MovieLens/Small-Data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Matrix Factorisation - Intuition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "ratings = pd.read_csv(\"data/ratings.csv\")\n",
    "items = pd.read_csv(\"data/items.csv\")\n",
    "users = pd.read_csv(\"data/users.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>user_id</th>\n",
       "      <th>movie_id</th>\n",
       "      <th>rating</th>\n",
       "      <th>unix_timestamp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>196</td>\n",
       "      <td>242</td>\n",
       "      <td>3</td>\n",
       "      <td>881250949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>186</td>\n",
       "      <td>302</td>\n",
       "      <td>3</td>\n",
       "      <td>891717742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>22</td>\n",
       "      <td>377</td>\n",
       "      <td>1</td>\n",
       "      <td>878887116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>244</td>\n",
       "      <td>51</td>\n",
       "      <td>2</td>\n",
       "      <td>880606923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>166</td>\n",
       "      <td>346</td>\n",
       "      <td>1</td>\n",
       "      <td>886397596</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  movie_id  rating  unix_timestamp\n",
       "0      196       242       3       881250949\n",
       "1      186       302       3       891717742\n",
       "2       22       377       1       878887116\n",
       "3      244        51       2       880606923\n",
       "4      166       346       1       886397596"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Small Data\n",
    "\n",
    "| movie_id | title                 |\n",
    "| -------: | :-------------------- |\n",
    "|      1   | Toy Story (1995)      |\n",
    "|     71   | Lion King, The (1994) |\n",
    "|     95   | Aladdin (1992)        |\n",
    "|     50   | Star Wars (1972)      |\n",
    "|    176   | Aliens (1986)         |\n",
    "|     82   | Jurassic Park (1993)  |\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.datasets import SampleData"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_users, sample_items, sample_ratings = SampleData(users, items, ratings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.preprocessing import EncodeUserItem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of users:  10\n",
      "Number of items:  6\n"
     ]
    }
   ],
   "source": [
    "interaction, n_users, n_items, user_encoder, item_encoder = EncodeUserItem(sample_ratings,\n",
    "                                                         \"user_id\", \"movie_id\", \"rating\", \"unix_timestamp\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>user_id</th>\n",
       "      <th>movie_id</th>\n",
       "      <th>RATING</th>\n",
       "      <th>TIMESTAMP</th>\n",
       "      <th>USER</th>\n",
       "      <th>ITEM</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>1052</td>\n",
       "      <td>2</td>\n",
       "      <td>50</td>\n",
       "      <td>5</td>\n",
       "      <td>888552084</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1090</td>\n",
       "      <td>8</td>\n",
       "      <td>50</td>\n",
       "      <td>5</td>\n",
       "      <td>879362124</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3672</td>\n",
       "      <td>6</td>\n",
       "      <td>95</td>\n",
       "      <td>2</td>\n",
       "      <td>883602133</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4280</td>\n",
       "      <td>1</td>\n",
       "      <td>82</td>\n",
       "      <td>5</td>\n",
       "      <td>878542589</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4596</td>\n",
       "      <td>12</td>\n",
       "      <td>82</td>\n",
       "      <td>4</td>\n",
       "      <td>879959610</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      user_id  movie_id  RATING  TIMESTAMP  USER  ITEM\n",
       "1052        2        50       5  888552084     1     1\n",
       "1090        8        50       5  879362124     4     1\n",
       "3672        6        95       2  883602133     2     4\n",
       "4280        1        82       5  878542589     0     3\n",
       "4596       12        82       4  879959610     6     3"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interaction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 5)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max_rating = interaction.RATING.max()\n",
    "min_rating = interaction.RATING.min()\n",
    "min_rating, max_rating"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.vis import InteractionVis, TrainTestVis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "axis": {
         "grid": false
        },
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-9794fa2be47445e32b4fd83ae56e13cb"
       },
       "datasets": {
        "data-9794fa2be47445e32b4fd83ae56e13cb": [
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 888552084,
          "USER": 1,
          "movie_id": 50,
          "user_id": 2
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 879362124,
          "USER": 4,
          "movie_id": 50,
          "user_id": 8
         },
         {
          "ITEM": 4,
          "RATING": 2,
          "TIMESTAMP": 883602133,
          "USER": 2,
          "movie_id": 95,
          "user_id": 6
         },
         {
          "ITEM": 3,
          "RATING": 5,
          "TIMESTAMP": 878542589,
          "USER": 0,
          "movie_id": 82,
          "user_id": 1
         },
         {
          "ITEM": 3,
          "RATING": 4,
          "TIMESTAMP": 879959610,
          "USER": 6,
          "movie_id": 82,
          "user_id": 12
         },
         {
          "ITEM": 2,
          "RATING": 4,
          "TIMESTAMP": 879959635,
          "USER": 6,
          "movie_id": 71,
          "user_id": 12
         },
         {
          "ITEM": 3,
          "RATING": 3,
          "TIMESTAMP": 891351471,
          "USER": 3,
          "movie_id": 82,
          "user_id": 7
         },
         {
          "ITEM": 2,
          "RATING": 4,
          "TIMESTAMP": 882398654,
          "USER": 7,
          "movie_id": 71,
          "user_id": 13
         },
         {
          "ITEM": 5,
          "RATING": 4,
          "TIMESTAMP": 877889130,
          "USER": 5,
          "movie_id": 176,
          "user_id": 10
         },
         {
          "ITEM": 5,
          "RATING": 1,
          "TIMESTAMP": 890881484,
          "USER": 8,
          "movie_id": 176,
          "user_id": 14
         },
         {
          "ITEM": 5,
          "RATING": 5,
          "TIMESTAMP": 879362233,
          "USER": 4,
          "movie_id": 176,
          "user_id": 8
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 890881557,
          "USER": 8,
          "movie_id": 50,
          "user_id": 14
         },
         {
          "ITEM": 1,
          "RATING": 4,
          "TIMESTAMP": 879959044,
          "USER": 6,
          "movie_id": 50,
          "user_id": 12
         },
         {
          "ITEM": 0,
          "RATING": 4,
          "TIMESTAMP": 877888877,
          "USER": 5,
          "movie_id": 1,
          "user_id": 10
         },
         {
          "ITEM": 5,
          "RATING": 3,
          "TIMESTAMP": 891350782,
          "USER": 3,
          "movie_id": 176,
          "user_id": 7
         },
         {
          "ITEM": 3,
          "RATING": 4,
          "TIMESTAMP": 877886912,
          "USER": 5,
          "movie_id": 82,
          "user_id": 10
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 877888545,
          "USER": 5,
          "movie_id": 50,
          "user_id": 10
         },
         {
          "ITEM": 5,
          "RATING": 5,
          "TIMESTAMP": 876892468,
          "USER": 0,
          "movie_id": 176,
          "user_id": 1
         },
         {
          "ITEM": 1,
          "RATING": 4,
          "TIMESTAMP": 883600842,
          "USER": 2,
          "movie_id": 50,
          "user_id": 6
         },
         {
          "ITEM": 4,
          "RATING": 5,
          "TIMESTAMP": 877728417,
          "USER": 9,
          "movie_id": 95,
          "user_id": 16
         },
         {
          "ITEM": 0,
          "RATING": 4,
          "TIMESTAMP": 888550871,
          "USER": 1,
          "movie_id": 1,
          "user_id": 2
         },
         {
          "ITEM": 4,
          "RATING": 5,
          "TIMESTAMP": 882140104,
          "USER": 7,
          "movie_id": 95,
          "user_id": 13
         },
         {
          "ITEM": 3,
          "RATING": 5,
          "TIMESTAMP": 879362356,
          "USER": 4,
          "movie_id": 82,
          "user_id": 8
         },
         {
          "ITEM": 0,
          "RATING": 3,
          "TIMESTAMP": 882140487,
          "USER": 7,
          "movie_id": 1,
          "user_id": 13
         },
         {
          "ITEM": 0,
          "RATING": 5,
          "TIMESTAMP": 874965758,
          "USER": 0,
          "movie_id": 1,
          "user_id": 1
         },
         {
          "ITEM": 0,
          "RATING": 5,
          "TIMESTAMP": 877717833,
          "USER": 9,
          "movie_id": 1,
          "user_id": 16
         },
         {
          "ITEM": 2,
          "RATING": 3,
          "TIMESTAMP": 876892425,
          "USER": 0,
          "movie_id": 71,
          "user_id": 1
         },
         {
          "ITEM": 2,
          "RATING": 4,
          "TIMESTAMP": 883601053,
          "USER": 2,
          "movie_id": 71,
          "user_id": 6
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 874965954,
          "USER": 0,
          "movie_id": 50,
          "user_id": 1
         },
         {
          "ITEM": 3,
          "RATING": 2,
          "TIMESTAMP": 882397503,
          "USER": 7,
          "movie_id": 82,
          "user_id": 13
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 882140001,
          "USER": 7,
          "movie_id": 50,
          "user_id": 13
         },
         {
          "ITEM": 0,
          "RATING": 4,
          "TIMESTAMP": 883599478,
          "USER": 2,
          "movie_id": 1,
          "user_id": 6
         },
         {
          "ITEM": 2,
          "RATING": 5,
          "TIMESTAMP": 877721071,
          "USER": 9,
          "movie_id": 71,
          "user_id": 16
         },
         {
          "ITEM": 2,
          "RATING": 5,
          "TIMESTAMP": 891352692,
          "USER": 3,
          "movie_id": 71,
          "user_id": 7
         },
         {
          "ITEM": 4,
          "RATING": 4,
          "TIMESTAMP": 875072303,
          "USER": 0,
          "movie_id": 95,
          "user_id": 1
         },
         {
          "ITEM": 5,
          "RATING": 3,
          "TIMESTAMP": 882140455,
          "USER": 7,
          "movie_id": 176,
          "user_id": 13
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "TIMESTAMP": 891351042,
          "USER": 3,
          "movie_id": 50,
          "user_id": 7
         }
        ]
       },
       "encoding": {
        "color": {
         "field": "RATING",
         "legend": {
          "direction": "horizontal",
          "orient": "bottom",
          "tickCount": 5,
          "titleOrient": "top"
         },
         "scale": {
          "nice": true,
          "scheme": "yellowgreenblue",
          "type": "bin-ordinal"
         },
         "type": "quantitative"
        },
        "x": {
         "axis": {
          "labelAngle": 0,
          "orient": "top"
         },
         "field": "ITEM",
         "type": "nominal"
        },
        "y": {
         "axis": {
          "orient": "left"
         },
         "field": "USER",
         "type": "nominal"
        }
       },
       "height": 300,
       "mark": "rect",
       "width": 180
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "InteractionVis(interaction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.preprocessing import RandomSplit, StratifiedSplit, ChronoSplit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = RandomSplit(interaction, [0.6, 0.4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "axis": {
         "grid": false
        },
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-72fd8a1dd6c8c9134511933e8d96f1fb"
       },
       "datasets": {
        "data-72fd8a1dd6c8c9134511933e8d96f1fb": [
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 888552084,
          "USER": 1,
          "movie_id": 50,
          "split_index": 0,
          "user_id": 2
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 879362124,
          "USER": 4,
          "movie_id": 50,
          "split_index": 0,
          "user_id": 8
         },
         {
          "ITEM": 4,
          "RATING": 2,
          "SPLIT": "train",
          "TIMESTAMP": 883602133,
          "USER": 2,
          "movie_id": 95,
          "split_index": 0,
          "user_id": 6
         },
         {
          "ITEM": 3,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 878542589,
          "USER": 0,
          "movie_id": 82,
          "split_index": 0,
          "user_id": 1
         },
         {
          "ITEM": 3,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 879959610,
          "USER": 6,
          "movie_id": 82,
          "split_index": 0,
          "user_id": 12
         },
         {
          "ITEM": 2,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 879959635,
          "USER": 6,
          "movie_id": 71,
          "split_index": 0,
          "user_id": 12
         },
         {
          "ITEM": 3,
          "RATING": 3,
          "SPLIT": "train",
          "TIMESTAMP": 891351471,
          "USER": 3,
          "movie_id": 82,
          "split_index": 0,
          "user_id": 7
         },
         {
          "ITEM": 2,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 882398654,
          "USER": 7,
          "movie_id": 71,
          "split_index": 0,
          "user_id": 13
         },
         {
          "ITEM": 5,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 877889130,
          "USER": 5,
          "movie_id": 176,
          "split_index": 0,
          "user_id": 10
         },
         {
          "ITEM": 5,
          "RATING": 1,
          "SPLIT": "train",
          "TIMESTAMP": 890881484,
          "USER": 8,
          "movie_id": 176,
          "split_index": 0,
          "user_id": 14
         },
         {
          "ITEM": 5,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 879362233,
          "USER": 4,
          "movie_id": 176,
          "split_index": 0,
          "user_id": 8
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 890881557,
          "USER": 8,
          "movie_id": 50,
          "split_index": 0,
          "user_id": 14
         },
         {
          "ITEM": 1,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 879959044,
          "USER": 6,
          "movie_id": 50,
          "split_index": 0,
          "user_id": 12
         },
         {
          "ITEM": 0,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 877888877,
          "USER": 5,
          "movie_id": 1,
          "split_index": 0,
          "user_id": 10
         },
         {
          "ITEM": 5,
          "RATING": 3,
          "SPLIT": "train",
          "TIMESTAMP": 891350782,
          "USER": 3,
          "movie_id": 176,
          "split_index": 0,
          "user_id": 7
         },
         {
          "ITEM": 3,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 877886912,
          "USER": 5,
          "movie_id": 82,
          "split_index": 0,
          "user_id": 10
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 877888545,
          "USER": 5,
          "movie_id": 50,
          "split_index": 0,
          "user_id": 10
         },
         {
          "ITEM": 5,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 876892468,
          "USER": 0,
          "movie_id": 176,
          "split_index": 0,
          "user_id": 1
         },
         {
          "ITEM": 1,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 883600842,
          "USER": 2,
          "movie_id": 50,
          "split_index": 0,
          "user_id": 6
         },
         {
          "ITEM": 4,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 877728417,
          "USER": 9,
          "movie_id": 95,
          "split_index": 0,
          "user_id": 16
         },
         {
          "ITEM": 0,
          "RATING": 4,
          "SPLIT": "train",
          "TIMESTAMP": 888550871,
          "USER": 1,
          "movie_id": 1,
          "split_index": 0,
          "user_id": 2
         },
         {
          "ITEM": 4,
          "RATING": 5,
          "SPLIT": "train",
          "TIMESTAMP": 882140104,
          "USER": 7,
          "movie_id": 95,
          "split_index": 0,
          "user_id": 13
         },
         {
          "ITEM": 3,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 879362356,
          "USER": 4,
          "movie_id": 82,
          "split_index": 1,
          "user_id": 8
         },
         {
          "ITEM": 0,
          "RATING": 3,
          "SPLIT": "test",
          "TIMESTAMP": 882140487,
          "USER": 7,
          "movie_id": 1,
          "split_index": 1,
          "user_id": 13
         },
         {
          "ITEM": 0,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 874965758,
          "USER": 0,
          "movie_id": 1,
          "split_index": 1,
          "user_id": 1
         },
         {
          "ITEM": 0,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 877717833,
          "USER": 9,
          "movie_id": 1,
          "split_index": 1,
          "user_id": 16
         },
         {
          "ITEM": 2,
          "RATING": 3,
          "SPLIT": "test",
          "TIMESTAMP": 876892425,
          "USER": 0,
          "movie_id": 71,
          "split_index": 1,
          "user_id": 1
         },
         {
          "ITEM": 2,
          "RATING": 4,
          "SPLIT": "test",
          "TIMESTAMP": 883601053,
          "USER": 2,
          "movie_id": 71,
          "split_index": 1,
          "user_id": 6
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 874965954,
          "USER": 0,
          "movie_id": 50,
          "split_index": 1,
          "user_id": 1
         },
         {
          "ITEM": 3,
          "RATING": 2,
          "SPLIT": "test",
          "TIMESTAMP": 882397503,
          "USER": 7,
          "movie_id": 82,
          "split_index": 1,
          "user_id": 13
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 882140001,
          "USER": 7,
          "movie_id": 50,
          "split_index": 1,
          "user_id": 13
         },
         {
          "ITEM": 0,
          "RATING": 4,
          "SPLIT": "test",
          "TIMESTAMP": 883599478,
          "USER": 2,
          "movie_id": 1,
          "split_index": 1,
          "user_id": 6
         },
         {
          "ITEM": 2,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 877721071,
          "USER": 9,
          "movie_id": 71,
          "split_index": 1,
          "user_id": 16
         },
         {
          "ITEM": 2,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 891352692,
          "USER": 3,
          "movie_id": 71,
          "split_index": 1,
          "user_id": 7
         },
         {
          "ITEM": 4,
          "RATING": 4,
          "SPLIT": "test",
          "TIMESTAMP": 875072303,
          "USER": 0,
          "movie_id": 95,
          "split_index": 1,
          "user_id": 1
         },
         {
          "ITEM": 5,
          "RATING": 3,
          "SPLIT": "test",
          "TIMESTAMP": 882140455,
          "USER": 7,
          "movie_id": 176,
          "split_index": 1,
          "user_id": 13
         },
         {
          "ITEM": 1,
          "RATING": 5,
          "SPLIT": "test",
          "TIMESTAMP": 891351042,
          "USER": 3,
          "movie_id": 50,
          "split_index": 1,
          "user_id": 7
         }
        ]
       },
       "encoding": {
        "color": {
         "field": "SPLIT",
         "legend": {
          "direction": "horizontal",
          "orient": "bottom",
          "tickCount": 5,
          "titleOrient": "top"
         },
         "scale": {
          "nice": true,
          "scheme": "darkred",
          "type": "ordinal"
         },
         "type": "ordinal"
        },
        "opacity": {
         "value": 1
        },
        "x": {
         "axis": {
          "labelAngle": 0,
          "orient": "top"
         },
         "field": "ITEM",
         "type": "nominal"
        },
        "y": {
         "axis": {
          "orient": "left"
         },
         "field": "USER",
         "type": "nominal"
        }
       },
       "height": 300,
       "mark": "rect",
       "width": 180
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TrainTestVis(train, test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.models import ExplicitMatrixFactorisationBias"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_factors = 2\n",
    "model = ExplicitMatrixFactorisationBias(n_users, n_items, n_factors, max_rating, min_rating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"ExplicitMatrixFactorisationBias\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "Item (InputLayer)               (None, 1)            0                                            \n",
      "__________________________________________________________________________________________________\n",
      "User (InputLayer)               (None, 1)            0                                            \n",
      "__________________________________________________________________________________________________\n",
      "ItemEmbedding (Embedding)       (None, 1, 2)         12          Item[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "UserEmbedding (Embedding)       (None, 1, 2)         20          User[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "FlattenItemE (Flatten)          (None, 2)            0           ItemEmbedding[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "FlattenUserE (Flatten)          (None, 2)            0           UserEmbedding[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "ItemBias (Embedding)            (None, 1, 1)         6           Item[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "UserBias (Embedding)            (None, 1, 1)         10          User[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "DotProduct (Dot)                (None, 1)            0           FlattenItemE[0][0]               \n",
      "                                                                 FlattenUserE[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "FlattenItemBiasE (Flatten)      (None, 1)            0           ItemBias[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "FlattenUserBiasE (Flatten)      (None, 1)            0           UserBias[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "AddBias (Add)                   (None, 1)            0           DotProduct[0][0]                 \n",
      "                                                                 FlattenItemBiasE[0][0]           \n",
      "                                                                 FlattenUserBiasE[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "activation_3 (Activation)       (None, 1)            0           AddBias[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "lambda_3 (Lambda)               (None, 1)            0           activation_3[0][0]               \n",
      "==================================================================================================\n",
      "Total params: 48\n",
      "Trainable params: 48\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 8.51 s, sys: 1.25 s, total: 9.75 s\n",
      "Wall time: 7.33 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "output = model.fit([train.USER, train.ITEM], train.RATING, shuffle=True, batch_size=1, epochs=100, verbose=0, \n",
    "                   validation_data=([test.USER, test.ITEM], test.RATING))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.vis import MetricsVis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-39203a5615d02b19756b81aee8e0e06e"
       },
       "datasets": {
        "data-39203a5615d02b19756b81aee8e0e06e": [
         {
          "batch": 1,
          "metrics": "loss",
          "type": "validation",
          "value": 2.3595998018378546
         },
         {
          "batch": 2,
          "metrics": "loss",
          "type": "validation",
          "value": 2.3363067559810586
         },
         {
          "batch": 3,
          "metrics": "loss",
          "type": "validation",
          "value": 2.315473632500895
         },
         {
          "batch": 4,
          "metrics": "loss",
          "type": "validation",
          "value": 2.2899373012594877
         },
         {
          "batch": 5,
          "metrics": "loss",
          "type": "validation",
          "value": 2.2667550797419
         },
         {
          "batch": 6,
          "metrics": "loss",
          "type": "validation",
          "value": 2.2430516773834825
         },
         {
          "batch": 7,
          "metrics": "loss",
          "type": "validation",
          "value": 2.218928933485101
         },
         {
          "batch": 8,
          "metrics": "loss",
          "type": "validation",
          "value": 2.1971759140801925
         },
         {
          "batch": 9,
          "metrics": "loss",
          "type": "validation",
          "value": 2.1741403366749483
         },
         {
          "batch": 10,
          "metrics": "loss",
          "type": "validation",
          "value": 2.1512243054807185
         },
         {
          "batch": 11,
          "metrics": "loss",
          "type": "validation",
          "value": 2.1290950644140443
         },
         {
          "batch": 12,
          "metrics": "loss",
          "type": "validation",
          "value": 2.106534293293953
         },
         {
          "batch": 13,
          "metrics": "loss",
          "type": "validation",
          "value": 2.085305488482118
         },
         {
          "batch": 14,
          "metrics": "loss",
          "type": "validation",
          "value": 2.06608888382713
         },
         {
          "batch": 15,
          "metrics": "loss",
          "type": "validation",
          "value": 2.043294706195593
         },
         {
          "batch": 16,
          "metrics": "loss",
          "type": "validation",
          "value": 2.0227368257939817
         },
         {
          "batch": 17,
          "metrics": "loss",
          "type": "validation",
          "value": 2.001574547464649
         },
         {
          "batch": 18,
          "metrics": "loss",
          "type": "validation",
          "value": 1.9831479653716086
         },
         {
          "batch": 19,
          "metrics": "loss",
          "type": "validation",
          "value": 1.9637038740019004
         },
         {
          "batch": 20,
          "metrics": "loss",
          "type": "validation",
          "value": 1.9433827514449755
         },
         {
          "batch": 21,
          "metrics": "loss",
          "type": "validation",
          "value": 1.924477122972409
         },
         {
          "batch": 22,
          "metrics": "loss",
          "type": "validation",
          "value": 1.9084426052868366
         },
         {
          "batch": 23,
          "metrics": "loss",
          "type": "validation",
          "value": 1.888773933549722
         },
         {
          "batch": 24,
          "metrics": "loss",
          "type": "validation",
          "value": 1.8722489669919014
         },
         {
          "batch": 25,
          "metrics": "loss",
          "type": "validation",
          "value": 1.8554005096356074
         },
         {
          "batch": 26,
          "metrics": "loss",
          "type": "validation",
          "value": 1.837032686173916
         },
         {
          "batch": 27,
          "metrics": "loss",
          "type": "validation",
          "value": 1.8239056438207626
         },
         {
          "batch": 28,
          "metrics": "loss",
          "type": "validation",
          "value": 1.8072071557243665
         },
         {
          "batch": 29,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7928590521216392
         },
         {
          "batch": 30,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7798539206385613
         },
         {
          "batch": 31,
          "metrics": "loss",
          "type": "validation",
          "value": 1.76434163103501
         },
         {
          "batch": 32,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7522001897295316
         },
         {
          "batch": 33,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7383466030160586
         },
         {
          "batch": 34,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7265618364016215
         },
         {
          "batch": 35,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7133323887983958
         },
         {
          "batch": 36,
          "metrics": "loss",
          "type": "validation",
          "value": 1.7028594354788462
         },
         {
          "batch": 37,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6935475885868072
         },
         {
          "batch": 38,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6811999986569086
         },
         {
          "batch": 39,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6694642623265585
         },
         {
          "batch": 40,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6635905812184015
         },
         {
          "batch": 41,
          "metrics": "loss",
          "type": "validation",
          "value": 1.654229222238064
         },
         {
          "batch": 42,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6444485902786254
         },
         {
          "batch": 43,
          "metrics": "loss",
          "type": "validation",
          "value": 1.639475513001283
         },
         {
          "batch": 44,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6298139462868373
         },
         {
          "batch": 45,
          "metrics": "loss",
          "type": "validation",
          "value": 1.624733842164278
         },
         {
          "batch": 46,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6165392364064852
         },
         {
          "batch": 47,
          "metrics": "loss",
          "type": "validation",
          "value": 1.611235468586286
         },
         {
          "batch": 48,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6066598532100518
         },
         {
          "batch": 49,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6010827884078025
         },
         {
          "batch": 50,
          "metrics": "loss",
          "type": "validation",
          "value": 1.593630864843726
         },
         {
          "batch": 51,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5906035035848618
         },
         {
          "batch": 52,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5869181635479133
         },
         {
          "batch": 53,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5833799857646227
         },
         {
          "batch": 54,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5794398999462524
         },
         {
          "batch": 55,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5756570187707741
         },
         {
          "batch": 56,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5735216255222138
         },
         {
          "batch": 57,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5735630884300917
         },
         {
          "batch": 58,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5709348098878158
         },
         {
          "batch": 59,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5683452395838686
         },
         {
          "batch": 60,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5700622662669048
         },
         {
          "batch": 61,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5644863976320873
         },
         {
          "batch": 62,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5676329120062291
         },
         {
          "batch": 63,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5623399190604688
         },
         {
          "batch": 64,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5654218966141342
         },
         {
          "batch": 65,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5631119190404812
         },
         {
          "batch": 66,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5612002942711114
         },
         {
          "batch": 67,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5632244041810432
         },
         {
          "batch": 68,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5652808836350838
         },
         {
          "batch": 69,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5647269854942958
         },
         {
          "batch": 70,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5628180871407191
         },
         {
          "batch": 71,
          "metrics": "loss",
          "type": "validation",
          "value": 1.567386752118667
         },
         {
          "batch": 72,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5655802324414254
         },
         {
          "batch": 73,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5678146824240684
         },
         {
          "batch": 74,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5695800950129828
         },
         {
          "batch": 75,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5684153576691946
         },
         {
          "batch": 76,
          "metrics": "loss",
          "type": "validation",
          "value": 1.569999777774016
         },
         {
          "batch": 77,
          "metrics": "loss",
          "type": "validation",
          "value": 1.572445565958818
         },
         {
          "batch": 78,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5735840489466986
         },
         {
          "batch": 79,
          "metrics": "loss",
          "type": "validation",
          "value": 1.577104076743126
         },
         {
          "batch": 80,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5781962886452674
         },
         {
          "batch": 81,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5793609728415807
         },
         {
          "batch": 82,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5824805667002997
         },
         {
          "batch": 83,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5825329651435216
         },
         {
          "batch": 84,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5856510122617087
         },
         {
          "batch": 85,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5913439214229583
         },
         {
          "batch": 86,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5906568348407746
         },
         {
          "batch": 87,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5921064227819444
         },
         {
          "batch": 88,
          "metrics": "loss",
          "type": "validation",
          "value": 1.59596655468146
         },
         {
          "batch": 89,
          "metrics": "loss",
          "type": "validation",
          "value": 1.5947074751059214
         },
         {
          "batch": 90,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6024138768513998
         },
         {
          "batch": 91,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6013300826152166
         },
         {
          "batch": 92,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6023160884777705
         },
         {
          "batch": 93,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6072221239407858
         },
         {
          "batch": 94,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6078603625297547
         },
         {
          "batch": 95,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6109299659729004
         },
         {
          "batch": 96,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6141657173633575
         },
         {
          "batch": 97,
          "metrics": "loss",
          "type": "validation",
          "value": 1.616324112812678
         },
         {
          "batch": 98,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6181227246920267
         },
         {
          "batch": 99,
          "metrics": "loss",
          "type": "validation",
          "value": 1.6203107078870138
         },
         {
          "batch": 100,
          "metrics": "loss",
          "type": "validation",
          "value": 1.623794561624527
         },
         {
          "batch": 1,
          "metrics": "loss",
          "type": "training",
          "value": 2.3013224284090525
         },
         {
          "batch": 2,
          "metrics": "loss",
          "type": "training",
          "value": 2.2691161685877224
         },
         {
          "batch": 3,
          "metrics": "loss",
          "type": "training",
          "value": 2.2341159943316597
         },
         {
          "batch": 4,
          "metrics": "loss",
          "type": "training",
          "value": 2.203760811545759
         },
         {
          "batch": 5,
          "metrics": "loss",
          "type": "training",
          "value": 2.1704795408163178
         },
         {
          "batch": 6,
          "metrics": "loss",
          "type": "training",
          "value": 2.1365029863187703
         },
         {
          "batch": 7,
          "metrics": "loss",
          "type": "training",
          "value": 2.1033236294804905
         },
         {
          "batch": 8,
          "metrics": "loss",
          "type": "training",
          "value": 2.069763127849993
         },
         {
          "batch": 9,
          "metrics": "loss",
          "type": "training",
          "value": 2.0371861916750773
         },
         {
          "batch": 10,
          "metrics": "loss",
          "type": "training",
          "value": 2.0033635041376825
         },
         {
          "batch": 11,
          "metrics": "loss",
          "type": "training",
          "value": 1.9691003483562781
         },
         {
          "batch": 12,
          "metrics": "loss",
          "type": "training",
          "value": 1.9355594221280874
         },
         {
          "batch": 13,
          "metrics": "loss",
          "type": "training",
          "value": 1.9035805509718413
         },
         {
          "batch": 14,
          "metrics": "loss",
          "type": "training",
          "value": 1.8679192427643598
         },
         {
          "batch": 15,
          "metrics": "loss",
          "type": "training",
          "value": 1.8335105669033178
         },
         {
          "batch": 16,
          "metrics": "loss",
          "type": "training",
          "value": 1.8000174591606166
         },
         {
          "batch": 17,
          "metrics": "loss",
          "type": "training",
          "value": 1.7653345656144666
         },
         {
          "batch": 18,
          "metrics": "loss",
          "type": "training",
          "value": 1.73219833555843
         },
         {
          "batch": 19,
          "metrics": "loss",
          "type": "training",
          "value": 1.6959928275636313
         },
         {
          "batch": 20,
          "metrics": "loss",
          "type": "training",
          "value": 1.6622610920167062
         },
         {
          "batch": 21,
          "metrics": "loss",
          "type": "training",
          "value": 1.6293730689001016
         },
         {
          "batch": 22,
          "metrics": "loss",
          "type": "training",
          "value": 1.5932400203124748
         },
         {
          "batch": 23,
          "metrics": "loss",
          "type": "training",
          "value": 1.5625795236026698
         },
         {
          "batch": 24,
          "metrics": "loss",
          "type": "training",
          "value": 1.5275884610177441
         },
         {
          "batch": 25,
          "metrics": "loss",
          "type": "training",
          "value": 1.4957981174123811
         },
         {
          "batch": 26,
          "metrics": "loss",
          "type": "training",
          "value": 1.4626306768676096
         },
         {
          "batch": 27,
          "metrics": "loss",
          "type": "training",
          "value": 1.4303240630066616
         },
         {
          "batch": 28,
          "metrics": "loss",
          "type": "training",
          "value": 1.3991307412295348
         },
         {
          "batch": 29,
          "metrics": "loss",
          "type": "training",
          "value": 1.3650824895382605
         },
         {
          "batch": 30,
          "metrics": "loss",
          "type": "training",
          "value": 1.3361086200351233
         },
         {
          "batch": 31,
          "metrics": "loss",
          "type": "training",
          "value": 1.3057018805676224
         },
         {
          "batch": 32,
          "metrics": "loss",
          "type": "training",
          "value": 1.2730245234402404
         },
         {
          "batch": 33,
          "metrics": "loss",
          "type": "training",
          "value": 1.2449263978791847
         },
         {
          "batch": 34,
          "metrics": "loss",
          "type": "training",
          "value": 1.2153161489298905
         },
         {
          "batch": 35,
          "metrics": "loss",
          "type": "training",
          "value": 1.186723468370143
         },
         {
          "batch": 36,
          "metrics": "loss",
          "type": "training",
          "value": 1.1592918695162304
         },
         {
          "batch": 37,
          "metrics": "loss",
          "type": "training",
          "value": 1.13158595521765
         },
         {
          "batch": 38,
          "metrics": "loss",
          "type": "training",
          "value": 1.1042939537120136
         },
         {
          "batch": 39,
          "metrics": "loss",
          "type": "training",
          "value": 1.0796793720430948
         },
         {
          "batch": 40,
          "metrics": "loss",
          "type": "training",
          "value": 1.0515632095259868
         },
         {
          "batch": 41,
          "metrics": "loss",
          "type": "training",
          "value": 1.0279137744695286
         },
         {
          "batch": 42,
          "metrics": "loss",
          "type": "training",
          "value": 1.0010566421657463
         },
         {
          "batch": 43,
          "metrics": "loss",
          "type": "training",
          "value": 0.9789737980939787
         },
         {
          "batch": 44,
          "metrics": "loss",
          "type": "training",
          "value": 0.955364637140354
         },
         {
          "batch": 45,
          "metrics": "loss",
          "type": "training",
          "value": 0.9316545430232178
         },
         {
          "batch": 46,
          "metrics": "loss",
          "type": "training",
          "value": 0.909601031167602
         },
         {
          "batch": 47,
          "metrics": "loss",
          "type": "training",
          "value": 0.8862795671024783
         },
         {
          "batch": 48,
          "metrics": "loss",
          "type": "training",
          "value": 0.8666291014206681
         },
         {
          "batch": 49,
          "metrics": "loss",
          "type": "training",
          "value": 0.8444258130832829
         },
         {
          "batch": 50,
          "metrics": "loss",
          "type": "training",
          "value": 0.8238591416738927
         },
         {
          "batch": 51,
          "metrics": "loss",
          "type": "training",
          "value": 0.8036125224633989
         },
         {
          "batch": 52,
          "metrics": "loss",
          "type": "training",
          "value": 0.7869317274413664
         },
         {
          "batch": 53,
          "metrics": "loss",
          "type": "training",
          "value": 0.7650453148579056
         },
         {
          "batch": 54,
          "metrics": "loss",
          "type": "training",
          "value": 0.7484154609793966
         },
         {
          "batch": 55,
          "metrics": "loss",
          "type": "training",
          "value": 0.7296215512535789
         },
         {
          "batch": 56,
          "metrics": "loss",
          "type": "training",
          "value": 0.7118577251041477
         },
         {
          "batch": 57,
          "metrics": "loss",
          "type": "training",
          "value": 0.6963252412705597
         },
         {
          "batch": 58,
          "metrics": "loss",
          "type": "training",
          "value": 0.6780754934403707
         },
         {
          "batch": 59,
          "metrics": "loss",
          "type": "training",
          "value": 0.6632074401396412
         },
         {
          "batch": 60,
          "metrics": "loss",
          "type": "training",
          "value": 0.6474162024890326
         },
         {
          "batch": 61,
          "metrics": "loss",
          "type": "training",
          "value": 0.6317600892560388
         },
         {
          "batch": 62,
          "metrics": "loss",
          "type": "training",
          "value": 0.6178098538604734
         },
         {
          "batch": 63,
          "metrics": "loss",
          "type": "training",
          "value": 0.6023489313877441
         },
         {
          "batch": 64,
          "metrics": "loss",
          "type": "training",
          "value": 0.5892421014948261
         },
         {
          "batch": 65,
          "metrics": "loss",
          "type": "training",
          "value": 0.5744218029186331
         },
         {
          "batch": 66,
          "metrics": "loss",
          "type": "training",
          "value": 0.5624754370698859
         },
         {
          "batch": 67,
          "metrics": "loss",
          "type": "training",
          "value": 0.5483085986586296
         },
         {
          "batch": 68,
          "metrics": "loss",
          "type": "training",
          "value": 0.5358850668917925
         },
         {
          "batch": 69,
          "metrics": "loss",
          "type": "training",
          "value": 0.5238740804390379
         },
         {
          "batch": 70,
          "metrics": "loss",
          "type": "training",
          "value": 0.5121232129796144
         },
         {
          "batch": 71,
          "metrics": "loss",
          "type": "training",
          "value": 0.5000097611945503
         },
         {
          "batch": 72,
          "metrics": "loss",
          "type": "training",
          "value": 0.48871137060632464
         },
         {
          "batch": 73,
          "metrics": "loss",
          "type": "training",
          "value": 0.4776896297222596
         },
         {
          "batch": 74,
          "metrics": "loss",
          "type": "training",
          "value": 0.4665011755233453
         },
         {
          "batch": 75,
          "metrics": "loss",
          "type": "training",
          "value": 0.456456836232519
         },
         {
          "batch": 76,
          "metrics": "loss",
          "type": "training",
          "value": 0.4461300003081719
         },
         {
          "batch": 77,
          "metrics": "loss",
          "type": "training",
          "value": 0.43630195868602395
         },
         {
          "batch": 78,
          "metrics": "loss",
          "type": "training",
          "value": 0.4257433903674775
         },
         {
          "batch": 79,
          "metrics": "loss",
          "type": "training",
          "value": 0.4169131620040968
         },
         {
          "batch": 80,
          "metrics": "loss",
          "type": "training",
          "value": 0.4077998477044854
         },
         {
          "batch": 81,
          "metrics": "loss",
          "type": "training",
          "value": 0.3982865500146628
         },
         {
          "batch": 82,
          "metrics": "loss",
          "type": "training",
          "value": 0.39006410217949783
         },
         {
          "batch": 83,
          "metrics": "loss",
          "type": "training",
          "value": 0.38103121013992597
         },
         {
          "batch": 84,
          "metrics": "loss",
          "type": "training",
          "value": 0.3724134687621369
         },
         {
          "batch": 85,
          "metrics": "loss",
          "type": "training",
          "value": 0.3647224415349642
         },
         {
          "batch": 86,
          "metrics": "loss",
          "type": "training",
          "value": 0.3568251856406939
         },
         {
          "batch": 87,
          "metrics": "loss",
          "type": "training",
          "value": 0.3492895796291784
         },
         {
          "batch": 88,
          "metrics": "loss",
          "type": "training",
          "value": 0.34135424074404663
         },
         {
          "batch": 89,
          "metrics": "loss",
          "type": "training",
          "value": 0.33398595375712664
         },
         {
          "batch": 90,
          "metrics": "loss",
          "type": "training",
          "value": 0.32701790423280647
         },
         {
          "batch": 91,
          "metrics": "loss",
          "type": "training",
          "value": 0.3196979194485307
         },
         {
          "batch": 92,
          "metrics": "loss",
          "type": "training",
          "value": 0.31284893948891707
         },
         {
          "batch": 93,
          "metrics": "loss",
          "type": "training",
          "value": 0.3057792665891942
         },
         {
          "batch": 94,
          "metrics": "loss",
          "type": "training",
          "value": 0.2997834361383205
         },
         {
          "batch": 95,
          "metrics": "loss",
          "type": "training",
          "value": 0.29339530207190284
         },
         {
          "batch": 96,
          "metrics": "loss",
          "type": "training",
          "value": 0.2868938151295879
         },
         {
          "batch": 97,
          "metrics": "loss",
          "type": "training",
          "value": 0.2813592610959445
         },
         {
          "batch": 98,
          "metrics": "loss",
          "type": "training",
          "value": 0.27488395580175246
         },
         {
          "batch": 99,
          "metrics": "loss",
          "type": "training",
          "value": 0.2700864557882596
         },
         {
          "batch": 100,
          "metrics": "loss",
          "type": "training",
          "value": 0.2633903234015055
         }
        ]
       },
       "facet": {
        "column": {
         "field": "metrics",
         "type": "nominal"
        }
       },
       "resolve": {
        "scale": {
         "y": "independent"
        }
       },
       "spec": {
        "layer": [
         {
          "encoding": {
           "color": {
            "field": "type",
            "type": "nominal"
           },
           "tooltip": [
            {
             "field": "batch",
             "type": "quantitative"
            },
            {
             "field": "value",
             "type": "quantitative"
            }
           ],
           "x": {
            "field": "batch",
            "type": "quantitative"
           },
           "y": {
            "field": "value",
            "type": "quantitative"
           }
          },
          "height": 300,
          "mark": {
           "size": 50,
           "type": "circle"
          },
          "width": 300
         },
         {
          "encoding": {
           "color": {
            "field": "type",
            "type": "nominal"
           },
           "x": {
            "field": "batch",
            "type": "quantitative"
           },
           "y": {
            "field": "value",
            "type": "quantitative"
           }
          },
          "height": 300,
          "mark": "line",
          "width": 300
         }
        ]
       }
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MetricsVis(output.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.recommend import UserEmbedding, ItemEmbedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [],
   "source": [
    "item_embedding = UserEmbedding(model, \"ItemEmbedding\")\n",
    "user_embedding = ItemEmbedding(model, \"UserEmbedding\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.4266383 ,  0.43679917],\n",
       "       [ 0.7709527 ,  1.0180782 ],\n",
       "       [ 0.49435946, -0.35138014],\n",
       "       [ 0.72814727, -0.51169   ],\n",
       "       [ 0.80276716, -0.93691117],\n",
       "       [ 0.88108987, -0.6649029 ]], dtype=float32)"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "item_embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.7634384 , -0.7496685 ],\n",
       "       [ 0.5777171 ,  0.602675  ],\n",
       "       [-0.24383129,  0.75017303],\n",
       "       [-0.14105083,  0.05877911],\n",
       "       [ 0.85818356,  0.08425549],\n",
       "       [ 0.80780387,  0.36935842],\n",
       "       [ 0.5184164 , -0.15546027],\n",
       "       [ 0.5371777 , -0.62514496],\n",
       "       [-0.1534093 ,  0.8556748 ],\n",
       "       [ 0.5974195 , -0.5847428 ]], dtype=float32)"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_embedding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "ename": "ImportError",
     "evalue": "cannot import name 'EmbeddingVis' from 'recoflow.vis' (/Users/amitkaps/miniconda3/envs/ml/lib/python3.7/site-packages/recoflow/vis.py)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m-----------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mImportError\u001b[0m                           Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-212-d6aa3bd8c158>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mrecoflow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvis\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mEmbeddingVis\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mImportError\u001b[0m: cannot import name 'EmbeddingVis' from 'recoflow.vis' (/Users/amitkaps/miniconda3/envs/ml/lib/python3.7/site-packages/recoflow/vis.py)"
     ]
    }
   ],
   "source": [
    "from recoflow.vis import EmbeddingVis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-3a162cda04c1558a214665aac001c2ca"
       },
       "datasets": {
        "data-3a162cda04c1558a214665aac001c2ca": [
         {
          "ITEM": 0,
          "dim": 0,
          "value": 0.42663830518722534
         },
         {
          "ITEM": 1,
          "dim": 0,
          "value": 0.7709527015686035
         },
         {
          "ITEM": 2,
          "dim": 0,
          "value": 0.49435946345329285
         },
         {
          "ITEM": 3,
          "dim": 0,
          "value": 0.7281472682952881
         },
         {
          "ITEM": 4,
          "dim": 0,
          "value": 0.8027671575546265
         },
         {
          "ITEM": 5,
          "dim": 0,
          "value": 0.8810898661613464
         },
         {
          "ITEM": 0,
          "dim": 1,
          "value": 0.43679916858673096
         },
         {
          "ITEM": 1,
          "dim": 1,
          "value": 1.0180782079696655
         },
         {
          "ITEM": 2,
          "dim": 1,
          "value": -0.3513801395893097
         },
         {
          "ITEM": 3,
          "dim": 1,
          "value": -0.5116900205612183
         },
         {
          "ITEM": 4,
          "dim": 1,
          "value": -0.9369111657142639
         },
         {
          "ITEM": 5,
          "dim": 1,
          "value": -0.664902925491333
         }
        ]
       },
       "encoding": {
        "color": {
         "field": "value",
         "legend": {
          "direction": "horizontal",
          "orient": "bottom",
          "tickCount": 5,
          "titleOrient": "top"
         },
         "scale": {
          "nice": true,
          "scheme": "yellowgreenblue",
          "type": "bin-ordinal"
         },
         "type": "quantitative"
        },
        "x": {
         "axis": {
          "labelAngle": 0,
          "orient": "top"
         },
         "field": "ITEM",
         "type": "nominal"
        },
        "y": {
         "axis": {
          "orient": "left"
         },
         "field": "dim",
         "type": "nominal"
        }
       },
       "height": 60,
       "mark": "rect",
       "width": 180
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "EmbeddingVis(item_embedding, n_factors, \"ITEM\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-e3e1ab306af7a6eb4a5ff1f9ae990b2b"
       },
       "datasets": {
        "data-e3e1ab306af7a6eb4a5ff1f9ae990b2b": [
         {
          "USER": 0,
          "dim": 0,
          "value": 0.7634384036064148
         },
         {
          "USER": 1,
          "dim": 0,
          "value": 0.577717125415802
         },
         {
          "USER": 2,
          "dim": 0,
          "value": -0.24383129179477692
         },
         {
          "USER": 3,
          "dim": 0,
          "value": -0.14105083048343658
         },
         {
          "USER": 4,
          "dim": 0,
          "value": 0.8581835627555847
         },
         {
          "USER": 5,
          "dim": 0,
          "value": 0.8078038692474365
         },
         {
          "USER": 6,
          "dim": 0,
          "value": 0.5184164047241211
         },
         {
          "USER": 7,
          "dim": 0,
          "value": 0.5371776819229126
         },
         {
          "USER": 8,
          "dim": 0,
          "value": -0.15340930223464966
         },
         {
          "USER": 9,
          "dim": 0,
          "value": 0.5974195003509521
         },
         {
          "USER": 0,
          "dim": 1,
          "value": -0.7496684789657593
         },
         {
          "USER": 1,
          "dim": 1,
          "value": 0.6026750206947327
         },
         {
          "USER": 2,
          "dim": 1,
          "value": 0.750173032283783
         },
         {
          "USER": 3,
          "dim": 1,
          "value": 0.05877910926938057
         },
         {
          "USER": 4,
          "dim": 1,
          "value": 0.08425549417734146
         },
         {
          "USER": 5,
          "dim": 1,
          "value": 0.3693584203720093
         },
         {
          "USER": 6,
          "dim": 1,
          "value": -0.15546026825904846
         },
         {
          "USER": 7,
          "dim": 1,
          "value": -0.6251449584960938
         },
         {
          "USER": 8,
          "dim": 1,
          "value": 0.8556748032569885
         },
         {
          "USER": 9,
          "dim": 1,
          "value": -0.5847427845001221
         }
        ]
       },
       "encoding": {
        "color": {
         "field": "value",
         "legend": {
          "direction": "horizontal",
          "orient": "bottom",
          "tickCount": 5,
          "titleOrient": "top"
         },
         "scale": {
          "nice": true,
          "scheme": "yellowgreenblue",
          "type": "bin-ordinal"
         },
         "type": "quantitative"
        },
        "x": {
         "axis": {
          "labelAngle": 0,
          "orient": "top"
         },
         "field": "dim",
         "type": "nominal"
        },
        "y": {
         "axis": {
          "orient": "left"
         },
         "field": "USER",
         "type": "nominal"
        }
       },
       "height": 300,
       "mark": "rect",
       "width": 60
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "EmbeddingVis(user_embedding, n_factors, \"USER\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "datasets": {
        "data-3a162cda04c1558a214665aac001c2ca": [
         {
          "ITEM": 0,
          "dim": 0,
          "value": 0.42663830518722534
         },
         {
          "ITEM": 1,
          "dim": 0,
          "value": 0.7709527015686035
         },
         {
          "ITEM": 2,
          "dim": 0,
          "value": 0.49435946345329285
         },
         {
          "ITEM": 3,
          "dim": 0,
          "value": 0.7281472682952881
         },
         {
          "ITEM": 4,
          "dim": 0,
          "value": 0.8027671575546265
         },
         {
          "ITEM": 5,
          "dim": 0,
          "value": 0.8810898661613464
         },
         {
          "ITEM": 0,
          "dim": 1,
          "value": 0.43679916858673096
         },
         {
          "ITEM": 1,
          "dim": 1,
          "value": 1.0180782079696655
         },
         {
          "ITEM": 2,
          "dim": 1,
          "value": -0.3513801395893097
         },
         {
          "ITEM": 3,
          "dim": 1,
          "value": -0.5116900205612183
         },
         {
          "ITEM": 4,
          "dim": 1,
          "value": -0.9369111657142639
         },
         {
          "ITEM": 5,
          "dim": 1,
          "value": -0.664902925491333
         }
        ],
        "data-e3e1ab306af7a6eb4a5ff1f9ae990b2b": [
         {
          "USER": 0,
          "dim": 0,
          "value": 0.7634384036064148
         },
         {
          "USER": 1,
          "dim": 0,
          "value": 0.577717125415802
         },
         {
          "USER": 2,
          "dim": 0,
          "value": -0.24383129179477692
         },
         {
          "USER": 3,
          "dim": 0,
          "value": -0.14105083048343658
         },
         {
          "USER": 4,
          "dim": 0,
          "value": 0.8581835627555847
         },
         {
          "USER": 5,
          "dim": 0,
          "value": 0.8078038692474365
         },
         {
          "USER": 6,
          "dim": 0,
          "value": 0.5184164047241211
         },
         {
          "USER": 7,
          "dim": 0,
          "value": 0.5371776819229126
         },
         {
          "USER": 8,
          "dim": 0,
          "value": -0.15340930223464966
         },
         {
          "USER": 9,
          "dim": 0,
          "value": 0.5974195003509521
         },
         {
          "USER": 0,
          "dim": 1,
          "value": -0.7496684789657593
         },
         {
          "USER": 1,
          "dim": 1,
          "value": 0.6026750206947327
         },
         {
          "USER": 2,
          "dim": 1,
          "value": 0.750173032283783
         },
         {
          "USER": 3,
          "dim": 1,
          "value": 0.05877910926938057
         },
         {
          "USER": 4,
          "dim": 1,
          "value": 0.08425549417734146
         },
         {
          "USER": 5,
          "dim": 1,
          "value": 0.3693584203720093
         },
         {
          "USER": 6,
          "dim": 1,
          "value": -0.15546026825904846
         },
         {
          "USER": 7,
          "dim": 1,
          "value": -0.6251449584960938
         },
         {
          "USER": 8,
          "dim": 1,
          "value": 0.8556748032569885
         },
         {
          "USER": 9,
          "dim": 1,
          "value": -0.5847427845001221
         }
        ]
       },
       "hconcat": [
        {
         "data": {
          "name": "data-e3e1ab306af7a6eb4a5ff1f9ae990b2b"
         },
         "encoding": {
          "color": {
           "field": "value",
           "legend": {
            "direction": "horizontal",
            "orient": "bottom",
            "tickCount": 5,
            "titleOrient": "top"
           },
           "scale": {
            "nice": true,
            "scheme": "yellowgreenblue",
            "type": "bin-ordinal"
           },
           "type": "quantitative"
          },
          "x": {
           "axis": {
            "labelAngle": 0,
            "orient": "top"
           },
           "field": "dim",
           "type": "nominal"
          },
          "y": {
           "axis": {
            "orient": "left"
           },
           "field": "USER",
           "type": "nominal"
          }
         },
         "height": 300,
         "mark": "rect",
         "width": 60
        },
        {
         "data": {
          "name": "data-3a162cda04c1558a214665aac001c2ca"
         },
         "encoding": {
          "color": {
           "field": "value",
           "legend": {
            "direction": "horizontal",
            "orient": "bottom",
            "tickCount": 5,
            "titleOrient": "top"
           },
           "scale": {
            "nice": true,
            "scheme": "yellowgreenblue",
            "type": "bin-ordinal"
           },
           "type": "quantitative"
          },
          "x": {
           "axis": {
            "labelAngle": 0,
            "orient": "top"
           },
           "field": "ITEM",
           "type": "nominal"
          },
          "y": {
           "axis": {
            "orient": "left"
           },
           "field": "dim",
           "type": "nominal"
          }
         },
         "height": 60,
         "mark": "rect",
         "width": 180
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 211,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "EmbeddingVis(user_embedding, n_factors, \"USER\") | EmbeddingVis(item_embedding, n_factors, \"ITEM\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [],
   "source": [
    "from recoflow.recommend import GetPredictions, GetRankingTopK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = GetPredictions(model, interaction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "axis": {
         "grid": false
        },
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-030053414facf0c95d4c0916054f902f"
       },
       "datasets": {
        "data-030053414facf0c95d4c0916054f902f": [
         {
          "ITEM": 1,
          "RATING": 4.525042533874512,
          "USER": 1
         },
         {
          "ITEM": 4,
          "RATING": 3.4519119262695312,
          "USER": 1
         },
         {
          "ITEM": 3,
          "RATING": 3.758251905441284,
          "USER": 1
         },
         {
          "ITEM": 2,
          "RATING": 3.684328556060791,
          "USER": 1
         },
         {
          "ITEM": 5,
          "RATING": 3.5420827865600586,
          "USER": 1
         },
         {
          "ITEM": 0,
          "RATING": 3.9941534996032715,
          "USER": 1
         },
         {
          "ITEM": 1,
          "RATING": 4.457489967346191,
          "USER": 4
         },
         {
          "ITEM": 4,
          "RATING": 4.163325786590576,
          "USER": 4
         },
         {
          "ITEM": 3,
          "RATING": 4.225610733032227,
          "USER": 4
         },
         {
          "ITEM": 2,
          "RATING": 4.069202899932861,
          "USER": 4
         },
         {
          "ITEM": 5,
          "RATING": 4.148058891296387,
          "USER": 4
         },
         {
          "ITEM": 0,
          "RATING": 4.033124923706055,
          "USER": 4
         },
         {
          "ITEM": 1,
          "RATING": 3.9940121173858643,
          "USER": 2
         },
         {
          "ITEM": 4,
          "RATING": 2.267279624938965,
          "USER": 2
         },
         {
          "ITEM": 3,
          "RATING": 2.696197509765625,
          "USER": 2
         },
         {
          "ITEM": 2,
          "RATING": 2.8249964714050293,
          "USER": 2
         },
         {
          "ITEM": 5,
          "RATING": 2.3303346633911133,
          "USER": 2
         },
         {
          "ITEM": 0,
          "RATING": 3.3698816299438477,
          "USER": 2
         },
         {
          "ITEM": 1,
          "RATING": 3.859039306640625,
          "USER": 0
         },
         {
          "ITEM": 4,
          "RATING": 4.533079624176025,
          "USER": 0
         },
         {
          "ITEM": 3,
          "RATING": 4.419366836547852,
          "USER": 0
         },
         {
          "ITEM": 2,
          "RATING": 4.226285934448242,
          "USER": 0
         },
         {
          "ITEM": 5,
          "RATING": 4.416132926940918,
          "USER": 0
         },
         {
          "ITEM": 0,
          "RATING": 3.696733236312866,
          "USER": 0
         },
         {
          "ITEM": 1,
          "RATING": 4.084014892578125,
          "USER": 6
         },
         {
          "ITEM": 4,
          "RATING": 4.037803649902344,
          "USER": 6
         },
         {
          "ITEM": 3,
          "RATING": 4.052545070648193,
          "USER": 6
         },
         {
          "ITEM": 2,
          "RATING": 3.9062767028808594,
          "USER": 6
         },
         {
          "ITEM": 5,
          "RATING": 3.951486587524414,
          "USER": 6
         },
         {
          "ITEM": 0,
          "RATING": 3.726719617843628,
          "USER": 6
         },
         {
          "ITEM": 1,
          "RATING": 3.471980333328247,
          "USER": 3
         },
         {
          "ITEM": 4,
          "RATING": 2.9769163131713867,
          "USER": 3
         },
         {
          "ITEM": 3,
          "RATING": 3.137287139892578,
          "USER": 3
         },
         {
          "ITEM": 2,
          "RATING": 3.133159637451172,
          "USER": 3
         },
         {
          "ITEM": 5,
          "RATING": 2.868990421295166,
          "USER": 3
         },
         {
          "ITEM": 0,
          "RATING": 3.130906343460083,
          "USER": 3
         },
         {
          "ITEM": 1,
          "RATING": 3.757328510284424,
          "USER": 7
         },
         {
          "ITEM": 4,
          "RATING": 4.356379508972168,
          "USER": 7
         },
         {
          "ITEM": 3,
          "RATING": 4.252635955810547,
          "USER": 7
         },
         {
          "ITEM": 2,
          "RATING": 4.072903633117676,
          "USER": 7
         },
         {
          "ITEM": 5,
          "RATING": 4.21536922454834,
          "USER": 7
         },
         {
          "ITEM": 0,
          "RATING": 3.5927414894104004,
          "USER": 7
         },
         {
          "ITEM": 1,
          "RATING": 4.548954963684082,
          "USER": 5
         },
         {
          "ITEM": 4,
          "RATING": 3.9097957611083984,
          "USER": 5
         },
         {
          "ITEM": 3,
          "RATING": 4.076635360717773,
          "USER": 5
         },
         {
          "ITEM": 2,
          "RATING": 3.9456300735473633,
          "USER": 5
         },
         {
          "ITEM": 5,
          "RATING": 3.9495255947113037,
          "USER": 5
         },
         {
          "ITEM": 0,
          "RATING": 4.078114032745361,
          "USER": 5
         },
         {
          "ITEM": 1,
          "RATING": 4.136589527130127,
          "USER": 8
         },
         {
          "ITEM": 4,
          "RATING": 2.263800621032715,
          "USER": 8
         },
         {
          "ITEM": 3,
          "RATING": 2.7295868396759033,
          "USER": 8
         },
         {
          "ITEM": 2,
          "RATING": 2.854670524597168,
          "USER": 8
         },
         {
          "ITEM": 5,
          "RATING": 2.358680486679077,
          "USER": 8
         },
         {
          "ITEM": 0,
          "RATING": 3.472015619277954,
          "USER": 8
         },
         {
          "ITEM": 1,
          "RATING": 3.7849059104919434,
          "USER": 9
         },
         {
          "ITEM": 4,
          "RATING": 4.33188533782959,
          "USER": 9
         },
         {
          "ITEM": 3,
          "RATING": 4.232998847961426,
          "USER": 9
         },
         {
          "ITEM": 2,
          "RATING": 4.044402122497559,
          "USER": 9
         },
         {
          "ITEM": 5,
          "RATING": 4.196937561035156,
          "USER": 9
         },
         {
          "ITEM": 0,
          "RATING": 3.5819289684295654,
          "USER": 9
         }
        ]
       },
       "encoding": {
        "color": {
         "field": "RATING",
         "legend": {
          "direction": "horizontal",
          "orient": "bottom",
          "tickCount": 5,
          "titleOrient": "top"
         },
         "scale": {
          "nice": true,
          "scheme": "yellowgreenblue",
          "type": "bin-ordinal"
         },
         "type": "quantitative"
        },
        "x": {
         "axis": {
          "labelAngle": 0,
          "orient": "top"
         },
         "field": "ITEM",
         "type": "nominal"
        },
        "y": {
         "axis": {
          "orient": "left"
         },
         "field": "USER",
         "type": "nominal"
        }
       },
       "height": 300,
       "mark": "rect",
       "width": 180
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "InteractionVis(predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [],
   "source": [
    "def SimilarityVis(item_embedding, user_embedding):\n",
    "    \n",
    "    item_embedding_df_wide = pd.DataFrame(item_embedding)\n",
    "    user_embedding_df_wide = pd.DataFrame(user_embedding)\n",
    "\n",
    "    item_embedding_df_wide.reset_index(inplace=True)\n",
    "    item_embedding_df_wide[\"idx\"] = item_embedding_df_wide[\"index\"].apply(lambda x: \"I\" + str(x))\n",
    "    item_embedding_df_wide.columns = [\"index\", \"X0\", \"X1\", \"idx\" ]\n",
    "\n",
    "    user_embedding_df_wide.reset_index(inplace=True)\n",
    "    user_embedding_df_wide[\"idx\"] = user_embedding_df_wide[\"index\"].apply(lambda x: \"U\" + str(x))\n",
    "    user_embedding_df_wide.columns = [\"index\", \"X0\", \"X1\", \"idx\" ]\n",
    "\n",
    "    embedding_df_wide = pd.concat([item_embedding_df_wide, user_embedding_df_wide])\n",
    "    \n",
    "    base = alt.Chart(edf).encode(\n",
    "        alt.X(\"X0:Q\", axis = alt.Axis(bandPosition = 0.5)),\n",
    "        alt.Y(\"X1:Q\", axis = alt.Axis(bandPosition = 0.5))\n",
    "    )\n",
    "\n",
    "    vis = base.mark_point(size=0) + base.mark_text().encode(text=\"idx\")\n",
    "    \n",
    "    return vis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "name": "data-be73b32a1f94598140b7129142a8488d"
       },
       "datasets": {
        "data-be73b32a1f94598140b7129142a8488d": [
         {
          "X0": 0.42663830518722534,
          "X1": 0.43679916858673096,
          "idx": "I0",
          "index": 0
         },
         {
          "X0": 0.7709527015686035,
          "X1": 1.0180782079696655,
          "idx": "I1",
          "index": 1
         },
         {
          "X0": 0.49435946345329285,
          "X1": -0.3513801395893097,
          "idx": "I2",
          "index": 2
         },
         {
          "X0": 0.7281472682952881,
          "X1": -0.5116900205612183,
          "idx": "I3",
          "index": 3
         },
         {
          "X0": 0.8027671575546265,
          "X1": -0.9369111657142639,
          "idx": "I4",
          "index": 4
         },
         {
          "X0": 0.8810898661613464,
          "X1": -0.664902925491333,
          "idx": "I5",
          "index": 5
         },
         {
          "X0": 0.7634384036064148,
          "X1": -0.7496684789657593,
          "idx": "U0",
          "index": 0
         },
         {
          "X0": 0.577717125415802,
          "X1": 0.6026750206947327,
          "idx": "U1",
          "index": 1
         },
         {
          "X0": -0.24383129179477692,
          "X1": 0.750173032283783,
          "idx": "U2",
          "index": 2
         },
         {
          "X0": -0.14105083048343658,
          "X1": 0.05877910926938057,
          "idx": "U3",
          "index": 3
         },
         {
          "X0": 0.8581835627555847,
          "X1": 0.08425549417734146,
          "idx": "U4",
          "index": 4
         },
         {
          "X0": 0.8078038692474365,
          "X1": 0.3693584203720093,
          "idx": "U5",
          "index": 5
         },
         {
          "X0": 0.5184164047241211,
          "X1": -0.15546026825904846,
          "idx": "U6",
          "index": 6
         },
         {
          "X0": 0.5371776819229126,
          "X1": -0.6251449584960938,
          "idx": "U7",
          "index": 7
         },
         {
          "X0": -0.15340930223464966,
          "X1": 0.8556748032569885,
          "idx": "U8",
          "index": 8
         },
         {
          "X0": 0.5974195003509521,
          "X1": -0.5847427845001221,
          "idx": "U9",
          "index": 9
         }
        ]
       },
       "layer": [
        {
         "encoding": {
          "x": {
           "axis": {
            "bandPosition": 0.5
           },
           "field": "X0",
           "type": "quantitative"
          },
          "y": {
           "axis": {
            "bandPosition": 0.5
           },
           "field": "X1",
           "type": "quantitative"
          }
         },
         "mark": {
          "size": 0,
          "type": "point"
         }
        },
        {
         "encoding": {
          "text": {
           "field": "idx",
           "type": "nominal"
          },
          "x": {
           "axis": {
            "bandPosition": 0.5
           },
           "field": "X0",
           "type": "quantitative"
          },
          "y": {
           "axis": {
            "bandPosition": 0.5
           },
           "field": "X1",
           "type": "quantitative"
          }
         },
         "mark": "text"
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 320,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SimilarityVis(item_embedding, user_embedding)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "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>index</th>\n",
       "      <th>X0</th>\n",
       "      <th>X1</th>\n",
       "      <th>idx</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.426638</td>\n",
       "      <td>0.436799</td>\n",
       "      <td>I0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.770953</td>\n",
       "      <td>1.018078</td>\n",
       "      <td>I1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0.494359</td>\n",
       "      <td>-0.351380</td>\n",
       "      <td>I2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0.728147</td>\n",
       "      <td>-0.511690</td>\n",
       "      <td>I3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>0.802767</td>\n",
       "      <td>-0.936911</td>\n",
       "      <td>I4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index        X0        X1 idx\n",
       "0      0  0.426638  0.436799  I0\n",
       "1      1  0.770953  1.018078  I1\n",
       "2      2  0.494359 -0.351380  I2\n",
       "3      3  0.728147 -0.511690  I3\n",
       "4      4  0.802767 -0.936911  I4"
      ]
     },
     "execution_count": 299,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "edf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [],
   "source": [
    "base = alt.Chart(edf).encode(\n",
    "    alt.X(\"X0:Q\", axis = alt.Axis(bandPosition = 0.5)),\n",
    "    alt.Y(\"X1:Q\", axis = alt.Axis(bandPosition = 0.5))\n",
    ")\n",
    "\n",
    "vis = base.mark_point(size=1) + base.mark_text().encode(text=\"idx\")"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}