN                             NaN   \n",
       "607                 NaN         NaN                             NaN   \n",
       "608                 4.5         3.5                             NaN   \n",
       "609                 NaN         NaN                             NaN   \n",
       "610                 NaN         2.0                             1.5   \n",
       "\n",
       "title   ¡Three Amigos! (1986)  À nous la liberté (Freedom for Us) (1931)  \n",
       "userId                                                                    \n",
       "1                         4.0                                        NaN  \n",
       "2                         NaN                                        NaN  \n",
       "3                         NaN                                        NaN  \n",
       "4                         NaN                                        NaN  \n",
       "5                         NaN                                        NaN  \n",
       "...                       ...                                        ...  \n",
       "606                       NaN                                        NaN  \n",
       "607                       NaN                                        NaN  \n",
       "608                       NaN                                        NaN  \n",
       "609                       NaN                                        NaN  \n",
       "610                       NaN                                        NaN  \n",
       "\n",
       "[610 rows x 9719 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings_pivot = data_ratings.pivot(index=\"userId\", columns=\"title\", values=\"rating\")\n",
    "data_ratings_pivot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Podívejme se nyní na hodnocení uživatele (nebo uživatelky) s id 2. Jeho hodnocení zjistíme jako řádek s id 2. Pozor na to, že sloupce v původní tabulce jsou čísla, musíme tedy do `loc` vložít hodnotu 2 jako číslo, tj. bez uvozovek. Pokud bychom zadali číslo s uvozovkami, modul `pandas` by hledal řetězec 2 a nic by nenašel, protože řetězec 2 je jiná hodnota než číslo 2.\n",
    "\n",
    "Vidíme, že máme celkem 29 hodnocení a jsou zastoupena jak vysoká, tak nízká hodnocení."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title\n",
       "The Drop (2014)                                         2.0\n",
       "Girl with the Dragon Tattoo, The (2011)                 2.5\n",
       "Zombieland (2009)                                       3.0\n",
       "Shawshank Redemption, The (1994)                        3.0\n",
       "Interstellar (2014)                                     3.0\n",
       "Exit Through the Gift Shop (2010)                       3.0\n",
       "Collateral (2004)                                       3.5\n",
       "Django Unchained (2012)                                 3.5\n",
       "Dark Knight Rises, The (2012)                           3.5\n",
       "Ex Machina (2015)                                       3.5\n",
       "Whiplash (2014)                                         4.0\n",
       "Tommy Boy (1995)                                        4.0\n",
       "Talladega Nights: The Ballad of Ricky Bobby (2006)      4.0\n",
       "Shutter Island (2010)                                   4.0\n",
       "Louis C.K.: Hilarious (2010)                            4.0\n",
       "Kill Bill: Vol. 1 (2003)                                4.0\n",
       "Departed, The (2006)                                    4.0\n",
       "Inception (2010)                                        4.0\n",
       "Gladiator (2000)                                        4.0\n",
       "Inglourious Basterds (2009)                             4.5\n",
       "Good Will Hunting (1997)                                4.5\n",
       "Town, The (2010)                                        4.5\n",
       "Dark Knight, The (2008)                                 4.5\n",
       "Mad Max: Fury Road (2015)                               5.0\n",
       "Inside Job (2010)                                       5.0\n",
       "Step Brothers (2008)                                    5.0\n",
       "The Jinx: The Life and Deaths of Robert Durst (2015)    5.0\n",
       "Warrior (2011)                                          5.0\n",
       "Wolf of Wall Street, The (2013)                         5.0\n",
       "Name: 2, dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings_pivot.loc[2].dropna().sort_values()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Velkým problémem dat tohoto typu je, že v nich je obrovské množství prázdných hodnot. Pro taková data se používá označení řídká (*sparse*). My si ukážeme velmi jednoduchý přístup k řešení tohoto problému. Nemůžeme nyní ale jen nahradit chybějící hodnoty 0, protože tím bychom zkresili průměrná hodnocení filmu. Využijeme následující postup.\n",
    "\n",
    "1. Pro každý film spočteme průměrné hodnocení.\n",
    "1. Od každého hodnocení odečteme průměrné hodnocení.\n",
    "1. Nyní můžeme prázdné hodnoty nahradit 0, aniž bychom ovlivnili průměrné hodnocení filmů."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title\n",
       "'71 (2014)                                   4.000000\n",
       "'Hellboy': The Seeds of Creation (2004)      4.000000\n",
       "'Round Midnight (1986)                       3.500000\n",
       "'Salem's Lot (2004)                          5.000000\n",
       "'Til There Was You (1997)                    4.000000\n",
       "                                               ...   \n",
       "eXistenZ (1999)                              3.863636\n",
       "xXx (2002)                                   2.770833\n",
       "xXx: State of the Union (2005)               2.000000\n",
       "¡Three Amigos! (1986)                        3.134615\n",
       "À nous la liberté (Freedom for Us) (1931)    1.000000\n",
       "Length: 9719, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avg_ratings = data_ratings_pivot.mean()\n",
    "avg_ratings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>title</th>\n",
       "      <th>'71 (2014)</th>\n",
       "      <th>'Hellboy': The Seeds of Creation (2004)</th>\n",
       "      <th>'Round Midnight (1986)</th>\n",
       "      <th>'Salem's Lot (2004)</th>\n",
       "      <th>'Til There Was You (1997)</th>\n",
       "      <th>'Tis the Season for Love (2015)</th>\n",
       "      <th>'burbs, The (1989)</th>\n",
       "      <th>'night Mother (1986)</th>\n",
       "      <th>(500) Days of Summer (2009)</th>\n",
       "      <th>*batteries not included (1987)</th>\n",
       "      <th>...</th>\n",
       "      <th>Zulu (2013)</th>\n",
       "      <th>[REC] (2007)</th>\n",
       "      <th>[REC]² (2009)</th>\n",
       "      <th>[REC]³ 3 Génesis (2012)</th>\n",
       "      <th>anohana: The Flower We Saw That Day - The Movie (2013)</th>\n",
       "      <th>eXistenZ (1999)</th>\n",
       "      <th>xXx (2002)</th>\n",
       "      <th>xXx: State of the Union (2005)</th>\n",
       "      <th>¡Three Amigos! (1986)</th>\n",
       "      <th>À nous la liberté (Freedom for Us) (1931)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>userId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.865385</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>606</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>607</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>608</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.636364</td>\n",
       "      <td>0.729167</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>609</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>610</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.166667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.0625</td>\n",
       "      <td>-0.166667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.770833</td>\n",
       "      <td>-0.5</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>610 rows × 9719 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "title   '71 (2014)  'Hellboy': The Seeds of Creation (2004)  \\\n",
       "userId                                                        \n",
       "1              0.0                                      0.0   \n",
       "2              0.0                                      0.0   \n",
       "3              0.0                                      0.0   \n",
       "4              0.0                                      0.0   \n",
       "5              0.0                                      0.0   \n",
       "...            ...                                      ...   \n",
       "606            0.0                                      0.0   \n",
       "607            0.0                                      0.0   \n",
       "608            0.0                                      0.0   \n",
       "609            0.0                                      0.0   \n",
       "610            0.0                                      0.0   \n",
       "\n",
       "title   'Round Midnight (1986)  'Salem's Lot (2004)  \\\n",
       "userId                                                \n",
       "1                          0.0                  0.0   \n",
       "2                          0.0                  0.0   \n",
       "3                          0.0                  0.0   \n",
       "4                          0.0                  0.0   \n",
       "5                          0.0                  0.0   \n",
       "...                        ...                  ...   \n",
       "606                        0.0                  0.0   \n",
       "607                        0.0                  0.0   \n",
       "608                        0.0                  0.0   \n",
       "609                        0.0                  0.0   \n",
       "610                        0.0                  0.0   \n",
       "\n",
       "title   'Til There Was You (1997)  'Tis the Season for Love (2015)  \\\n",
       "userId                                                               \n",
       "1                             0.0                              0.0   \n",
       "2                             0.0                              0.0   \n",
       "3                             0.0                              0.0   \n",
       "4                             0.0                              0.0   \n",
       "5                             0.0                              0.0   \n",
       "...                           ...                              ...   \n",
       "606                           0.0                              0.0   \n",
       "607                           0.0                              0.0   \n",
       "608                           0.0                              0.0   \n",
       "609                           0.0                              0.0   \n",
       "610                           0.0                              0.0   \n",
       "\n",
       "title   'burbs, The (1989)  'night Mother (1986)  (500) Days of Summer (2009)  \\\n",
       "userId                                                                          \n",
       "1                      0.0                   0.0                     0.000000   \n",
       "2                      0.0                   0.0                     0.000000   \n",
       "3                      0.0                   0.0                     0.000000   \n",
       "4                      0.0                   0.0                     0.000000   \n",
       "5                      0.0                   0.0                     0.000000   \n",
       "...                    ...                   ...                          ...   \n",
       "606                    0.0                   0.0                     0.000000   \n",
       "607                    0.0                   0.0                     0.000000   \n",
       "608                    0.0                   0.0                     0.000000   \n",
       "609                    0.0                   0.0                     0.000000   \n",
       "610                    0.0                   0.0                    -0.166667   \n",
       "\n",
       "title   *batteries not included (1987)  ...  Zulu (2013)  [REC] (2007)  \\\n",
       "userId                                  ...                              \n",
       "1                                  0.0  ...          0.0        0.0000   \n",
       "2                                  0.0  ...          0.0        0.0000   \n",
       "3                                  0.0  ...          0.0        0.0000   \n",
       "4                                  0.0  ...          0.0        0.0000   \n",
       "5                                  0.0  ...          0.0        0.0000   \n",
       "...                                ...  ...          ...           ...   \n",
       "606                                0.0  ...          0.0        0.0000   \n",
       "607                                0.0  ...          0.0        0.0000   \n",
       "608                                0.0  ...          0.0        0.0000   \n",
       "609                                0.0  ...          0.0        0.0000   \n",
       "610                                0.0  ...          0.0       -0.0625   \n",
       "\n",
       "title   [REC]² (2009)  [REC]³ 3 Génesis (2012)  \\\n",
       "userId                                           \n",
       "1            0.000000                      0.0   \n",
       "2            0.000000                      0.0   \n",
       "3            0.000000                      0.0   \n",
       "4            0.000000                      0.0   \n",
       "5            0.000000                      0.0   \n",
       "...               ...                      ...   \n",
       "606          0.000000                      0.0   \n",
       "607          0.000000                      0.0   \n",
       "608          0.000000                      0.0   \n",
       "609          0.000000                      0.0   \n",
       "610         -0.166667                      0.0   \n",
       "\n",
       "title   anohana: The Flower We Saw That Day - The Movie (2013)  \\\n",
       "userId                                                           \n",
       "1                                                     0.0        \n",
       "2                                                     0.0        \n",
       "3                                                     0.0        \n",
       "4                                                     0.0        \n",
       "5                                                     0.0        \n",
       "...                                                   ...        \n",
       "606                                                   0.0        \n",
       "607                                                   0.0        \n",
       "608                                                   0.0        \n",
       "609                                                   0.0        \n",
       "610                                                   0.0        \n",
       "\n",
       "title   eXistenZ (1999)  xXx (2002)  xXx: State of the Union (2005)  \\\n",
       "userId                                                                \n",
       "1              0.000000    0.000000                             0.0   \n",
       "2              0.000000    0.000000                             0.0   \n",
       "3              0.000000    0.000000                             0.0   \n",
       "4              0.000000    0.000000                             0.0   \n",
       "5              0.000000    0.000000                             0.0   \n",
       "...                 ...         ...                             ...   \n",
       "606            0.000000    0.000000                             0.0   \n",
       "607            0.000000    0.000000                             0.0   \n",
       "608            0.636364    0.729167                             0.0   \n",
       "609            0.000000    0.000000                             0.0   \n",
       "610            0.000000   -0.770833                            -0.5   \n",
       "\n",
       "title   ¡Three Amigos! (1986)  À nous la liberté (Freedom for Us) (1931)  \n",
       "userId                                                                    \n",
       "1                    0.865385                                        0.0  \n",
       "2                    0.000000                                        0.0  \n",
       "3                    0.000000                                        0.0  \n",
       "4                    0.000000                                        0.0  \n",
       "5                    0.000000                                        0.0  \n",
       "...                       ...                                        ...  \n",
       "606                  0.000000                                        0.0  \n",
       "607                  0.000000                                        0.0  \n",
       "608                  0.000000                                        0.0  \n",
       "609                  0.000000                                        0.0  \n",
       "610                  0.000000                                        0.0  \n",
       "\n",
       "[610 rows x 9719 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings_pivot_standardized = data_ratings_pivot.sub(avg_ratings, axis=1)\n",
    "data_ratings_pivot_standardized = data_ratings_pivot_standardized.fillna(0)\n",
    "data_ratings_pivot_standardized"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Dopočítání hodnot\n",
    "\n",
    "Podívejme se třeba na hodnocení filmu Pulp Fiction od uživatele s id 1 a na něm si ukážeme, jak vznikly hodnoty v tabulce `data_ratings_pivot_standardized`. Film dostal od uživatele hodnocení 3 body. Nyní vidíme, že má hodnocení -1.197068. Kde se toto číslo vzalo? Rozeberme si postup výpočtu."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.0"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings_pivot.loc[1, \"Pulp Fiction (1994)\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "V tabulce `data_ratings_pivot_standardized` je hodnocení filmu Pulp Fiction -1.197068403908795."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1.197068403908795"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings_pivot_standardized.loc[1, \"Pulp Fiction (1994)\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Podívejme se na průměrné hodnocení tohoto filmu v původních tabulce `avg_ratings`. To má hodnotu 4.197068403908795."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.197068403908795"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "avg_ratings.loc[\"Pulp Fiction (1994)\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "K číslu se dopočítáme, pokud vezmeme hodnocení 3 a odečteme od něj průměrné hodnocení filmu. Tím získáme hodnotu -1.197068403908795, což je stejná hodnota, jaká byla v "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1.197068403908795"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "3 - 4.197068403908795"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Podobnost uživatelů\n",
    "\n",
    "Naším cílem nyní bude vyhledat k našemu uživateli č. 1 uživatel a uživatelky, kteří jsou mu (jí) nejvíce podobní. K výpočtu podobnosti opět můžeme využít kosinovou vzdálenost. Jednou z jejích hodnot je, že jí nevadí obrovské množství chybějících hodnot, jimi ovlivněná není. Pro výpočet podobnosti uživatelů využijeme tabulku se standardizovanými hodnoceními, tj. tabulku `data_ratings_pivot_standardized`. Výsledná data opět převedeme na tabulku."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>userId</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>601</th>\n",
       "      <th>602</th>\n",
       "      <th>603</th>\n",
       "      <th>604</th>\n",
       "      <th>605</th>\n",
       "      <th>606</th>\n",
       "      <th>607</th>\n",
       "      <th>608</th>\n",
       "      <th>609</th>\n",
       "      <th>610</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>userId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.005147</td>\n",
       "      <td>-0.045052</td>\n",
       "      <td>0.005082</td>\n",
       "      <td>0.018204</td>\n",
       "      <td>0.009896</td>\n",
       "      <td>-0.011553</td>\n",
       "      <td>0.027520</td>\n",
       "      <td>0.029854</td>\n",
       "      <td>-0.019043</td>\n",
       "      <td>...</td>\n",
       "      <td>0.030035</td>\n",
       "      <td>0.000167</td>\n",
       "      <td>-0.078620</td>\n",
       "      <td>-0.016158</td>\n",
       "      <td>-0.064228</td>\n",
       "      <td>-0.017452</td>\n",
       "      <td>0.036311</td>\n",
       "      <td>-0.092671</td>\n",
       "      <td>-0.005583</td>\n",
       "      <td>0.052643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.005147</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.018459</td>\n",
       "      <td>0.088285</td>\n",
       "      <td>-0.009396</td>\n",
       "      <td>0.008564</td>\n",
       "      <td>-0.035152</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.002004</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.062581</td>\n",
       "      <td>-0.019606</td>\n",
       "      <td>-0.003878</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.022969</td>\n",
       "      <td>-0.015052</td>\n",
       "      <td>0.000654</td>\n",
       "      <td>0.049596</td>\n",
       "      <td>0.051937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.045052</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.013393</td>\n",
       "      <td>-0.034685</td>\n",
       "      <td>0.015053</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.034564</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.064929</td>\n",
       "      <td>-0.013092</td>\n",
       "      <td>0.026124</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001096</td>\n",
       "      <td>-0.053438</td>\n",
       "      <td>-0.000739</td>\n",
       "      <td>0.007190</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.019865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.005082</td>\n",
       "      <td>-0.018459</td>\n",
       "      <td>-0.013393</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.059430</td>\n",
       "      <td>0.013483</td>\n",
       "      <td>0.070708</td>\n",
       "      <td>0.025238</td>\n",
       "      <td>-0.009606</td>\n",
       "      <td>0.102568</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.082818</td>\n",
       "      <td>0.057897</td>\n",
       "      <td>-0.015017</td>\n",
       "      <td>-0.005770</td>\n",
       "      <td>0.035938</td>\n",
       "      <td>-0.020737</td>\n",
       "      <td>0.011569</td>\n",
       "      <td>-0.029545</td>\n",
       "      <td>0.006883</td>\n",
       "      <td>-0.034780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.018204</td>\n",
       "      <td>0.088285</td>\n",
       "      <td>-0.034685</td>\n",
       "      <td>-0.059430</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.001382</td>\n",
       "      <td>0.003460</td>\n",
       "      <td>-0.128999</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.029349</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.010424</td>\n",
       "      <td>-0.064583</td>\n",
       "      <td>0.014181</td>\n",
       "      <td>-0.029109</td>\n",
       "      <td>-0.057051</td>\n",
       "      <td>0.010026</td>\n",
       "      <td>-0.001006</td>\n",
       "      <td>-0.024182</td>\n",
       "      <td>0.118276</td>\n",
       "      <td>-0.012240</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 610 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "userId       1         2         3         4         5         6         7    \\\n",
       "userId                                                                         \n",
       "1       1.000000  0.005147 -0.045052  0.005082  0.018204  0.009896 -0.011553   \n",
       "2       0.005147  1.000000  0.000000 -0.018459  0.088285 -0.009396  0.008564   \n",
       "3      -0.045052  0.000000  1.000000 -0.013393 -0.034685  0.015053  0.000000   \n",
       "4       0.005082 -0.018459 -0.013393  1.000000 -0.059430  0.013483  0.070708   \n",
       "5       0.018204  0.088285 -0.034685 -0.059430  1.000000 -0.001382  0.003460   \n",
       "\n",
       "userId       8         9         10   ...       601       602       603  \\\n",
       "userId                                ...                                 \n",
       "1       0.027520  0.029854 -0.019043  ...  0.030035  0.000167 -0.078620   \n",
       "2      -0.035152  0.000000 -0.002004  ... -0.062581 -0.019606 -0.003878   \n",
       "3      -0.034564  0.000000  0.000000  ... -0.064929 -0.013092  0.026124   \n",
       "4       0.025238 -0.009606  0.102568  ... -0.082818  0.057897 -0.015017   \n",
       "5      -0.128999  0.000000 -0.029349  ... -0.010424 -0.064583  0.014181   \n",
       "\n",
       "userId       604       605       606       607       608       609       610  \n",
       "userId                                                                        \n",
       "1      -0.016158 -0.064228 -0.017452  0.036311 -0.092671 -0.005583  0.052643  \n",
       "2       0.000000  0.000000  0.022969 -0.015052  0.000654  0.049596  0.051937  \n",
       "3       0.000000  0.001096 -0.053438 -0.000739  0.007190  0.000000  0.019865  \n",
       "4      -0.005770  0.035938 -0.020737  0.011569 -0.029545  0.006883 -0.034780  \n",
       "5      -0.029109 -0.057051  0.010026 -0.001006 -0.024182  0.118276 -0.012240  \n",
       "\n",
       "[5 rows x 610 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similarities = cosine_similarity(data_ratings_pivot_standardized)\n",
    "similarities = pd.DataFrame(similarities, index=data_ratings_pivot_standardized.index, columns=data_ratings_pivot_standardized.index)\n",
    "similarities.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Zobrazíme si, kteří uživatelé do 1 nejvíce podobných uživatelů patří. Nemáme žádný klíč k tomu, kolik uživatelů vybrat, zobrazíme tady 10 nejbližších. Do nich se počítá i sám uživatel 1, protože pro něj je kosinová podobnost 1 (tj. nejvyšší možná podobnost)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "userId\n",
       "2      1.000000\n",
       "189    0.174249\n",
       "145    0.170194\n",
       "524    0.120279\n",
       "468    0.104534\n",
       "564    0.103556\n",
       "379    0.093442\n",
       "5      0.088285\n",
       "329    0.087958\n",
       "378    0.086594\n",
       "Name: 2, dtype: float64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_users = 10\n",
    "user_1_similarities = similarities.loc[2]\n",
    "user_1_similarities = user_1_similarities.sort_values(ascending=False)\n",
    "user_1_similarities.head(n_users)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pomocí `iloc[]` nyní vyberu `n_users` nejvíce podobných uživatelů, samotného uživatele 1 už vyřadíme."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([189, 145, 524, 468, 564, 379, 5, 329, 378, 272], dtype='int64', name='userId')"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similar_users = user_1_similarities.iloc[1:n_users+1].index\n",
    "similar_users\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "V tabulce níže máme hodnocení filmů od 10 vybraných uživatelů. Aby bylo hodnocení relevantní, vybereme pouze filmy, které byly ohodnoceny alespoň od 3 těchto uživatelů. K tomu využijeme metodu `dropna()`, kterou již známe, ale tentokrát ji použijeme poměrně nezvykle.\n",
    "\n",
    "- Jako hodnotu parametru `axis` použijeme `1`, tj. chceme, aby metoda odebírala sloupečky, nikoli řádky. My totiž z výběru chceme vyřadit filmy, které mají málo hodnocení, a filmy jsou v tabulce jako sloupce.\n",
    "- Nechceme ale ponechat pouze filmy hodnocené od všech vybraných uživatelů, to by nám v tabulce moc filmů nezbylo. Proto použijeme parametr `tresh=3` (zkratka od anglického výrazu Treshold), který ponechá sloupečky, které mají alespoň tři hodnoty."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>title</th>\n",
       "      <th>Ace Ventura: Pet Detective (1994)</th>\n",
       "      <th>Aladdin (1992)</th>\n",
       "      <th>Apollo 13 (1995)</th>\n",
       "      <th>Batman (1989)</th>\n",
       "      <th>Batman Forever (1995)</th>\n",
       "      <th>Beauty and the Beast (1991)</th>\n",
       "      <th>Braveheart (1995)</th>\n",
       "      <th>Clear and Present Danger (1994)</th>\n",
       "      <th>Cliffhanger (1993)</th>\n",
       "      <th>Clueless (1995)</th>\n",
       "      <th>...</th>\n",
       "      <th>Seven (a.k.a. Se7en) (1995)</th>\n",
       "      <th>Shawshank Redemption, The (1994)</th>\n",
       "      <th>Silence of the Lambs, The (1991)</th>\n",
       "      <th>Stargate (1994)</th>\n",
       "      <th>Terminator 2: Judgment Day (1991)</th>\n",
       "      <th>Toy Story (1995)</th>\n",
       "      <th>True Lies (1994)</th>\n",
       "      <th>Truman Show, The (1998)</th>\n",
       "      <th>Usual Suspects, The (1995)</th>\n",
       "      <th>Waterworld (1995)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>userId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>189</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>524</th>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>468</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>564</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>379</th>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>329</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>378</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>4.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>272</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 36 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "title   Ace Ventura: Pet Detective (1994)  Aladdin (1992)  Apollo 13 (1995)  \\\n",
       "userId                                                                        \n",
       "189                                   NaN             NaN               NaN   \n",
       "145                                   2.0             4.0               3.0   \n",
       "524                                   3.0             4.0               5.0   \n",
       "468                                   3.0             3.0               5.0   \n",
       "564                                   NaN             NaN               NaN   \n",
       "379                                   2.0             NaN               5.0   \n",
       "5                                     3.0             4.0               3.0   \n",
       "329                                   NaN             NaN               NaN   \n",
       "378                                   NaN             NaN               NaN   \n",
       "272                                   NaN             NaN               NaN   \n",
       "\n",
       "title   Batman (1989)  Batman Forever (1995)  Beauty and the Beast (1991)  \\\n",
       "userId                                                                      \n",
       "189               NaN                    NaN                          NaN   \n",
       "145               3.0                    3.0                          4.0   \n",
       "524               3.0                    3.0                          NaN   \n",
       "468               3.0                    3.0                          5.0   \n",
       "564               NaN                    NaN                          NaN   \n",
       "379               3.0                    3.0                          NaN   \n",
       "5                 3.0                    3.0                          5.0   \n",
       "329               NaN                    NaN                          NaN   \n",
       "378               NaN                    NaN                          NaN   \n",
       "272               NaN                    NaN                          NaN   \n",
       "\n",
       "title   Braveheart (1995)  Clear and Present Danger (1994)  \\\n",
       "userId                                                       \n",
       "189                   NaN                              NaN   \n",
       "145                   NaN                              NaN   \n",
       "524                   3.0                              5.0   \n",
       "468                   4.0                              2.0   \n",
       "564                   NaN                              NaN   \n",
       "379                   5.0                              4.0   \n",
       "5                     4.0                              3.0   \n",
       "329                   NaN                              NaN   \n",
       "378                   NaN                              NaN   \n",
       "272                   NaN                              NaN   \n",
       "\n",
       "title   Cliffhanger (1993)  Clueless (1995)  ...  Seven (a.k.a. Se7en) (1995)  \\\n",
       "userId                                       ...                                \n",
       "189                    NaN              NaN  ...                          NaN   \n",
       "145                    3.0              NaN  ...                          NaN   \n",
       "524                    NaN              NaN  ...                          5.0   \n",
       "468                    3.0              5.0  ...                          5.0   \n",
       "564                    NaN              3.5  ...                          NaN   \n",
       "379                    3.0              NaN  ...                          3.0   \n",
       "5                      NaN              3.0  ...                          NaN   \n",
       "329                    NaN              NaN  ...                          NaN   \n",
       "378                    NaN              NaN  ...                          NaN   \n",
       "272                    NaN              NaN  ...                          NaN   \n",
       "\n",
       "title   Shawshank Redemption, The (1994)  Silence of the Lambs, The (1991)  \\\n",
       "userId                                                                       \n",
       "189                                  4.0                               4.0   \n",
       "145                                  3.0                               NaN   \n",
       "524                                  1.0                               4.0   \n",
       "468                                  3.0                               NaN   \n",
       "564                                  3.0                               NaN   \n",
       "379                                  3.0                               NaN   \n",
       "5                                    3.0                               NaN   \n",
       "329                                  2.0                               2.0   \n",
       "378                                  4.5                               4.5   \n",
       "272                                  4.0                               NaN   \n",
       "\n",
       "title   Stargate (1994)  Terminator 2: Judgment Day (1991)  Toy Story (1995)  \\\n",
       "userId                                                                         \n",
       "189                 NaN                                NaN               NaN   \n",
       "145                 2.0                                NaN               5.0   \n",
       "524                 3.0                                5.0               4.0   \n",
       "468                 NaN                                NaN               4.0   \n",
       "564                 NaN                                NaN               NaN   \n",
       "379                 3.0                                4.0               NaN   \n",
       "5                   2.0                                3.0               4.0   \n",
       "329                 NaN                                NaN               NaN   \n",
       "378                 NaN                                NaN               4.5   \n",
       "272                 NaN                                3.5               NaN   \n",
       "\n",
       "title   True Lies (1994)  Truman Show, The (1998)  Usual Suspects, The (1995)  \\\n",
       "userId                                                                          \n",
       "189                  NaN                      NaN                         NaN   \n",
       "145                  3.0                      NaN                         5.0   \n",
       "524                  4.0                      NaN                         2.0   \n",
       "468                  4.0                      NaN                         NaN   \n",
       "564                  NaN                      1.0                         NaN   \n",
       "379                  4.0                      NaN                         NaN   \n",
       "5                    2.0                      NaN                         4.0   \n",
       "329                  NaN                      NaN                         2.0   \n",
       "378                  NaN                      4.0                         NaN   \n",
       "272                  NaN                      4.0                         NaN   \n",
       "\n",
       "title   Waterworld (1995)  \n",
       "userId                     \n",
       "189                   NaN  \n",
       "145                   3.0  \n",
       "524                   1.0  \n",
       "468                   3.0  \n",
       "564                   NaN  \n",
       "379                   3.0  \n",
       "5                     NaN  \n",
       "329                   NaN  \n",
       "378                   NaN  \n",
       "272                   NaN  \n",
       "\n",
       "[10 rows x 36 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_ratings_pivot_similar = data_ratings_pivot.loc[similar_users]\n",
    "data_ratings_pivot_similar = data_ratings_pivot_similar.dropna(axis=1, thresh=3)\n",
    "data_ratings_pivot_similar"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nakonec spočítáme průměrná hodnocení pro jednotlivé filmy a seřadíme je. U těchto filmů tedy můžeme předpokládat větší šanci, že by se mohly líbit i uživateli 2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title\n",
       "Dances with Wolves (1990)                                    4.750000\n",
       "Beauty and the Beast (1991)                                  4.666667\n",
       "Fugitive, The (1993)                                         4.666667\n",
       "Seven (a.k.a. Se7en) (1995)                                  4.333333\n",
       "Toy Story (1995)                                             4.300000\n",
       "Schindler's List (1993)                                      4.250000\n",
       "Apollo 13 (1995)                                             4.200000\n",
       "Forrest Gump (1994)                                          4.000000\n",
       "Braveheart (1995)                                            4.000000\n",
       "Terminator 2: Judgment Day (1991)                            3.875000\n",
       "Clueless (1995)                                              3.833333\n",
       "Crimson Tide (1995)                                          3.750000\n",
       "Aladdin (1992)                                               3.750000\n",
       "Mask, The (1994)                                             3.666667\n",
       "Outbreak (1995)                                              3.666667\n",
       "Silence of the Lambs, The (1991)                             3.625000\n",
       "Pulp Fiction (1994)                                          3.500000\n",
       "Lion King, The (1994)                                        3.500000\n",
       "Clear and Present Danger (1994)                              3.500000\n",
       "Lord of the Rings: The Fellowship of the Ring, The (2001)    3.500000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "average_ratins_by_similar_users = data_ratings_pivot_similar.mean().sort_values(ascending=False)\n",
    "average_ratins_by_similar_users.head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Čtení na doma: Odebrání filmů, které uživatel již viděl\n",
    "\n",
    "Většině uživatelů asi nemá smysl doporučovat filmy, které už viděli, protože dají přednost něčemu novému. Můžeme tedy z našeho seznamu odebrat filmy, které už uživatel 2 ohodnotil. Nejprve si zjistíme, které to jsou. Pokud vybereme z tabulky hodnocení uživatele 2 a odebereme prázdné hodnoty, pak indexy vybrané série budou představovat filmy, které uživatel již viděl. Výsledek si uložíme do proměnné `seen_by_user_2`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Collateral (2004)', 'Dark Knight Rises, The (2012)',\n",
       "       'Dark Knight, The (2008)', 'Departed, The (2006)',\n",
       "       'Django Unchained (2012)', 'Ex Machina (2015)',\n",
       "       'Exit Through the Gift Shop (2010)',\n",
       "       'Girl with the Dragon Tattoo, The (2011)', 'Gladiator (2000)',\n",
       "       'Good Will Hunting (1997)', 'Inception (2010)',\n",
       "       'Inglourious Basterds (2009)', 'Inside Job (2010)',\n",
       "       'Interstellar (2014)', 'Kill Bill: Vol. 1 (2003)',\n",
       "       'Louis C.K.: Hilarious (2010)', 'Mad Max: Fury Road (2015)',\n",
       "       'Shawshank Redemption, The (1994)', 'Shutter Island (2010)',\n",
       "       'Step Brothers (2008)',\n",
       "       'Talladega Nights: The Ballad of Ricky Bobby (2006)', 'The Drop (2014)',\n",
       "       'The Jinx: The Life and Deaths of Robert Durst (2015)',\n",
       "       'Tommy Boy (1995)', 'Town, The (2010)', 'Warrior (2011)',\n",
       "       'Whiplash (2014)', 'Wolf of Wall Street, The (2013)',\n",
       "       'Zombieland (2009)'],\n",
       "      dtype='object', name='title')"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seen_by_user_2 = data_ratings_pivot.loc[2].dropna().index\n",
    "seen_by_user_2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nakonec tyto filmy odebereme z filmů, které chceme doporučit. Pomocí dotazu vybereme ze série `average_ratins_by_similar_users` položky, jejichž index (i u této série jsou index názvy filmů) není v  `seen_by_user_2`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title\n",
       "Dances with Wolves (1990)                                    4.750000\n",
       "Beauty and the Beast (1991)                                  4.666667\n",
       "Fugitive, The (1993)                                         4.666667\n",
       "Seven (a.k.a. Se7en) (1995)                                  4.333333\n",
       "Toy Story (1995)                                             4.300000\n",
       "Schindler's List (1993)                                      4.250000\n",
       "Apollo 13 (1995)                                             4.200000\n",
       "Forrest Gump (1994)                                          4.000000\n",
       "Braveheart (1995)                                            4.000000\n",
       "Terminator 2: Judgment Day (1991)                            3.875000\n",
       "Clueless (1995)                                              3.833333\n",
       "Crimson Tide (1995)                                          3.750000\n",
       "Aladdin (1992)                                               3.750000\n",
       "Mask, The (1994)                                             3.666667\n",
       "Outbreak (1995)                                              3.666667\n",
       "Silence of the Lambs, The (1991)                             3.625000\n",
       "Pulp Fiction (1994)                                          3.500000\n",
       "Lion King, The (1994)                                        3.500000\n",
       "Clear and Present Danger (1994)                              3.500000\n",
       "Lord of the Rings: The Fellowship of the Ring, The (2001)    3.500000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "average_ratins_by_similar_users = average_ratins_by_similar_users[~average_ratins_by_similar_users.index.isin(seen_by_user_2)]\n",
    "average_ratins_by_similar_users.head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Zdroje a odkazy\n",
    "\n",
    "- [Understanding Cosine Similarity in Python with Scikit-Learn](https://memgraph.com/blog/cosine-similarity-python-scikit-learn)\n",
    "- [What is the 'More Like This' section?](https://help.imdb.com/article/imdb/discover-watch/what-is-the-more-like-this-section/GPE7SPGZREKKY7YN?ref_=cons_tt_rec_lm#)\n",
    "- [Beginner Tutorial: Recommender Systems in Python](https://www.datacamp.com/tutorial/recommender-systems-python)\n",
    "- [Build a Recommendation Engine With Collaborative Filtering](https://realpython.com/build-recommendation-engine-collaborative-filtering/)\n",
    "- [How Recommendation Systems Tackle the Cold Start Problem](https://www.yusp.com/blog-posts/cold-start-problem/)\n",
    "- [Overcoming Echo Chambers in Recommendation Systems (Using Movie Ratings)](https://towardsdatascience.com/overcoming-echo-chambers-in-recommendation-systems-using-movie-ratings-3686f50fc053) (placená část Towards Data Science)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cvičení\n",
    "\n",
    "Náš systém se orientoval čistě podle popisu děje filmu. Pokud nemáme k dispozici nic jiného, než slovní popis, je často nutné pracovat s ním. V našem datasetu ale máme ale další strukturovaná data, například žánr filmu. Ten může být užitečný. Vraťme se třeba k filmu The Fugitive. Uživatel nebo uživatelka, kterému (které) se tento film líbil, může preferovat filmy s podobnými žánry. Hypotetická romantická komedie o uprchlém vězni, který se na útěku zamiluje, může být méně lákavá než jiný akční thriller s detektivní zápletkou.\n",
    "\n",
    "Žánry filmu máme v datech uvedeny ve sloupečcích, které začínaji `gen_`. \n",
    "\n",
    "Načti si nejprve znovu data a ponech si v nich pouze sloupce se žánry pomocí kódu níže."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gen_Action</th>\n",
       "      <th>gen_Adventure</th>\n",
       "      <th>gen_Animation</th>\n",
       "      <th>gen_Comedy</th>\n",
       "      <th>gen_Crime</th>\n",
       "      <th>gen_Documentary</th>\n",
       "      <th>gen_Drama</th>\n",
       "      <th>gen_Family</th>\n",
       "      <th>gen_Fantasy</th>\n",
       "      <th>gen_Foreign</th>\n",
       "      <th>gen_History</th>\n",
       "      <th>gen_Horror</th>\n",
       "      <th>gen_Music</th>\n",
       "      <th>gen_Mystery</th>\n",
       "      <th>gen_Romance</th>\n",
       "      <th>gen_Science Fiction</th>\n",
       "      <th>gen_TV Movie</th>\n",
       "      <th>gen_Thriller</th>\n",
       "      <th>gen_War</th>\n",
       "      <th>gen_Western</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Avatar</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pirates of the Caribbean: At World's End</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spectre</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Dark Knight Rises</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>John Carter</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          gen_Action  gen_Adventure  \\\n",
       "title                                                                 \n",
       "Avatar                                           1.0            1.0   \n",
       "Pirates of the Caribbean: At World's End         1.0            1.0   \n",
       "Spectre                                          1.0            1.0   \n",
       "The Dark Knight Rises                            1.0            0.0   \n",
       "John Carter                                      1.0            1.0   \n",
       "\n",
       "                                          gen_Animation  gen_Comedy  \\\n",
       "title                                                                 \n",
       "Avatar                                              0.0         0.0   \n",
       "Pirates of the Caribbean: At World's End            0.0         0.0   \n",
       "Spectre                                             0.0         0.0   \n",
       "The Dark Knight Rises                               0.0         0.0   \n",
       "John Carter                                         0.0         0.0   \n",
       "\n",
       "                                          gen_Crime  gen_Documentary  \\\n",
       "title                                                                  \n",
       "Avatar                                          0.0              0.0   \n",
       "Pirates of the Caribbean: At World's End        0.0              0.0   \n",
       "Spectre                                         1.0              0.0   \n",
       "The Dark Knight Rises                           1.0              0.0   \n",
       "John Carter                                     0.0              0.0   \n",
       "\n",
       "                                          gen_Drama  gen_Family  gen_Fantasy  \\\n",
       "title                                                                          \n",
       "Avatar                                          0.0         0.0          1.0   \n",
       "Pirates of the Caribbean: At World's End        0.0         0.0          1.0   \n",
       "Spectre                                         0.0         0.0          0.0   \n",
       "The Dark Knight Rises                           1.0         0.0          0.0   \n",
       "John Carter                                     0.0         0.0          0.0   \n",
       "\n",
       "                                          gen_Foreign  gen_History  \\\n",
       "title                                                                \n",
       "Avatar                                            0.0          0.0   \n",
       "Pirates of the Caribbean: At World's End          0.0          0.0   \n",
       "Spectre                                           0.0          0.0   \n",
       "The Dark Knight Rises                             0.0          0.0   \n",
       "John Carter                                       0.0          0.0   \n",
       "\n",
       "                                          gen_Horror  gen_Music  gen_Mystery  \\\n",
       "title                                                                          \n",
       "Avatar                                           0.0        0.0          0.0   \n",
       "Pirates of the Caribbean: At World's End         0.0        0.0          0.0   \n",
       "Spectre                                          0.0        0.0          0.0   \n",
       "The Dark Knight Rises                            0.0        0.0          0.0   \n",
       "John Carter                                      0.0        0.0          0.0   \n",
       "\n",
       "                                          gen_Romance  gen_Science Fiction  \\\n",
       "title                                                                        \n",
       "Avatar                                            0.0                  1.0   \n",
       "Pirates of the Caribbean: At World's End          0.0                  0.0   \n",
       "Spectre                                           0.0                  0.0   \n",
       "The Dark Knight Rises                             0.0                  0.0   \n",
       "John Carter                                       0.0                  1.0   \n",
       "\n",
       "                                          gen_TV Movie  gen_Thriller  gen_War  \\\n",
       "title                                                                           \n",
       "Avatar                                             0.0           0.0      0.0   \n",
       "Pirates of the Caribbean: At World's End           0.0           0.0      0.0   \n",
       "Spectre                                            0.0           0.0      0.0   \n",
       "The Dark Knight Rises                              0.0           1.0      0.0   \n",
       "John Carter                                        0.0           0.0      0.0   \n",
       "\n",
       "                                          gen_Western  \n",
       "title                                                  \n",
       "Avatar                                            0.0  \n",
       "Pirates of the Caribbean: At World's End          0.0  \n",
       "Spectre                                           0.0  \n",
       "The Dark Knight Rises                             0.0  \n",
       "John Carter                                       0.0  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"tmdb_5000_movies.csv\")\n",
    "# Nastavíme index\n",
    "data = data.set_index(\"title\")\n",
    "# Ponecháme v datech pouze sloupce od gen_Action po gen_Western\n",
    "data_genres = data.loc[:, \"gen_Action\":\"gen_Western\"]\n",
    "data_genres.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Podívej se na žánry filmu The Fugitive s využitím `.loc`.\n",
    "\n",
    "Žárny filmů vyzkoušíme porovnat nikoli pomocí cosinové vzdálenosti, ale vyzkoušíme Jaccardovovo skóre. Jaccardovo skóre je založeno na následujícím vzorečku: (počet žánrů, které mají filmy společné / počet žánrů, které se vyskytují alespoň u jednoho filmu).\n",
    "\n",
    "Uvažuj třeba hypotetický film A, který má žánr Comedy a Romance, a film B, který má žárn Comedy a History. Oba filmy mají společný jeden žánr (Comedy). Celkem se u obou filmů objevují žárny Comedy, Romance a History, tj. tři žánry. Jaccardovo skóre pro tyto dva filmy je 1/3.\n",
    "\n",
    "Proveď import funkce `jaccard_score` pomocí příkazu níže."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import jaccard_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nyní zkus spočítat Jaccardovo skóre pro filmy The Fugitive a Witness. Uděláš to tak, že funkci `jaccard_score` dáš dva parametry - sérii s žánry filmu The Fugitive a sérii s žánry filmu Witness. Sérii s žánry The Fugitive už jsi získala v předchozím kroku a sérii s žánry Witness získáš stejným způsobem.\n",
    "\n",
    "Měla by ti vyjít hodnota přibližně 0.2857.\n",
    "\n",
    "Dále zkus to samé pro film The Fugitive a Titanic. Mělo by ti vyjít přibližně 0.142856.\n",
    "\n",
    "Nyní zkus vypočítat Jaccardovy vzdálenosti pro všechny filmy. Proveď následující importy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.spatial.distance import pdist, squareform"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nejprve použij funkci `pdist`, která vypočítá Jaccardovy vzdálenosti pro všechny dvojice filmů. Funkci dej jako první parametr matici `data` a jako druhý parametr `metric=\"jaccard\"`. Výsledek ulož do proměnné `jaccard_distances`. Výsledkem je jednorozměrné pole.\n",
    "\n",
    "Jako druhý krok použij funkci `squareform`, která převede jednorozměrné pole na dvourozměrnou matici. Funkci volej pouze s jedním parametrem - proměnnou `jaccard_distances`. Výsledek opět ulož do proměnné s názvem `jaccard_distances`. Nakonec je je potřeba spustit tento řádek\n",
    "\n",
    "```py\n",
    "jaccard_distances = 1 - jaccard_distances\n",
    "```\n",
    "\n",
    "Funkce totiž počítá vzdálenosti, nikoli Jaccardovo skóre. Pro Jaccardovo skóre a Jaccardovu vzdálenost platí vztah: Jaccardovo skóre = 1 - Jaccardova vzdálenost.\n",
    "\n",
    "Výsledek poté převeď na pandas tabulku stejným způsobem, jako jsme to dělali v lekci. Jako index (parametr `index`) a názvy sloupců (parametr `columns`) použij indexy z původní tabulky data (tj. `data.index`).\n",
    "\n",
    "Poté vyber sloupec s filmem The Fugitive a seřaď hodnoty v něm od největší po nejmenší. Filmy nahoře jsou nyní nejvíce podobné a algortimus postavený na žánru filmu by je doporučil.\n",
    "\n",
    "### Bonus\n",
    "\n",
    "Podívej se na kód níže. Obsahuje metody a funkce, které jsme si ukazovali dříve, především v prvních třech lekcích. Dokážeš si vzpomenout, co tyto funkce dělají, a odhadnout výsledek? Pokud ne, podívej se na výstup níže."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['release_date'] = pd.to_datetime(data['release_date'])\n",
    "data['release_year'] = data['release_date'].dt.year\n",
    "\n",
    "bins = [1979, 1989, 1999, 2009, 2019, 2029]\n",
    "labels = ['1980s', '1990s', '2000s', '2010s', '2020s']\n",
    "\n",
    "data['decade'] = pd.cut(data['release_year'], bins=bins, labels=labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Níže je nápověda, která zobrazuje nově přidaný sloupec `decade`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>release_date</th>\n",
       "      <th>release_year</th>\n",
       "      <th>decade</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>title</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Avatar</th>\n",
       "      <td>2009-12-10</td>\n",
       "      <td>2009.0</td>\n",
       "      <td>2000s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pirates of the Caribbean: At World's End</th>\n",
       "      <td>2007-05-19</td>\n",
       "      <td>2007.0</td>\n",
       "      <td>2000s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spectre</th>\n",
       "      <td>2015-10-26</td>\n",
       "      <td>2015.0</td>\n",
       "      <td>2010s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Dark Knight Rises</th>\n",
       "      <td>2012-07-16</td>\n",
       "      <td>2012.0</td>\n",
       "      <td>2010s</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>John Carter</th>\n",
       "      <td>2012-03-07</td>\n",
       "      <td>2012.0</td>\n",
       "      <td>2010s</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         release_date  release_year decade\n",
       "title                                                                     \n",
       "Avatar                                     2009-12-10        2009.0  2000s\n",
       "Pirates of the Caribbean: At World's End   2007-05-19        2007.0  2000s\n",
       "Spectre                                    2015-10-26        2015.0  2010s\n",
       "The Dark Knight Rises                      2012-07-16        2012.0  2010s\n",
       "John Carter                                2012-03-07        2012.0  2010s"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[['release_date', 'release_year', 'decade']].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Níže jsou komentáře k jednotlivým řádkům"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Převod sloupce 'release_date' na datový typ datetime\n",
    "data['release_date'] = pd.to_datetime(data['release_date'])\n",
    "\n",
    "# Extrahování roku z 'release_date' a uložení do nového sloupce 'release_year'\n",
    "data['release_year'] = data['release_date'].dt.year\n",
    "\n",
    "# Definice rozmezí pro dekády\n",
    "bins = [1979, 1989, 1999, 2009, 2019, 2029]\n",
    "\n",
    "# Definice popisek pro jednotlivé dekády\n",
    "labels = ['1980s', '1990s', '2000s', '2010s', '2020s']\n",
    "\n",
    "# Kategorizace roku vydání filmu do dekád a přidání do nového sloupce 'decade'\n",
    "data['decade'] = pd.cut(data['release_year'], bins=bins, labels=labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dekádu jsme zjišťovali, protože každá dekáda měla z pohledu kinematografie jistá specifika a například v devadesátých letech se natáčely filmy jinak, než jak se natáčí dnes. A pokud se někomu líbí spíše starší filmy, pravděpodobně mu bude lepší doporučit nějaký starší film. Upravíme tedy náš systém tak, aby doporučoval pouze filmy ze stjené dekády.\n",
    "\n",
    "Nejprve do tabulky `jaccard_distances` vlož sloupec `decade` z tabulky `data`. Pokud si nejsi jistý(á), jak to provést, použij řádek níže.\n",
    "\n",
    "```py\n",
    "jaccard_distances[\"decade\"] = data[\"decade\"]\n",
    "```\n",
    "\n",
    "Dále se vraťme k filmu The Fugitive. Z lekce víc, z jaké dekády tento film pochází, ale můžeš to zkusit zjistit pomocí `.loc`. Dále vytvoř tabulku `movies_from_decade`, která bude obsahovat pouze filmy ze stejné dekády jako The Fugitive. Nakonec z této tabulky vyber filmy, které jsou svými žánry nejvíce podobné filme The Fugitive. K tomu můžeš použít stejný kód jaký na závěru předchozího příkladu, pouze jej použij s tabulkou `movies_from_decade`."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}