train_run_models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://www.kaggle.com/datasets/grouplens/movielens-20m-dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "!if [ ! -d movielens-20m-dataset ]; then kaggle datasets download -d grouplens/movielens-20m-dataset; fi\n",
    "\n",
    "!if [ -f movielens-20m-dataset.zip ]; then unzip movielens-20m-dataset.zip; rm movielens-20m-dataset.zip; mkdir movielens-20m-dataset; mv *.csv movielens-20m-dataset/; fi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from datetime import datetime\n",
    "import os\n",
    "import sys\n",
    "import shutil\n",
    "\n",
    "import optuna\n",
    "\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "from sklearn.utils import shuffle\n",
    "\n",
    "from pyspark.mllib.recommendation import ALS, MatrixFactorizationModel, Rating\n",
    "from pyspark import SparkContext, SparkConf\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "plt.style.use('seaborn-v0_8-whitegrid')"
   ]
  },
  {
   "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>userId</th>\n",
       "      <th>movieId</th>\n",
       "      <th>rating</th>\n",
       "      <th>timestamp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:53:47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>29</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:31:16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>32</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:33:39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>47</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:32:07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>50</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:29:40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000258</th>\n",
       "      <td>138493</td>\n",
       "      <td>68954</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-11-13 15:42:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000259</th>\n",
       "      <td>138493</td>\n",
       "      <td>69526</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-12-03 18:31:48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000260</th>\n",
       "      <td>138493</td>\n",
       "      <td>69644</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2009-12-07 18:10:57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000261</th>\n",
       "      <td>138493</td>\n",
       "      <td>70286</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2009-11-13 15:42:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000262</th>\n",
       "      <td>138493</td>\n",
       "      <td>71619</td>\n",
       "      <td>2.5</td>\n",
       "      <td>2009-10-17 20:25:36</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>20000263 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          userId  movieId  rating           timestamp\n",
       "0              1        2     3.5 2005-04-02 23:53:47\n",
       "1              1       29     3.5 2005-04-02 23:31:16\n",
       "2              1       32     3.5 2005-04-02 23:33:39\n",
       "3              1       47     3.5 2005-04-02 23:32:07\n",
       "4              1       50     3.5 2005-04-02 23:29:40\n",
       "...          ...      ...     ...                 ...\n",
       "20000258  138493    68954     4.5 2009-11-13 15:42:00\n",
       "20000259  138493    69526     4.5 2009-12-03 18:31:48\n",
       "20000260  138493    69644     3.0 2009-12-07 18:10:57\n",
       "20000261  138493    70286     5.0 2009-11-13 15:42:24\n",
       "20000262  138493    71619     2.5 2009-10-17 20:25:36\n",
       "\n",
       "[20000263 rows x 4 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('movielens-20m-dataset/rating.csv', parse_dates=['timestamp'])\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 20000263 entries, 0 to 20000262\n",
      "Data columns (total 4 columns):\n",
      " #   Column     Dtype         \n",
      "---  ------     -----         \n",
      " 0   userId     int64         \n",
      " 1   movieId    int64         \n",
      " 2   rating     float64       \n",
      " 3   timestamp  datetime64[ns]\n",
      "dtypes: datetime64[ns](1), float64(1), int64(2)\n",
      "memory usage: 610.4 MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create sequential IDs for users and movies, that begin at 0. This is needed by some models. Original IDs are sparse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>movieId</th>\n",
       "      <th>rating</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>userIdseq</th>\n",
       "      <th>movieIdseq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:53:47</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>29</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:31:16</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>32</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:33:39</td>\n",
       "      <td>0</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>47</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:32:07</td>\n",
       "      <td>0</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>50</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2005-04-02 23:29:40</td>\n",
       "      <td>0</td>\n",
       "      <td>49</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000258</th>\n",
       "      <td>138493</td>\n",
       "      <td>68954</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-11-13 15:42:00</td>\n",
       "      <td>138492</td>\n",
       "      <td>13754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000259</th>\n",
       "      <td>138493</td>\n",
       "      <td>69526</td>\n",
       "      <td>4.5</td>\n",
       "      <td>2009-12-03 18:31:48</td>\n",
       "      <td>138492</td>\n",
       "      <td>13862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000260</th>\n",
       "      <td>138493</td>\n",
       "      <td>69644</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2009-12-07 18:10:57</td>\n",
       "      <td>138492</td>\n",
       "      <td>13875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000261</th>\n",
       "      <td>138493</td>\n",
       "      <td>70286</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2009-11-13 15:42:24</td>\n",
       "      <td>138492</td>\n",
       "      <td>13993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000262</th>\n",
       "      <td>138493</td>\n",
       "      <td>71619</td>\n",
       "      <td>2.5</td>\n",
       "      <td>2009-10-17 20:25:36</td>\n",
       "      <td>138492</td>\n",
       "      <td>14277</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>20000263 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          userId  movieId  rating           timestamp  userIdseq  movieIdseq\n",
       "0              1        2     3.5 2005-04-02 23:53:47          0           1\n",
       "1              1       29     3.5 2005-04-02 23:31:16          0          28\n",
       "2              1       32     3.5 2005-04-02 23:33:39          0          31\n",
       "3              1       47     3.5 2005-04-02 23:32:07          0          46\n",
       "4              1       50     3.5 2005-04-02 23:29:40          0          49\n",
       "...          ...      ...     ...                 ...        ...         ...\n",
       "20000258  138493    68954     4.5 2009-11-13 15:42:00     138492       13754\n",
       "20000259  138493    69526     4.5 2009-12-03 18:31:48     138492       13862\n",
       "20000260  138493    69644     3.0 2009-12-07 18:10:57     138492       13875\n",
       "20000261  138493    70286     5.0 2009-11-13 15:42:24     138492       13993\n",
       "20000262  138493    71619     2.5 2009-10-17 20:25:36     138492       14277\n",
       "\n",
       "[20000263 rows x 6 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['userIdseq'] = pd.Categorical(df['userId'])\n",
    "df['userIdseq'] = df['userIdseq'].cat.codes\n",
    "df['movieIdseq'] = pd.Categorical(df['movieId'])\n",
    "df['movieIdseq'] = df['movieIdseq'].cat.codes\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>movieId</th>\n",
       "      <th>rating</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>userIdseq</th>\n",
       "      <th>movieIdseq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>28507</td>\n",
       "      <td>1176</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1995-01-09 11:46:44</td>\n",
       "      <td>28506</td>\n",
       "      <td>1153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>131160</td>\n",
       "      <td>1079</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1995-01-09 11:46:49</td>\n",
       "      <td>131159</td>\n",
       "      <td>1057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>131160</td>\n",
       "      <td>47</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1995-01-09 11:46:49</td>\n",
       "      <td>131159</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>131160</td>\n",
       "      <td>21</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1995-01-09 11:46:49</td>\n",
       "      <td>131159</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>85252</td>\n",
       "      <td>45</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1996-01-29 00:00:00</td>\n",
       "      <td>85251</td>\n",
       "      <td>44</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000258</th>\n",
       "      <td>53930</td>\n",
       "      <td>118706</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2015-03-31 06:00:51</td>\n",
       "      <td>53929</td>\n",
       "      <td>24832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000259</th>\n",
       "      <td>16978</td>\n",
       "      <td>2093</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2015-03-31 06:03:17</td>\n",
       "      <td>16977</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000260</th>\n",
       "      <td>89081</td>\n",
       "      <td>55232</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2015-03-31 06:11:26</td>\n",
       "      <td>89080</td>\n",
       "      <td>12118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000261</th>\n",
       "      <td>89081</td>\n",
       "      <td>52458</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2015-03-31 06:11:28</td>\n",
       "      <td>89080</td>\n",
       "      <td>11782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20000262</th>\n",
       "      <td>87586</td>\n",
       "      <td>7151</td>\n",
       "      <td>3.5</td>\n",
       "      <td>2015-03-31 06:40:02</td>\n",
       "      <td>87585</td>\n",
       "      <td>7039</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>20000263 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          userId  movieId  rating           timestamp  userIdseq  movieIdseq\n",
       "0          28507     1176     4.0 1995-01-09 11:46:44      28506        1153\n",
       "1         131160     1079     3.0 1995-01-09 11:46:49     131159        1057\n",
       "2         131160       47     5.0 1995-01-09 11:46:49     131159          46\n",
       "3         131160       21     3.0 1995-01-09 11:46:49     131159          20\n",
       "4          85252       45     3.0 1996-01-29 00:00:00      85251          44\n",
       "...          ...      ...     ...                 ...        ...         ...\n",
       "20000258   53930   118706     3.5 2015-03-31 06:00:51      53929       24832\n",
       "20000259   16978     2093     3.5 2015-03-31 06:03:17      16977        2009\n",
       "20000260   89081    55232     3.5 2015-03-31 06:11:26      89080       12118\n",
       "20000261   89081    52458     4.0 2015-03-31 06:11:28      89080       11782\n",
       "20000262   87586     7151     3.5 2015-03-31 06:40:02      87585        7039\n",
       "\n",
       "[20000263 rows x 6 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sort_values(by=['timestamp'], inplace=True)\n",
    "df.reset_index(drop=True, inplace=True)\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create translation dictionaries from sequential IDs back to original IDs. Models operate in sequential ID space. Results must be brought back into the original ID space."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "user_idseq2id = df[['userId', 'userIdseq']].drop_duplicates(\n",
    "    subset=['userIdseq']\n",
    ")\n",
    "movie_idseq2id = df[['movieId', 'movieIdseq']].drop_duplicates(\n",
    "    subset=['movieIdseq']\n",
    ")\n",
    "user_idseq2id = pd.Series(\n",
    "    user_idseq2id['userId'].values, index=user_idseq2id['userIdseq']\n",
    ").to_dict()\n",
    "movie_idseq2id = pd.Series(\n",
    "    movie_idseq2id['movieId'].values, index=movie_idseq2id['movieIdseq']\n",
    ").to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nr. of users:        138,493\n",
      "nr. of movies:       26,744\n",
      "nr. of observations: 20,000,263\n"
     ]
    }
   ],
   "source": [
    "N = len(set(df['userIdseq'].unique().tolist()))\n",
    "M = len(set(df['movieIdseq'].unique().tolist()))\n",
    "N_ratings = len(df)\n",
    "\n",
    "print(f'nr. of users:        {N:,}')\n",
    "print(f'nr. of movies:       {M:,}')\n",
    "print(f'nr. of observations: {N_ratings:,}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train/test split. Use a calendar date for the split."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_1st_day = '2014-01-01'\n",
    "\n",
    "df_train = df.loc[df['timestamp'] < train_1st_day]\n",
    "df_test = df.loc[df['timestamp'] >= train_1st_day]\n",
    "\n",
    "df_seq_train_fn = 'movielens-20m-dataset/rating-seq-train.csv'\n",
    "df_seq_test_fn = 'movielens-20m-dataset/rating-seq-test.csv'\n",
    "\n",
    "if not os.path.isfile(df_seq_train_fn):\n",
    "    df_train.to_csv(df_seq_train_fn, index=False)\n",
    "if not os.path.isfile(df_seq_test_fn):\n",
    "    df_test.to_csv(df_seq_test_fn, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nr. of observations training:   19,153,489\n",
      "nr. of observations testing:    846,774\n",
      "training set size / total size: 0.9576618567465838\n"
     ]
    }
   ],
   "source": [
    "user_ids_train = df_train['userIdseq'].values\n",
    "movie_ids_train = df_train['movieIdseq'].values\n",
    "ratings_train = df_train['rating'].values / df['rating'].max()\n",
    "\n",
    "user_ids_test = df_test['userIdseq'].values\n",
    "movie_ids_test = df_test['movieIdseq'].values\n",
    "ratings_test = df_test['rating'].values / df['rating'].max()\n",
    "\n",
    "train_fraq = len(user_ids_train) / len(df)\n",
    "\n",
    "print(f'nr. of observations training:   {len(user_ids_train):,}')\n",
    "print(f'nr. of observations testing:    {len(user_ids_test):,}')\n",
    "print(f'training set size / total size: {train_fraq}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load additional info about movies. Not used by the models (for now)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>genres</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>movieId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Toy Story (1995)</td>\n",
       "      <td>Adventure|Animation|Children|Comedy|Fantasy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Jumanji (1995)</td>\n",
       "      <td>Adventure|Children|Fantasy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Grumpier Old Men (1995)</td>\n",
       "      <td>Comedy|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Waiting to Exhale (1995)</td>\n",
       "      <td>Comedy|Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Father of the Bride Part II (1995)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131254</th>\n",
       "      <td>Kein Bund für's Leben (2007)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131256</th>\n",
       "      <td>Feuer, Eis &amp; Dosenbier (2002)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131258</th>\n",
       "      <td>The Pirates (2014)</td>\n",
       "      <td>Adventure</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131260</th>\n",
       "      <td>Rentun Ruusu (2001)</td>\n",
       "      <td>(no genres listed)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131262</th>\n",
       "      <td>Innocence (2014)</td>\n",
       "      <td>Adventure|Fantasy|Horror</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>27278 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      title  \\\n",
       "movieId                                       \n",
       "1                          Toy Story (1995)   \n",
       "2                            Jumanji (1995)   \n",
       "3                   Grumpier Old Men (1995)   \n",
       "4                  Waiting to Exhale (1995)   \n",
       "5        Father of the Bride Part II (1995)   \n",
       "...                                     ...   \n",
       "131254         Kein Bund für's Leben (2007)   \n",
       "131256        Feuer, Eis & Dosenbier (2002)   \n",
       "131258                   The Pirates (2014)   \n",
       "131260                  Rentun Ruusu (2001)   \n",
       "131262                     Innocence (2014)   \n",
       "\n",
       "                                              genres  \n",
       "movieId                                               \n",
       "1        Adventure|Animation|Children|Comedy|Fantasy  \n",
       "2                         Adventure|Children|Fantasy  \n",
       "3                                     Comedy|Romance  \n",
       "4                               Comedy|Drama|Romance  \n",
       "5                                             Comedy  \n",
       "...                                              ...  \n",
       "131254                                        Comedy  \n",
       "131256                                        Comedy  \n",
       "131258                                     Adventure  \n",
       "131260                            (no genres listed)  \n",
       "131262                      Adventure|Fantasy|Horror  \n",
       "\n",
       "[27278 rows x 2 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movie_df = pd.read_csv('movielens-20m-dataset/movie.csv', index_col='movieId')\n",
    "movie_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PyTorch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Main model definition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RSModel(nn.Module):\n",
    "    def __init__(\n",
    "        self,\n",
    "        n_users,\n",
    "        n_items,\n",
    "        embed_dim=20,\n",
    "        n_hidden1=20,\n",
    "        n_hidden2=20,\n",
    "        activation='sigmoid',\n",
    "    ):\n",
    "        super(RSModel, self).__init__()\n",
    "        self.N = n_users\n",
    "        self.M = n_items\n",
    "        self.D = embed_dim\n",
    "        self.n_hidden1 = n_hidden1\n",
    "        self.n_hidden2 = n_hidden2\n",
    "        self.activation = activation\n",
    "\n",
    "        self.emb_users = nn.Embedding(self.N, self.D)\n",
    "        self.emb_items = nn.Embedding(self.M, self.D)\n",
    "\n",
    "        \"\"\"\n",
    "        # set the weights since N(0, 1) leads to poor results\n",
    "        self.emb_users.weight.data = nn.Parameter(\n",
    "            torch.Tensor(np.random.randn(self.N, self.D_users) * 0.01))\n",
    "        self.emb_items.weight.data = nn.Parameter(\n",
    "            torch.Tensor(np.random.randn(self.M, self.D_items) * 0.01))\n",
    "        \"\"\"\n",
    "\n",
    "        self.fc1 = nn.Linear(2 * self.D, self.n_hidden1)\n",
    "        self.fc2 = nn.Linear(self.n_hidden1, self.n_hidden2)\n",
    "        self.fc_out = nn.Linear(self.n_hidden2, 1)\n",
    "\n",
    "    def forward(self, u, b):\n",
    "        u = self.emb_users(u)\n",
    "        b = self.emb_items(b)\n",
    "        out = torch.cat((u, b), 1)\n",
    "\n",
    "        out = self.fc1(out)\n",
    "        if self.activation == 'relu':\n",
    "            out = F.relu(out)\n",
    "        elif self.activation == 'leaky_relu':\n",
    "            out = F.leaky_relu(out)\n",
    "        elif self.activation == 'sigmoid':\n",
    "            out = F.sigmoid(out)\n",
    "        elif self.activation == 'tanh':\n",
    "            out = F.tanh(out)\n",
    "        else:\n",
    "            raise Exception(f'unknown activation {self.activation}')\n",
    "\n",
    "        out = self.fc2(out)\n",
    "        if self.activation == 'relu':\n",
    "            out = F.relu(out)\n",
    "        elif self.activation == 'leaky_relu':\n",
    "            out = F.leaky_relu(out)\n",
    "        elif self.activation == 'sigmoid':\n",
    "            out = F.sigmoid(out)\n",
    "        elif self.activation == 'tanh':\n",
    "            out = F.tanh(out)\n",
    "        else:\n",
    "            raise Exception(f'unknown activation {self.activation}')\n",
    "\n",
    "        out = self.fc_out(out)\n",
    "        return out"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Gradient descent function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def batch_gd_direct(\n",
    "    model,\n",
    "    device,\n",
    "    criterion,\n",
    "    optimizer,\n",
    "    train_tuple,\n",
    "    test_tuple,\n",
    "    epochs=24,\n",
    "    batch_size=512,\n",
    "    progress=True,\n",
    "    quit_after=5,\n",
    "):\n",
    "    # Using a torch dataloader is slower than using tensors directly\n",
    "    # so we use tensors here.\n",
    "    # First, extract the numpy arrays:\n",
    "    user_ids_train, movie_ids_train, ratings_train = train_tuple\n",
    "    user_ids_test, movie_ids_test, ratings_test = test_tuple\n",
    "\n",
    "    losses_train = np.full(epochs, np.inf)\n",
    "    losses_test = np.full(epochs, np.inf)\n",
    "\n",
    "    batches_per_epoch_train = int(np.ceil(len(user_ids_train) / batch_size))\n",
    "    batches_per_epoch_test = int(np.ceil(len(user_ids_test) / batch_size))\n",
    "\n",
    "    for it in range(epochs):\n",
    "        t_start = datetime.now()\n",
    "        loss_train = []\n",
    "\n",
    "        # shuffle train data\n",
    "        user_ids_train, movie_ids_train, ratings_train = shuffle(\n",
    "            user_ids_train, movie_ids_train, ratings_train, random_state=it\n",
    "        )\n",
    "\n",
    "        for j in range(batches_per_epoch_train):\n",
    "            # create batch\n",
    "            users_batch = user_ids_train[j * batch_size : (j + 1) * batch_size]\n",
    "            movies_batch = movie_ids_train[\n",
    "                j * batch_size : (j + 1) * batch_size\n",
    "            ]\n",
    "            ratings_batch = ratings_train[j * batch_size : (j + 1) * batch_size]\n",
    "\n",
    "            # convert to tensor\n",
    "            users_batch = torch.from_numpy(users_batch).long()\n",
    "            movies_batch = torch.from_numpy(movies_batch).long()\n",
    "            ratings_batch = torch.from_numpy(ratings_batch)\n",
    "            # reshape targets\n",
    "            ratings_batch = ratings_batch.view(-1, 1).float()\n",
    "\n",
    "            # move batch to GPU\n",
    "            users_batch, movies_batch, ratings_batch = (\n",
    "                users_batch.to(device),\n",
    "                movies_batch.to(device),\n",
    "                ratings_batch.to(device),\n",
    "            )\n",
    "\n",
    "            # forward pass\n",
    "            optimizer.zero_grad()\n",
    "            outputs = model(users_batch, movies_batch)\n",
    "            loss = criterion(outputs, ratings_batch)\n",
    "\n",
    "            # backward pass\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            loss_train.append(loss.item())\n",
    "\n",
    "        losses_train[it] = np.mean(loss_train)\n",
    "\n",
    "        loss_test = []\n",
    "\n",
    "        for j in range(batches_per_epoch_test):\n",
    "            # create batch\n",
    "            users_batch = user_ids_test[j * batch_size : (j + 1) * batch_size]\n",
    "            movies_batch = movie_ids_test[j * batch_size : (j + 1) * batch_size]\n",
    "            ratings_batch = ratings_test[j * batch_size : (j + 1) * batch_size]\n",
    "\n",
    "            # convert to tensor\n",
    "            users_batch = torch.from_numpy(users_batch).long()\n",
    "            movies_batch = torch.from_numpy(movies_batch).long()\n",
    "            ratings_batch = torch.from_numpy(ratings_batch)\n",
    "            # reshape targets\n",
    "            ratings_batch = ratings_batch.view(-1, 1).float()\n",
    "\n",
    "            # move batch to GPU\n",
    "            users_batch, movies_batch, ratings_batch = (\n",
    "                users_batch.to(device),\n",
    "                movies_batch.to(device),\n",
    "                ratings_batch.to(device),\n",
    "            )\n",
    "\n",
    "            outputs = model(users_batch, movies_batch)\n",
    "            loss = criterion(outputs, ratings_batch)\n",
    "            loss_test.append(loss.item())\n",
    "\n",
    "        losses_test[it] = np.mean(loss_test)\n",
    "\n",
    "        dt = datetime.now() - t_start\n",
    "        if progress:\n",
    "            print(\n",
    "                f'epoch: {(1+it):3n}/{epochs}, train loss: {losses_train[it]:.8f}, test loss: {losses_test[it]:.8f}, duration: {dt}'\n",
    "            )\n",
    "        best_epoch = np.argmin(losses_test)\n",
    "        if best_epoch == it:\n",
    "            torch.save(model.state_dict(), 'torch_recsys.pt')\n",
    "        if it - best_epoch >= quit_after:\n",
    "            break\n",
    "\n",
    "    return losses_train[: it + 1], losses_test[: it + 1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create and evaluate a model, given some parameters. Should probably be merged in the objective function below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_torch(\n",
    "    device,\n",
    "    train_tuple,\n",
    "    test_tuple,\n",
    "    N,\n",
    "    M,\n",
    "    D=20,\n",
    "    H1=20,\n",
    "    H2=20,\n",
    "    activation='sigmoid',\n",
    "    lr_sgd=0.03,\n",
    "    momentum=0.9,\n",
    "    epochs=24,\n",
    "    progress=False,\n",
    "    quit_after=5,\n",
    "):\n",
    "\n",
    "    torch.manual_seed(0)\n",
    "    model = RSModel(N, M, D, H1, H2, activation)\n",
    "    model.to(device)\n",
    "\n",
    "    criterion = nn.MSELoss()\n",
    "    # optimizer = torch.optim.Adam(model.parameters(), lr=0.001)\n",
    "    optimizer = torch.optim.SGD(\n",
    "        model.parameters(), lr=lr_sgd, momentum=momentum\n",
    "    )\n",
    "\n",
    "    batch_size = 512\n",
    "\n",
    "    losses_train, losses_test = batch_gd_direct(\n",
    "        model,\n",
    "        device,\n",
    "        criterion,\n",
    "        optimizer,\n",
    "        train_tuple,\n",
    "        test_tuple,\n",
    "        epochs,\n",
    "        batch_size,\n",
    "        progress,\n",
    "        quit_after,\n",
    "    )\n",
    "    return losses_test.min()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Optuna objective function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def objective_torch(trial, device, train_tuple, test_tuple, N, M):\n",
    "    D = trial.suggest_int('D_embedding', 2, 1000, step=1, log=True)\n",
    "    H1 = trial.suggest_int(f'Hidden1', 2, 10000, step=1, log=True)\n",
    "    H2 = trial.suggest_int(f'Hidden2', 2, 10000, step=1, log=True)\n",
    "    activation = trial.suggest_categorical(\n",
    "        'activation', ['relu', 'leaky_relu', 'sigmoid', 'tanh']\n",
    "    )\n",
    "    lr_sgd = trial.suggest_float('lr_sgd', 0.00001, 0.01, log=True)\n",
    "    momentum = trial.suggest_float('momentum', 0.5, 0.999, log=True)\n",
    "\n",
    "    loss = evaluate_torch(\n",
    "        device,\n",
    "        train_tuple,\n",
    "        test_tuple,\n",
    "        N,\n",
    "        M,\n",
    "        D=D,\n",
    "        H1=H1,\n",
    "        H2=H2,\n",
    "        activation=activation,\n",
    "        lr_sgd=lr_sgd,\n",
    "        momentum=momentum,\n",
    "        epochs=24,\n",
    "        progress=False,\n",
    "        quit_after=5,\n",
    "    )\n",
    "    torch.cuda.empty_cache()\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "device(type='cuda', index=0)"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')\n",
    "display(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[I 2024-02-19 12:42:02,588] Using an existing study with name 'Movie Lens 20m' instead of creating a new one.\n"
     ]
    }
   ],
   "source": [
    "study_torch = optuna.create_study(\n",
    "    storage='sqlite:///optuna_recsys_torch.db',\n",
    "    sampler=optuna.samplers.TPESampler(n_startup_trials=10, seed=0),\n",
    "    study_name='Movie Lens 20m',\n",
    "    direction='minimize',\n",
    "    load_if_exists=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hyperparameter optimization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_pytorch_model = 'torch_recsys_best.pt'\n",
    "\n",
    "if False:\n",
    "    # if optimization is run, assume the best model will become obsolete\n",
    "    #\n",
    "    # TODO: not the best strategy. Model management should be handled in an Optuna hook,\n",
    "    # so training the best model again would not be needed.\n",
    "    try:\n",
    "        os.remove(best_pytorch_model)\n",
    "    except Exception as e:\n",
    "        pass\n",
    "    study_torch.optimize(\n",
    "        lambda trial: objective_torch(\n",
    "            trial,\n",
    "            device,\n",
    "            (user_ids_train, movie_ids_train, ratings_train),\n",
    "            (user_ids_test, movie_ids_test, ratings_test),\n",
    "            N,\n",
    "            M,\n",
    "        ),\n",
    "        n_trials=None,\n",
    "        timeout=9 * 3600,\n",
    "        n_jobs=1,\n",
    "        gc_after_trial=True,\n",
    "        show_progress_bar=True,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'D_embedding': 71,\n",
       " 'Hidden1': 245,\n",
       " 'Hidden2': 259,\n",
       " 'activation': 'leaky_relu',\n",
       " 'lr_sgd': 0.0007657998731223536,\n",
       " 'momentum': 0.997086550671604}"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "study_torch.best_params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.038680934534152514"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "study_torch.best_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = optuna.visualization.plot_optimization_history(study_torch)\n",
    "fig.update_layout(\n",
    "    title='PyTorch Model Optimization History Plot', width=1200, height=400\n",
    ")\n",
    "fig.show('png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = optuna.visualization.plot_slice(study_torch)\n",
    "fig.update_layout(title='PyTorch Model Slice Plot', width=2000, height=400)\n",
    "fig.show('png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Spark"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create Spark context and train/test RDDs. Train model. Evaluate its performance on test data. Return performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_spark(\n",
    "    df_seq_train_fn, df_seq_test_fn, K=10, epochs=10, lambda_=0.01\n",
    "):\n",
    "    # set python executables\n",
    "    os.environ['PYSPARK_PYTHON'] = sys.executable\n",
    "    os.environ['PYSPARK_DRIVER_PYTHON'] = sys.executable\n",
    "\n",
    "    # Memory and multiprocessing are fine-tuned for 64 GB RAM single system\n",
    "    #\n",
    "    # increase memory\n",
    "    SparkContext.setSystemProperty('spark.driver.memory', '40g')\n",
    "    # SparkContext.setSystemProperty('spark.executor.memory', '40g')\n",
    "    SparkContext.setSystemProperty('spark.memory.fraction', '0.8')\n",
    "    # create context\n",
    "    # local[...] controls the number of multiprocessing workers\n",
    "    # local[*] = use all CPUs\n",
    "    sc = SparkContext(\"local[3]\", \"Movie Lens 20m\")\n",
    "\n",
    "    # load and extract data\n",
    "    data_train = sc.textFile(df_seq_train_fn)\n",
    "    header_train = data_train.first()\n",
    "    data_train = data_train.filter(lambda row: row != header_train)\n",
    "\n",
    "    data_test = sc.textFile(df_seq_test_fn)\n",
    "    header_test = data_test.first()\n",
    "    data_test = data_test.filter(lambda row: row != header_test)\n",
    "\n",
    "    ratings_train = data_train.map(lambda l: l.split(',')).map(\n",
    "        lambda l: Rating(int(l[4]), int(l[5]), float(l[2]))\n",
    "    )\n",
    "    ratings_test = data_test.map(lambda l: l.split(',')).map(\n",
    "        lambda l: Rating(int(l[4]), int(l[5]), float(l[2]))\n",
    "    )\n",
    "\n",
    "    # train model\n",
    "    model = ALS.train(\n",
    "        ratings=ratings_train,\n",
    "        rank=K,\n",
    "        iterations=epochs,\n",
    "        lambda_=lambda_,\n",
    "        seed=0,\n",
    "    )\n",
    "\n",
    "    # evaluate model\n",
    "    # extract test features\n",
    "    x = ratings_test.map(lambda p: (p[0], p[1]))\n",
    "    # make predictions\n",
    "    p = model.predictAll(x).map(lambda r: ((r[0], r[1]), r[2]))\n",
    "    # join with ground-truth labels\n",
    "    ratesAndPreds = ratings_test.map(lambda r: ((r[0], r[1]), r[2])).join(p)\n",
    "    # compute MSE\n",
    "    # divide by 5 to keep the same [0, 1] scale like the other models\n",
    "    mse = ratesAndPreds.map(lambda r: (r[1][0] / 5 - r[1][1] / 5) ** 2).mean()\n",
    "    # force evaluation\n",
    "    mse = mse * 1.0\n",
    "\n",
    "    sc.stop()\n",
    "    return mse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Optuna objective function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def objective_spark(trial, df_seq_train_fn, df_seq_test_fn, epochs=10):\n",
    "    rank = trial.suggest_int('rank', 1, 100, step=1, log=True)\n",
    "    lambda_ = trial.suggest_float('lambda', 0.005, 0.5, log=True)\n",
    "\n",
    "    loss = evaluate_spark(\n",
    "        df_seq_train_fn, df_seq_test_fn, rank, epochs, lambda_\n",
    "    )\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hyperparameter optimization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[I 2024-02-19 12:42:04,457] Using an existing study with name 'Movie Lens 20m' instead of creating a new one.\n"
     ]
    }
   ],
   "source": [
    "study_spark = optuna.create_study(\n",
    "    storage='sqlite:///optuna_recsys_spark.db',\n",
    "    sampler=optuna.samplers.TPESampler(n_startup_trials=20, seed=0),\n",
    "    study_name='Movie Lens 20m',\n",
    "    direction='minimize',\n",
    "    load_if_exists=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_spark_model = 'spark_recsys_best'\n",
    "\n",
    "if False:\n",
    "    # if optimization is run, assume the best model will become obsolete\n",
    "    shutil.rmtree(best_spark_model, ignore_errors=True)\n",
    "    study_spark.optimize(\n",
    "        lambda trial: objective_spark(\n",
    "            trial, df_seq_train_fn, df_seq_test_fn, epochs=10\n",
    "        ),\n",
    "        n_trials=None,\n",
    "        timeout=1.5 * 3600,\n",
    "        n_jobs=1,\n",
    "        gc_after_trial=True,\n",
    "        show_progress_bar=True,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'rank': 19, 'lambda': 0.07607513547547576}"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "study_spark.best_params"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.02614118493694747"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "study_spark.best_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = optuna.visualization.plot_optimization_history(study_spark)\n",
    "fig.update_layout(\n",
    "    title='Spark Model Optimization History Plot', width=1200, height=400\n",
    ")\n",
    "fig.show('png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = optuna.visualization.plot_slice(study_spark)\n",
    "fig.update_layout(title='Spark Model Slice Plot', width=720, height=400)\n",
    "fig.show('png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Make recommendations\n",
    "\n",
    "Use the best models, determined above, to make movie recommendations for a given user."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare data\n",
    "\n",
    "Take a user ID, find the movies they have watched.\n",
    "\n",
    "Then find all the movies they have not watched - these will be used to run predictions (make recommendations)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 30 rated movies by user 28507:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movieId</th>\n",
       "      <th>rating</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1632701</th>\n",
       "      <td>720</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Wallace &amp; Gromit: The Best of Aardman Animatio...</td>\n",
       "      <td>Adventure|Animation|Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1939379</th>\n",
       "      <td>1446</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Kolya (Kolja) (1996)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1637326</th>\n",
       "      <td>541</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Blade Runner (1982)</td>\n",
       "      <td>Action|Sci-Fi|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1606103</th>\n",
       "      <td>36</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Dead Man Walking (1995)</td>\n",
       "      <td>Crime|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1606115</th>\n",
       "      <td>745</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Wallace &amp; Gromit: A Close Shave (1995)</td>\n",
       "      <td>Animation|Children|Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1632709</th>\n",
       "      <td>1148</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Wallace &amp; Gromit: The Wrong Trousers (1993)</td>\n",
       "      <td>Animation|Children|Comedy|Crime</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1606163</th>\n",
       "      <td>1366</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Crucible, The (1996)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1632713</th>\n",
       "      <td>307</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Three Colors: Blue (Trois couleurs: Bleu) (1993)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1632696</th>\n",
       "      <td>290</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Once Were Warriors (1994)</td>\n",
       "      <td>Crime|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1632694</th>\n",
       "      <td>527</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Schindler's List (1993)</td>\n",
       "      <td>Drama|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639419</th>\n",
       "      <td>1132</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Manon of the Spring (Manon des sources) (1986)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639574</th>\n",
       "      <td>1172</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Cinema Paradiso (Nuovo cinema Paradiso) (1989)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639426</th>\n",
       "      <td>1269</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Arsenic and Old Lace (1944)</td>\n",
       "      <td>Comedy|Mystery|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639425</th>\n",
       "      <td>1175</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Delicatessen (1991)</td>\n",
       "      <td>Comedy|Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639423</th>\n",
       "      <td>1300</td>\n",
       "      <td>4.0</td>\n",
       "      <td>My Life as a Dog (Mitt liv som hund) (1985)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639420</th>\n",
       "      <td>1197</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Princess Bride, The (1987)</td>\n",
       "      <td>Action|Adventure|Comedy|Fantasy|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639413</th>\n",
       "      <td>21</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Get Shorty (1995)</td>\n",
       "      <td>Comedy|Crime|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639416</th>\n",
       "      <td>1185</td>\n",
       "      <td>4.0</td>\n",
       "      <td>My Left Foot (1989)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639414</th>\n",
       "      <td>1198</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Raiders of the Lost Ark (Indiana Jones and the...</td>\n",
       "      <td>Action|Adventure</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639587</th>\n",
       "      <td>1299</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Killing Fields, The (1984)</td>\n",
       "      <td>Drama|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639411</th>\n",
       "      <td>919</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Wizard of Oz, The (1939)</td>\n",
       "      <td>Adventure|Children|Fantasy|Musical</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639410</th>\n",
       "      <td>1096</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Sophie's Choice (1982)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639409</th>\n",
       "      <td>1131</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Jean de Florette (1986)</td>\n",
       "      <td>Drama|Mystery</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639408</th>\n",
       "      <td>1136</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Monty Python and the Holy Grail (1975)</td>\n",
       "      <td>Adventure|Comedy|Fantasy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639407</th>\n",
       "      <td>969</td>\n",
       "      <td>4.0</td>\n",
       "      <td>African Queen, The (1951)</td>\n",
       "      <td>Adventure|Comedy|Romance|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639405</th>\n",
       "      <td>534</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Shadowlands (1993)</td>\n",
       "      <td>Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1637331</th>\n",
       "      <td>986</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Fly Away Home (1996)</td>\n",
       "      <td>Adventure|Children</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1637329</th>\n",
       "      <td>302</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Queen Margot (Reine Margot, La) (1994)</td>\n",
       "      <td>Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1639583</th>\n",
       "      <td>1242</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Glory (1989)</td>\n",
       "      <td>Drama|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1176</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Double Life of Veronique, The (Double Vie de V...</td>\n",
       "      <td>Drama|Fantasy|Romance</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         movieId  rating                                              title  \\\n",
       "1632701      720     5.0  Wallace & Gromit: The Best of Aardman Animatio...   \n",
       "1939379     1446     5.0                               Kolya (Kolja) (1996)   \n",
       "1637326      541     5.0                                Blade Runner (1982)   \n",
       "1606103       36     5.0                            Dead Man Walking (1995)   \n",
       "1606115      745     5.0             Wallace & Gromit: A Close Shave (1995)   \n",
       "1632709     1148     5.0        Wallace & Gromit: The Wrong Trousers (1993)   \n",
       "1606163     1366     5.0                               Crucible, The (1996)   \n",
       "1632713      307     5.0   Three Colors: Blue (Trois couleurs: Bleu) (1993)   \n",
       "1632696      290     5.0                          Once Were Warriors (1994)   \n",
       "1632694      527     5.0                            Schindler's List (1993)   \n",
       "1639419     1132     4.0     Manon of the Spring (Manon des sources) (1986)   \n",
       "1639574     1172     4.0     Cinema Paradiso (Nuovo cinema Paradiso) (1989)   \n",
       "1639426     1269     4.0                        Arsenic and Old Lace (1944)   \n",
       "1639425     1175     4.0                                Delicatessen (1991)   \n",
       "1639423     1300     4.0        My Life as a Dog (Mitt liv som hund) (1985)   \n",
       "1639420     1197     4.0                         Princess Bride, The (1987)   \n",
       "1639413       21     4.0                                  Get Shorty (1995)   \n",
       "1639416     1185     4.0                                My Left Foot (1989)   \n",
       "1639414     1198     4.0  Raiders of the Lost Ark (Indiana Jones and the...   \n",
       "1639587     1299     4.0                         Killing Fields, The (1984)   \n",
       "1639411      919     4.0                           Wizard of Oz, The (1939)   \n",
       "1639410     1096     4.0                             Sophie's Choice (1982)   \n",
       "1639409     1131     4.0                            Jean de Florette (1986)   \n",
       "1639408     1136     4.0             Monty Python and the Holy Grail (1975)   \n",
       "1639407      969     4.0                          African Queen, The (1951)   \n",
       "1639405      534     4.0                                 Shadowlands (1993)   \n",
       "1637331      986     4.0                               Fly Away Home (1996)   \n",
       "1637329      302     4.0             Queen Margot (Reine Margot, La) (1994)   \n",
       "1639583     1242     4.0                                       Glory (1989)   \n",
       "0           1176     4.0  Double Life of Veronique, The (Double Vie de V...   \n",
       "\n",
       "                                          genres  \n",
       "1632701               Adventure|Animation|Comedy  \n",
       "1939379                             Comedy|Drama  \n",
       "1637326                   Action|Sci-Fi|Thriller  \n",
       "1606103                              Crime|Drama  \n",
       "1606115                Animation|Children|Comedy  \n",
       "1632709          Animation|Children|Comedy|Crime  \n",
       "1606163                                    Drama  \n",
       "1632713                                    Drama  \n",
       "1632696                              Crime|Drama  \n",
       "1632694                                Drama|War  \n",
       "1639419                                    Drama  \n",
       "1639574                                    Drama  \n",
       "1639426                  Comedy|Mystery|Thriller  \n",
       "1639425                     Comedy|Drama|Romance  \n",
       "1639423                             Comedy|Drama  \n",
       "1639420  Action|Adventure|Comedy|Fantasy|Romance  \n",
       "1639413                    Comedy|Crime|Thriller  \n",
       "1639416                                    Drama  \n",
       "1639414                         Action|Adventure  \n",
       "1639587                                Drama|War  \n",
       "1639411       Adventure|Children|Fantasy|Musical  \n",
       "1639410                                    Drama  \n",
       "1639409                            Drama|Mystery  \n",
       "1639408                 Adventure|Comedy|Fantasy  \n",
       "1639407             Adventure|Comedy|Romance|War  \n",
       "1639405                            Drama|Romance  \n",
       "1637331                       Adventure|Children  \n",
       "1637329                            Drama|Romance  \n",
       "1639583                                Drama|War  \n",
       "0                          Drama|Fantasy|Romance  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_entry = 0\n",
    "userId = df['userId'].to_list()[n_entry]\n",
    "userIdseq = df['userIdseq'].to_list()[n_entry]\n",
    "movies_watched = df[df['userId'] == userId]['movieIdseq'].unique()\n",
    "\n",
    "movies_unwatched = df[~df['movieIdseq'].isin(movies_watched)][\n",
    "    'movieIdseq'\n",
    "].unique()\n",
    "user_ids_np = np.full(movies_unwatched.shape[0], userIdseq)\n",
    "\n",
    "n_rec = 30\n",
    "print(f'Top {n_rec} rated movies by user {userId}:')\n",
    "top_rated = (\n",
    "    df[df['userId'] == userId][['movieId', 'rating']]\n",
    "    .sort_values(by='rating', ascending=False)\n",
    "    .join(movie_df, on='movieId')\n",
    "    .head(n_rec)\n",
    ")\n",
    "top_rated"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PyTorch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Either load the best model if it exists, or train one from scratch using the best hyperparameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found best trained model torch_recsys_best.pt, loading it now.\n"
     ]
    }
   ],
   "source": [
    "torch.manual_seed(0)\n",
    "model_torch = RSModel(\n",
    "    N,\n",
    "    M,\n",
    "    study_torch.best_params['D_embedding'],\n",
    "    study_torch.best_params['Hidden1'],\n",
    "    study_torch.best_params['Hidden2'],\n",
    "    study_torch.best_params['activation'],\n",
    ")\n",
    "\n",
    "if not os.path.isfile(best_pytorch_model):\n",
    "    print('Best model not found, need to train it first.')\n",
    "    model_torch.to(device)\n",
    "    criterion = nn.MSELoss()\n",
    "    optimizer = torch.optim.SGD(\n",
    "        model_torch.parameters(),\n",
    "        lr=study_torch.best_params['lr_sgd'],\n",
    "        momentum=study_torch.best_params['momentum'],\n",
    "    )\n",
    "    _ = batch_gd_direct(\n",
    "        model_torch,\n",
    "        device,\n",
    "        criterion,\n",
    "        optimizer,\n",
    "        (user_ids_train, movie_ids_train, ratings_train),\n",
    "        (user_ids_test, movie_ids_test, ratings_test),\n",
    "    )\n",
    "    shutil.copy('torch_recsys.pt', best_pytorch_model)\n",
    "else:\n",
    "    print(f'Found best trained model {best_pytorch_model}, loading it now.')\n",
    "    model_torch.load_state_dict(torch.load(best_pytorch_model))\n",
    "    model_torch.to(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run predictions on movies not watched by the user."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 30 recommendations for userId 28507:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movieId</th>\n",
       "      <th>rating_pred</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>103380</td>\n",
       "      <td>4.835875</td>\n",
       "      <td>A Coming-Out Party (1961)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>90084</td>\n",
       "      <td>4.598035</td>\n",
       "      <td>Experience, The (Tadjrebeh) (1973)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>124077</td>\n",
       "      <td>4.583772</td>\n",
       "      <td>The Trip to Bountiful (2014)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>126921</td>\n",
       "      <td>4.531214</td>\n",
       "      <td>The Fox and the Hound 2 (2006)</td>\n",
       "      <td>Adventure|Animation|Children|Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>90843</td>\n",
       "      <td>4.530509</td>\n",
       "      <td>Lavatory Lovestory (Ubornaya istoriya - lyubov...</td>\n",
       "      <td>Animation</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>113306</td>\n",
       "      <td>4.520823</td>\n",
       "      <td>Dragon Missile, The (Fei long zhan) (1976)</td>\n",
       "      <td>Action</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>99748</td>\n",
       "      <td>4.483615</td>\n",
       "      <td>I Got Next (2011)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>77816</td>\n",
       "      <td>4.462348</td>\n",
       "      <td>Lord, Save Us from Your Followers (2008)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>101801</td>\n",
       "      <td>4.457100</td>\n",
       "      <td>Skylab, Le (2011)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>111789</td>\n",
       "      <td>4.452935</td>\n",
       "      <td>Gulliver's Travels (1977)</td>\n",
       "      <td>Adventure|Animation</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>116859</td>\n",
       "      <td>4.444232</td>\n",
       "      <td>Remember Sunday (2013)</td>\n",
       "      <td>Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>858</td>\n",
       "      <td>4.442510</td>\n",
       "      <td>Godfather, The (1972)</td>\n",
       "      <td>Crime|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>47054</td>\n",
       "      <td>4.416222</td>\n",
       "      <td>Each Dawn I Die (1939)</td>\n",
       "      <td>Crime|Drama|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>79565</td>\n",
       "      <td>4.385439</td>\n",
       "      <td>I Remember (Eu Me Lembro) (2005)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>89052</td>\n",
       "      <td>4.381869</td>\n",
       "      <td>Cool School, The (2008)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>108120</td>\n",
       "      <td>4.381587</td>\n",
       "      <td>Turtle's Tale 2: Sammy's Escape from Paradise,...</td>\n",
       "      <td>Adventure|Animation|Children</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>104854</td>\n",
       "      <td>4.377420</td>\n",
       "      <td>Coward, The (Kapurush) (1965)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>129769</td>\n",
       "      <td>4.375969</td>\n",
       "      <td>Covert Action (1978)</td>\n",
       "      <td>(no genres listed)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>117924</td>\n",
       "      <td>4.374772</td>\n",
       "      <td>The Bloody Olive (1997)</td>\n",
       "      <td>Comedy|Crime|Film-Noir</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>101754</td>\n",
       "      <td>4.373071</td>\n",
       "      <td>Chasing Beauty (2013)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>88329</td>\n",
       "      <td>4.368734</td>\n",
       "      <td>Madeleine (1950)</td>\n",
       "      <td>Crime|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>127390</td>\n",
       "      <td>4.364296</td>\n",
       "      <td>Family Guy Presents: Blue Harvest (2007)</td>\n",
       "      <td>Animation|Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>8624</td>\n",
       "      <td>4.360486</td>\n",
       "      <td>Freedom Downtime (2001)</td>\n",
       "      <td>Adventure|Crime|Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>89056</td>\n",
       "      <td>4.351525</td>\n",
       "      <td>Company: Original Cast Album (1970)</td>\n",
       "      <td>Documentary|Musical</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>110330</td>\n",
       "      <td>4.343377</td>\n",
       "      <td>Me and you (io e te) (2012)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>53741</td>\n",
       "      <td>4.323250</td>\n",
       "      <td>Comedy of Power (Ivresse du pouvoir, L') (2006)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>50</td>\n",
       "      <td>4.322368</td>\n",
       "      <td>Usual Suspects, The (1995)</td>\n",
       "      <td>Crime|Mystery|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>87220</td>\n",
       "      <td>4.319832</td>\n",
       "      <td>Violent Saturday (1955)</td>\n",
       "      <td>Crime|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>84300</td>\n",
       "      <td>4.319046</td>\n",
       "      <td>Whisperers, The (1967)</td>\n",
       "      <td>Drama|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>70792</td>\n",
       "      <td>4.317989</td>\n",
       "      <td>Vengeance Valley (1951)</td>\n",
       "      <td>Western</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    movieId  rating_pred                                              title  \\\n",
       "0    103380     4.835875                          A Coming-Out Party (1961)   \n",
       "1     90084     4.598035                 Experience, The (Tadjrebeh) (1973)   \n",
       "2    124077     4.583772                       The Trip to Bountiful (2014)   \n",
       "3    126921     4.531214                     The Fox and the Hound 2 (2006)   \n",
       "4     90843     4.530509  Lavatory Lovestory (Ubornaya istoriya - lyubov...   \n",
       "5    113306     4.520823         Dragon Missile, The (Fei long zhan) (1976)   \n",
       "6     99748     4.483615                                  I Got Next (2011)   \n",
       "7     77816     4.462348           Lord, Save Us from Your Followers (2008)   \n",
       "8    101801     4.457100                                  Skylab, Le (2011)   \n",
       "9    111789     4.452935                          Gulliver's Travels (1977)   \n",
       "10   116859     4.444232                             Remember Sunday (2013)   \n",
       "11      858     4.442510                              Godfather, The (1972)   \n",
       "12    47054     4.416222                             Each Dawn I Die (1939)   \n",
       "13    79565     4.385439                   I Remember (Eu Me Lembro) (2005)   \n",
       "14    89052     4.381869                            Cool School, The (2008)   \n",
       "15   108120     4.381587  Turtle's Tale 2: Sammy's Escape from Paradise,...   \n",
       "16   104854     4.377420                      Coward, The (Kapurush) (1965)   \n",
       "17   129769     4.375969                               Covert Action (1978)   \n",
       "18   117924     4.374772                            The Bloody Olive (1997)   \n",
       "19   101754     4.373071                              Chasing Beauty (2013)   \n",
       "20    88329     4.368734                                   Madeleine (1950)   \n",
       "21   127390     4.364296           Family Guy Presents: Blue Harvest (2007)   \n",
       "22     8624     4.360486                            Freedom Downtime (2001)   \n",
       "23    89056     4.351525                Company: Original Cast Album (1970)   \n",
       "24   110330     4.343377                        Me and you (io e te) (2012)   \n",
       "25    53741     4.323250    Comedy of Power (Ivresse du pouvoir, L') (2006)   \n",
       "26       50     4.322368                         Usual Suspects, The (1995)   \n",
       "27    87220     4.319832                            Violent Saturday (1955)   \n",
       "28    84300     4.319046                             Whisperers, The (1967)   \n",
       "29    70792     4.317989                            Vengeance Valley (1951)   \n",
       "\n",
       "                                 genres  \n",
       "0                          Comedy|Drama  \n",
       "1                                 Drama  \n",
       "2                                 Drama  \n",
       "3   Adventure|Animation|Children|Comedy  \n",
       "4                             Animation  \n",
       "5                                Action  \n",
       "6                           Documentary  \n",
       "7                           Documentary  \n",
       "8                          Comedy|Drama  \n",
       "9                   Adventure|Animation  \n",
       "10                        Drama|Romance  \n",
       "11                          Crime|Drama  \n",
       "12                 Crime|Drama|Thriller  \n",
       "13                                Drama  \n",
       "14                          Documentary  \n",
       "15         Adventure|Animation|Children  \n",
       "16                                Drama  \n",
       "17                   (no genres listed)  \n",
       "18               Comedy|Crime|Film-Noir  \n",
       "19                          Documentary  \n",
       "20                          Crime|Drama  \n",
       "21                     Animation|Comedy  \n",
       "22          Adventure|Crime|Documentary  \n",
       "23                  Documentary|Musical  \n",
       "24                                Drama  \n",
       "25                                Drama  \n",
       "26               Crime|Mystery|Thriller  \n",
       "27                          Crime|Drama  \n",
       "28                       Drama|Thriller  \n",
       "29                              Western  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_ids_batch = torch.from_numpy(user_ids_np).long().to(device)\n",
    "movie_ids_batch = torch.from_numpy(movies_unwatched).long().to(device)\n",
    "\n",
    "with torch.no_grad():\n",
    "    pred_t = model_torch(user_ids_batch, movie_ids_batch)\n",
    "predictions = pred_t.cpu().numpy().flatten() * df['rating'].max()\n",
    "\n",
    "sort_idx = np.argsort(-predictions)\n",
    "recommendations_torch = pd.DataFrame(\n",
    "    {\n",
    "        'movieId': [movie_idseq2id[id] for id in movies_unwatched[sort_idx]],\n",
    "        'rating_pred': predictions[sort_idx].tolist(),\n",
    "    }\n",
    ")\n",
    "recommendations_torch = recommendations_torch.join(movie_df, on='movieId')\n",
    "print(f'Top {n_rec} recommendations for userId {userId}:')\n",
    "recommendations_torch.head(n_rec)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "del model_torch\n",
    "torch.cuda.empty_cache()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spark"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Either load the best model if it exists, or train one from scratch using the best hyperparameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/lib/python3.11/subprocess.py:1883: RuntimeWarning:\n",
      "\n",
      "os.fork() was called. os.fork() is incompatible with multithreaded code, and JAX is multithreaded, so this will likely lead to a deadlock.\n",
      "\n",
      "24/02/19 12:42:08 WARN Utils: Your hostname, media resolves to a loopback address: 127.0.1.1; using 192.168.1.20 instead (on interface enp39s0)\n",
      "24/02/19 12:42:08 WARN Utils: Set SPARK_LOCAL_IP if you need to bind to another address\n",
      "Setting default log level to \"WARN\".\n",
      "To adjust logging level use sc.setLogLevel(newLevel). For SparkR, use setLogLevel(newLevel).\n",
      "24/02/19 12:42:08 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found best trained model spark_recsys_best, loading it now.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "24/02/19 12:42:11 WARN MatrixFactorizationModel: User factor does not have a partitioner. Prediction on individual records could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModel: User factor is not cached. Prediction could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModel: Product factor does not have a partitioner. Prediction on individual records could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModel: Product factor is not cached. Prediction could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModelWrapper: User factor does not have a partitioner. Prediction on individual records could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModelWrapper: User factor is not cached. Prediction could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModelWrapper: Product factor does not have a partitioner. Prediction on individual records could be slow.\n",
      "24/02/19 12:42:11 WARN MatrixFactorizationModelWrapper: Product factor is not cached. Prediction could be slow.\n"
     ]
    }
   ],
   "source": [
    "os.environ['PYSPARK_PYTHON'] = sys.executable\n",
    "os.environ['PYSPARK_DRIVER_PYTHON'] = sys.executable\n",
    "sc = SparkContext(\"local[3]\", \"Movie Lens 20m\")\n",
    "\n",
    "if not os.path.isdir(best_spark_model):\n",
    "    print('Best model not found, need to train it first.')\n",
    "    data_train = sc.textFile(df_seq_train_fn)\n",
    "    header_train = data_train.first()\n",
    "    data_train = data_train.filter(lambda row: row != header_train)\n",
    "    ratings_train = data_train.map(lambda l: l.split(',')).map(\n",
    "        lambda l: Rating(int(l[4]), int(l[5]), float(l[2]))\n",
    "    )\n",
    "\n",
    "    model_spark = ALS.train(\n",
    "        ratings=ratings_train,\n",
    "        rank=study_spark.best_params['rank'],\n",
    "        iterations=10,\n",
    "        lambda_=study_spark.best_params['lambda'],\n",
    "        seed=0,\n",
    "    )\n",
    "    model_spark.save(sc, best_spark_model)\n",
    "else:\n",
    "    print(f'Found best trained model {best_spark_model}, loading it now.')\n",
    "    model_spark = MatrixFactorizationModel.load(sc, best_spark_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run predictions on the movies not watched by the user."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 30 recommendations for userId 28507:\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movieId</th>\n",
       "      <th>rating_pred</th>\n",
       "      <th>title</th>\n",
       "      <th>genres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>13079</th>\n",
       "      <td>98595</td>\n",
       "      <td>5.042134</td>\n",
       "      <td>Peppermint Soda (Diabolo menthe) (1977)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8737</th>\n",
       "      <td>81117</td>\n",
       "      <td>4.913878</td>\n",
       "      <td>Moth, The (Cma) (1980)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17473</th>\n",
       "      <td>106048</td>\n",
       "      <td>4.887680</td>\n",
       "      <td>Four Days in July (1985)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9015</th>\n",
       "      <td>99675</td>\n",
       "      <td>4.746409</td>\n",
       "      <td>Eat Sleep Die (Äta sova dö) (2012)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12016</th>\n",
       "      <td>86237</td>\n",
       "      <td>4.648244</td>\n",
       "      <td>Connections (1978)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12487</th>\n",
       "      <td>81072</td>\n",
       "      <td>4.625854</td>\n",
       "      <td>Konopielka (1982)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3096</th>\n",
       "      <td>5814</td>\n",
       "      <td>4.582462</td>\n",
       "      <td>Rising Place, The (2002)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4574</th>\n",
       "      <td>107408</td>\n",
       "      <td>4.565295</td>\n",
       "      <td>Only Old Men Are Going to Battle (V boy idut o...</td>\n",
       "      <td>Comedy|Drama|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6810</th>\n",
       "      <td>72235</td>\n",
       "      <td>4.494393</td>\n",
       "      <td>Between the Devil and the Deep Blue Sea (1995)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>73365</td>\n",
       "      <td>4.468267</td>\n",
       "      <td>Dead Time: Kala (2007)</td>\n",
       "      <td>Crime|Fantasy|Film-Noir|Thriller</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17099</th>\n",
       "      <td>58530</td>\n",
       "      <td>4.461205</td>\n",
       "      <td>On the Outs (2004)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19658</th>\n",
       "      <td>98328</td>\n",
       "      <td>4.460504</td>\n",
       "      <td>Chronicle of My Mother (Waga haha no ki) (2011)</td>\n",
       "      <td>Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21810</th>\n",
       "      <td>102107</td>\n",
       "      <td>4.444936</td>\n",
       "      <td>American Winter (2013)</td>\n",
       "      <td>Documentary|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6683</th>\n",
       "      <td>107623</td>\n",
       "      <td>4.444936</td>\n",
       "      <td>2013 Rock and Roll Hall of Fame Induction Cere...</td>\n",
       "      <td>Documentary|Musical</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11903</th>\n",
       "      <td>104103</td>\n",
       "      <td>4.444936</td>\n",
       "      <td>Miss You Can Do It (2013)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2834</th>\n",
       "      <td>105128</td>\n",
       "      <td>4.444936</td>\n",
       "      <td>Casting By (2012)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4938</th>\n",
       "      <td>107382</td>\n",
       "      <td>4.444936</td>\n",
       "      <td>Whoopi Goldberg Presents Moms Mabley (2013)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17125</th>\n",
       "      <td>101862</td>\n",
       "      <td>4.437551</td>\n",
       "      <td>50 Children: The Rescue Mission of Mr. And Mrs...</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21013</th>\n",
       "      <td>90286</td>\n",
       "      <td>4.422490</td>\n",
       "      <td>I Am My Films - A Portrait of Werner Herzog (W...</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5932</th>\n",
       "      <td>101188</td>\n",
       "      <td>4.403591</td>\n",
       "      <td>Central Park (1991)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6721</th>\n",
       "      <td>106158</td>\n",
       "      <td>4.403382</td>\n",
       "      <td>La discrète (1990)</td>\n",
       "      <td>Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20016</th>\n",
       "      <td>98173</td>\n",
       "      <td>4.398966</td>\n",
       "      <td>Caine Mutiny Court-Martial, The (1988)</td>\n",
       "      <td>Drama|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10665</th>\n",
       "      <td>62206</td>\n",
       "      <td>4.398791</td>\n",
       "      <td>Supermarket Woman (Sûpâ no onna) (1996)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4412</th>\n",
       "      <td>101717</td>\n",
       "      <td>4.388157</td>\n",
       "      <td>Elusive Summer of '68, The (Varljivo leto '68)...</td>\n",
       "      <td>Comedy|Drama|Romance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8711</th>\n",
       "      <td>72035</td>\n",
       "      <td>4.379599</td>\n",
       "      <td>Dimensions of Dialogue (Moznosti dialogu) (1982)</td>\n",
       "      <td>Animation|Comedy|Fantasy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7658</th>\n",
       "      <td>105130</td>\n",
       "      <td>4.359897</td>\n",
       "      <td>Notebook, The (A nagy füzet) (2013)</td>\n",
       "      <td>Drama|War</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2675</th>\n",
       "      <td>31154</td>\n",
       "      <td>4.359214</td>\n",
       "      <td>Heartlands (2002)</td>\n",
       "      <td>Comedy|Drama</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13751</th>\n",
       "      <td>102602</td>\n",
       "      <td>4.356692</td>\n",
       "      <td>Mimino (1977)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6971</th>\n",
       "      <td>88570</td>\n",
       "      <td>4.353338</td>\n",
       "      <td>Welfare (1975)</td>\n",
       "      <td>Documentary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18707</th>\n",
       "      <td>94280</td>\n",
       "      <td>4.333029</td>\n",
       "      <td>Spud (2010)</td>\n",
       "      <td>Comedy</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       movieId  rating_pred  \\\n",
       "13079    98595     5.042134   \n",
       "8737     81117     4.913878   \n",
       "17473   106048     4.887680   \n",
       "9015     99675     4.746409   \n",
       "12016    86237     4.648244   \n",
       "12487    81072     4.625854   \n",
       "3096      5814     4.582462   \n",
       "4574    107408     4.565295   \n",
       "6810     72235     4.494393   \n",
       "237      73365     4.468267   \n",
       "17099    58530     4.461205   \n",
       "19658    98328     4.460504   \n",
       "21810   102107     4.444936   \n",
       "6683    107623     4.444936   \n",
       "11903   104103     4.444936   \n",
       "2834    105128     4.444936   \n",
       "4938    107382     4.444936   \n",
       "17125   101862     4.437551   \n",
       "21013    90286     4.422490   \n",
       "5932    101188     4.403591   \n",
       "6721    106158     4.403382   \n",
       "20016    98173     4.398966   \n",
       "10665    62206     4.398791   \n",
       "4412    101717     4.388157   \n",
       "8711     72035     4.379599   \n",
       "7658    105130     4.359897   \n",
       "2675     31154     4.359214   \n",
       "13751   102602     4.356692   \n",
       "6971     88570     4.353338   \n",
       "18707    94280     4.333029   \n",
       "\n",
       "                                                   title  \\\n",
       "13079            Peppermint Soda (Diabolo menthe) (1977)   \n",
       "8737                              Moth, The (Cma) (1980)   \n",
       "17473                           Four Days in July (1985)   \n",
       "9015                  Eat Sleep Die (Äta sova dö) (2012)   \n",
       "12016                                 Connections (1978)   \n",
       "12487                                  Konopielka (1982)   \n",
       "3096                            Rising Place, The (2002)   \n",
       "4574   Only Old Men Are Going to Battle (V boy idut o...   \n",
       "6810      Between the Devil and the Deep Blue Sea (1995)   \n",
       "237                               Dead Time: Kala (2007)   \n",
       "17099                                 On the Outs (2004)   \n",
       "19658    Chronicle of My Mother (Waga haha no ki) (2011)   \n",
       "21810                             American Winter (2013)   \n",
       "6683   2013 Rock and Roll Hall of Fame Induction Cere...   \n",
       "11903                          Miss You Can Do It (2013)   \n",
       "2834                                   Casting By (2012)   \n",
       "4938         Whoopi Goldberg Presents Moms Mabley (2013)   \n",
       "17125  50 Children: The Rescue Mission of Mr. And Mrs...   \n",
       "21013  I Am My Films - A Portrait of Werner Herzog (W...   \n",
       "5932                                 Central Park (1991)   \n",
       "6721                                  La discrète (1990)   \n",
       "20016             Caine Mutiny Court-Martial, The (1988)   \n",
       "10665            Supermarket Woman (Sûpâ no onna) (1996)   \n",
       "4412   Elusive Summer of '68, The (Varljivo leto '68)...   \n",
       "8711    Dimensions of Dialogue (Moznosti dialogu) (1982)   \n",
       "7658                 Notebook, The (A nagy füzet) (2013)   \n",
       "2675                                   Heartlands (2002)   \n",
       "13751                                      Mimino (1977)   \n",
       "6971                                      Welfare (1975)   \n",
       "18707                                        Spud (2010)   \n",
       "\n",
       "                                 genres  \n",
       "13079                      Comedy|Drama  \n",
       "8737                              Drama  \n",
       "17473                      Comedy|Drama  \n",
       "9015                              Drama  \n",
       "12016                       Documentary  \n",
       "12487                            Comedy  \n",
       "3096                              Drama  \n",
       "4574                   Comedy|Drama|War  \n",
       "6810                              Drama  \n",
       "237    Crime|Fantasy|Film-Noir|Thriller  \n",
       "17099                             Drama  \n",
       "19658                             Drama  \n",
       "21810                 Documentary|Drama  \n",
       "6683                Documentary|Musical  \n",
       "11903                       Documentary  \n",
       "2834                        Documentary  \n",
       "4938                        Documentary  \n",
       "17125                       Documentary  \n",
       "21013                       Documentary  \n",
       "5932                        Documentary  \n",
       "6721                      Drama|Romance  \n",
       "20016                         Drama|War  \n",
       "10665                            Comedy  \n",
       "4412               Comedy|Drama|Romance  \n",
       "8711           Animation|Comedy|Fantasy  \n",
       "7658                          Drama|War  \n",
       "2675                       Comedy|Drama  \n",
       "13751                            Comedy  \n",
       "6971                        Documentary  \n",
       "18707                            Comedy  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_movie_batch = sc.parallelize(\n",
    "    zip(user_ids_np.tolist(), movies_unwatched.tolist())\n",
    ")\n",
    "out = model_spark.predictAll(user_movie_batch)\n",
    "df_spark_pred = pd.DataFrame(\n",
    "    out.map(lambda x: [x[1], x[2]]).collect(),\n",
    "    columns=['movieId', 'rating_pred'],\n",
    ")\n",
    "df_spark_pred['movieId'] = df_spark_pred['movieId'].apply(\n",
    "    lambda x: movie_idseq2id[x]\n",
    ")\n",
    "df_spark_pred.sort_values(by='rating_pred', ascending=False, inplace=True)\n",
    "recommendations_spark = df_spark_pred.join(movie_df, on='movieId')\n",
    "\n",
    "print(f'Top {n_rec} recommendations for userId {userId}:')\n",
    "recommendations_spark.head(n_rec)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "del model_spark\n",
    "sc.stop()\n",
    "del sc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Partial overlap between PyTorch and Spark in terms of genres, but no overlap in terms of actual recommended movies.\n",
    "\n",
    "Spark has much lower loss in testing, so its recommendations are likely better."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}