6_resultados.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "header = ['userId', 'movieId', 'rating', 'timestamp']\n",
    "ratings = pd.read_csv('./ml-100k/u.data', sep='\\t', names=header)\n",
    "\n",
    "u_unique = ratings.userId.unique()\n",
    "user2Idx = {o:i+1 for i,o in enumerate(u_unique)}\n",
    "\n",
    "m_unique = ratings.movieId.unique()\n",
    "movie2Idx = {o:i+1 for i,o in enumerate(m_unique)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "ratings.userId = ratings.userId.apply(lambda x: user2Idx[x])\n",
    "ratings.movieId = ratings.movieId.apply(lambda x: movie2Idx[x])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(80000, 20000)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_split = 20000\n",
    "ratings_train = ratings[n_split:]\n",
    "ratings_val = ratings[:n_split]\n",
    "len(ratings_train), len(ratings_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "943 1682 943 1650\n"
     ]
    }
   ],
   "source": [
    "n_users = int(ratings.userId.nunique())\n",
    "n_movies = int(ratings.movieId.nunique())\n",
    "n_users_train = int(ratings_train.userId.nunique())\n",
    "n_movies_train = int(ratings_train.movieId.nunique())\n",
    "print(n_users, n_movies, n_users_train, n_movies_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5, 1, 3.52835)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max_rating = ratings_train['rating'].max()\n",
    "min_rating = ratings_train['rating'].min()\n",
    "av_rating = ratings_train['rating'].mean()\n",
    "max_rating, min_rating, av_rating"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Definición RED"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.layers import Input, Embedding, Flatten, Dropout, Concatenate, Dense, Activation, Lambda\n",
    "from keras import Model\n",
    "from keras.optimizers import Adam"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "Item (InputLayer)               (None, 1)            0                                            \n",
      "__________________________________________________________________________________________________\n",
      "User (InputLayer)               (None, 1)            0                                            \n",
      "__________________________________________________________________________________________________\n",
      "Movie-Embedding (Embedding)     (None, 1, 8)         13464       Item[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "User-Embedding (Embedding)      (None, 1, 5)         4720        User[0][0]                       \n",
      "__________________________________________________________________________________________________\n",
      "FlattenMovies (Flatten)         (None, 8)            0           Movie-Embedding[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "FlattenUsers (Flatten)          (None, 5)            0           User-Embedding[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "Concat (Concatenate)            (None, 13)           0           FlattenMovies[0][0]              \n",
      "                                                                 FlattenUsers[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "FullyConnected-1 (Dense)        (None, 50)           700         Concat[0][0]                     \n",
      "__________________________________________________________________________________________________\n",
      "Activation (Dense)              (None, 1)            51          FullyConnected-1[0][0]           \n",
      "__________________________________________________________________________________________________\n",
      "lambda_1 (Lambda)               (None, 1)            0           Activation[0][0]                 \n",
      "==================================================================================================\n",
      "Total params: 18,935\n",
      "Trainable params: 18,935\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "n_latent_factors_user = 5\n",
    "n_latent_factors_movie = 8\n",
    "movie_input = Input(shape=[1],name='Item')\n",
    "movie_embedding = Embedding(n_movies + 1, n_latent_factors_movie, name='Movie-Embedding')(movie_input)\n",
    "movie_vec = Flatten(name='FlattenMovies')(movie_embedding)\n",
    "# movie_vec = Dropout(0.2)(movie_vec)\n",
    "\n",
    "\n",
    "user_input = Input(shape=[1],name='User')\n",
    "user_vec = Flatten(name='FlattenUsers')(Embedding(n_users + 1, \n",
    "                                                  n_latent_factors_user,name='User-Embedding')(user_input))\n",
    "# user_vec = Dropout(0.2)(user_vec)\n",
    "\n",
    "\n",
    "concat = Concatenate(name='Concat')([movie_vec, user_vec])\n",
    "# concat = Dropout(0.2)(concat)\n",
    "\n",
    "x = Dense(50,name='FullyConnected-1', activation='relu')(concat)\n",
    "#x = Dropout(0.5)(x)\n",
    "#x = Dense(50,name='FullyConnected-1', activation='relu')(concat)\n",
    "#x = Dropout(0.5)(x)\n",
    "\n",
    "\n",
    "## Se pueden sacar las siguientes dos lineas para no forzar a sigmoidea\n",
    "x = Dense(1, activation='sigmoid',name='Activation')(x)\n",
    "x = Lambda(lambda z: (max_rating - min_rating) * z + min_rating)(x)\n",
    "##\n",
    "\n",
    "model = Model([user_input, movie_input], x)\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras.backend as K \n",
    "def root_mean_squared_error(y_true, y_pred):\n",
    "    return K.sqrt(K.mean(K.square(y_pred - y_true))) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights('weights.hdf5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile('adam', loss= 'mean_squared_error', metrics=[root_mean_squared_error])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20000/20000 [==============================] - 0s 19us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.8864154481887817, 0.9323507642745972]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate([ratings_val.userId, ratings_val.movieId], ratings_val.rating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "80000/80000 [==============================] - 1s 13us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.7083571329832077, 0.8335200972318649]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate([ratings_train.userId, ratings_train.movieId], ratings_train.rating)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Obtengo embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "movie_embeddings_layer = model.layers[2]\n",
    "user_embeddings_layer = model.layers[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Movie-Embedding', 'User-Embedding')"
      ]
     },
     "execution_count": 180,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movie_embeddings_layer.name, user_embeddings_layer.name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1683, 8), (944, 5))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movie_embeddings_matrix = movie_embeddings_layer.get_weights()[0]\n",
    "user_embeddings_matrix = user_embeddings_layer.get_weights()[0]\n",
    "movie_embeddings_matrix.shape, user_embeddings_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.0014859 , -0.04829219,  0.0136039 ,  0.01360333,  0.03192383],\n",
       "       [-0.10639615,  0.0577417 ,  0.03024676,  0.06648126, -0.00818836],\n",
       "       [ 0.00644937,  0.05087709, -0.1267155 ,  0.07583879,  0.00484064],\n",
       "       ...,\n",
       "       [-0.04635098, -0.01193763, -0.03389605,  0.19602741, -0.05608499],\n",
       "       [ 0.10910267, -0.07419512,  0.03692925,  0.08315375,  0.02582754],\n",
       "       [ 0.07779792,  0.1508785 ,  0.00313076, -0.06561682,  0.11070072]],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_embeddings_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.14363556, -0.11988138,  0.15550709, -0.10945612,  0.04878137,\n",
       "        -0.07205785,  0.06160824, -0.01302086],\n",
       "       [-0.07785374, -0.11487766,  0.06631942, -0.01807967,  0.08839604,\n",
       "        -0.08385522,  0.1278895 , -0.16648261]], dtype=float32)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movie_embeddings_matrix[1:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Los puedo obtener definiendo un modelo nuevo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_test_emb = Model([movie_input], [movie_embedding])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[-0.12120856,  0.14663969,  0.19157135,  0.09968016,\n",
       "          0.14670104,  0.12186183,  0.04760471,  0.09871168]]],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 184,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_test_emb.predict([2])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Nearest Neighbors "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import NearestNeighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "nbrs = NearestNeighbors(n_neighbors=10, metric='cosine').fit(movie_embeddings_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "neighbors = nbrs.kneighbors([movie_embeddings_matrix[2]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[0.        , 0.03244066, 0.03463501, 0.03577471, 0.04553968,\n",
       "         0.04791397, 0.05075437, 0.05386627, 0.05777127, 0.05930173]],\n",
       "       dtype=float32),\n",
       " array([[   2,  688,  724, 1047,   46, 1272,  755,  242, 1553,  992]]))"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.06508597, -0.07309216,  0.05873515, -0.03105282,  0.04879123,\n",
       "       -0.04319927,  0.05170346, -0.08999793], dtype=float32)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movie_embeddings_matrix[688]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ordenar por ratings de peliculas no vistas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userId</th>\n",
       "      <th>movieId</th>\n",
       "      <th>rating</th>\n",
       "      <th>timestamp</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>881250949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>940</th>\n",
       "      <td>1</td>\n",
       "      <td>529</td>\n",
       "      <td>4</td>\n",
       "      <td>881251863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1133</th>\n",
       "      <td>1</td>\n",
       "      <td>378</td>\n",
       "      <td>4</td>\n",
       "      <td>881251728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1812</th>\n",
       "      <td>1</td>\n",
       "      <td>523</td>\n",
       "      <td>3</td>\n",
       "      <td>881251274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1896</th>\n",
       "      <td>1</td>\n",
       "      <td>432</td>\n",
       "      <td>5</td>\n",
       "      <td>881251793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2374</th>\n",
       "      <td>1</td>\n",
       "      <td>835</td>\n",
       "      <td>5</td>\n",
       "      <td>881252017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6910</th>\n",
       "      <td>1</td>\n",
       "      <td>381</td>\n",
       "      <td>4</td>\n",
       "      <td>881251021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7517</th>\n",
       "      <td>1</td>\n",
       "      <td>330</td>\n",
       "      <td>4</td>\n",
       "      <td>881251820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7842</th>\n",
       "      <td>1</td>\n",
       "      <td>551</td>\n",
       "      <td>5</td>\n",
       "      <td>881251911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10017</th>\n",
       "      <td>1</td>\n",
       "      <td>84</td>\n",
       "      <td>4</td>\n",
       "      <td>881251793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10254</th>\n",
       "      <td>1</td>\n",
       "      <td>633</td>\n",
       "      <td>2</td>\n",
       "      <td>881252056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10981</th>\n",
       "      <td>1</td>\n",
       "      <td>87</td>\n",
       "      <td>4</td>\n",
       "      <td>881251955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13733</th>\n",
       "      <td>1</td>\n",
       "      <td>290</td>\n",
       "      <td>5</td>\n",
       "      <td>881250949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14606</th>\n",
       "      <td>1</td>\n",
       "      <td>364</td>\n",
       "      <td>3</td>\n",
       "      <td>881252172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16834</th>\n",
       "      <td>1</td>\n",
       "      <td>439</td>\n",
       "      <td>5</td>\n",
       "      <td>881252017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17102</th>\n",
       "      <td>1</td>\n",
       "      <td>390</td>\n",
       "      <td>5</td>\n",
       "      <td>881251753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17830</th>\n",
       "      <td>1</td>\n",
       "      <td>650</td>\n",
       "      <td>4</td>\n",
       "      <td>881251702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18853</th>\n",
       "      <td>1</td>\n",
       "      <td>948</td>\n",
       "      <td>4</td>\n",
       "      <td>881252128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21605</th>\n",
       "      <td>1</td>\n",
       "      <td>424</td>\n",
       "      <td>3</td>\n",
       "      <td>881251842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22271</th>\n",
       "      <td>1</td>\n",
       "      <td>292</td>\n",
       "      <td>3</td>\n",
       "      <td>881251911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22773</th>\n",
       "      <td>1</td>\n",
       "      <td>11</td>\n",
       "      <td>2</td>\n",
       "      <td>881251577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23189</th>\n",
       "      <td>1</td>\n",
       "      <td>1007</td>\n",
       "      <td>4</td>\n",
       "      <td>881252110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24030</th>\n",
       "      <td>1</td>\n",
       "      <td>180</td>\n",
       "      <td>3</td>\n",
       "      <td>881251728</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25726</th>\n",
       "      <td>1</td>\n",
       "      <td>752</td>\n",
       "      <td>3</td>\n",
       "      <td>881251045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32721</th>\n",
       "      <td>1</td>\n",
       "      <td>488</td>\n",
       "      <td>3</td>\n",
       "      <td>881251884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33536</th>\n",
       "      <td>1</td>\n",
       "      <td>666</td>\n",
       "      <td>3</td>\n",
       "      <td>881251753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35197</th>\n",
       "      <td>1</td>\n",
       "      <td>93</td>\n",
       "      <td>4</td>\n",
       "      <td>881251843</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36281</th>\n",
       "      <td>1</td>\n",
       "      <td>513</td>\n",
       "      <td>5</td>\n",
       "      <td>881251753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41539</th>\n",
       "      <td>1</td>\n",
       "      <td>1046</td>\n",
       "      <td>3</td>\n",
       "      <td>881251642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42384</th>\n",
       "      <td>1</td>\n",
       "      <td>673</td>\n",
       "      <td>4</td>\n",
       "      <td>881251601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50147</th>\n",
       "      <td>1</td>\n",
       "      <td>657</td>\n",
       "      <td>4</td>\n",
       "      <td>881252090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52726</th>\n",
       "      <td>1</td>\n",
       "      <td>222</td>\n",
       "      <td>5</td>\n",
       "      <td>881251820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56628</th>\n",
       "      <td>1</td>\n",
       "      <td>433</td>\n",
       "      <td>2</td>\n",
       "      <td>881251955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59165</th>\n",
       "      <td>1</td>\n",
       "      <td>366</td>\n",
       "      <td>3</td>\n",
       "      <td>881251955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59607</th>\n",
       "      <td>1</td>\n",
       "      <td>322</td>\n",
       "      <td>2</td>\n",
       "      <td>881251820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60199</th>\n",
       "      <td>1</td>\n",
       "      <td>467</td>\n",
       "      <td>4</td>\n",
       "      <td>881251143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60706</th>\n",
       "      <td>1</td>\n",
       "      <td>303</td>\n",
       "      <td>4</td>\n",
       "      <td>881251954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78787</th>\n",
       "      <td>1</td>\n",
       "      <td>492</td>\n",
       "      <td>3</td>\n",
       "      <td>881250949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87863</th>\n",
       "      <td>1</td>\n",
       "      <td>522</td>\n",
       "      <td>1</td>\n",
       "      <td>881252305</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       userId  movieId  rating  timestamp\n",
       "0           1        1       3  881250949\n",
       "940         1      529       4  881251863\n",
       "1133        1      378       4  881251728\n",
       "1812        1      523       3  881251274\n",
       "1896        1      432       5  881251793\n",
       "2374        1      835       5  881252017\n",
       "6910        1      381       4  881251021\n",
       "7517        1      330       4  881251820\n",
       "7842        1      551       5  881251911\n",
       "10017       1       84       4  881251793\n",
       "10254       1      633       2  881252056\n",
       "10981       1       87       4  881251955\n",
       "13733       1      290       5  881250949\n",
       "14606       1      364       3  881252172\n",
       "16834       1      439       5  881252017\n",
       "17102       1      390       5  881251753\n",
       "17830       1      650       4  881251702\n",
       "18853       1      948       4  881252128\n",
       "21605       1      424       3  881251842\n",
       "22271       1      292       3  881251911\n",
       "22773       1       11       2  881251577\n",
       "23189       1     1007       4  881252110\n",
       "24030       1      180       3  881251728\n",
       "25726       1      752       3  881251045\n",
       "32721       1      488       3  881251884\n",
       "33536       1      666       3  881251753\n",
       "35197       1       93       4  881251843\n",
       "36281       1      513       5  881251753\n",
       "41539       1     1046       3  881251642\n",
       "42384       1      673       4  881251601\n",
       "50147       1      657       4  881252090\n",
       "52726       1      222       5  881251820\n",
       "56628       1      433       2  881251955\n",
       "59165       1      366       3  881251955\n",
       "59607       1      322       2  881251820\n",
       "60199       1      467       4  881251143\n",
       "60706       1      303       4  881251954\n",
       "78787       1      492       3  881250949\n",
       "87863       1      522       1  881252305"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Todas las peliculas que el usuario 1 califico\n",
    "\n",
    "ratings[ratings['userId'] == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[   1    2    3 ... 1680 1681 1682]\n",
      "[1 1 1 ... 1 1 1]\n"
     ]
    }
   ],
   "source": [
    "all_movie_idxs = np.linspace(1, ratings.movieId.max(), n_movies, dtype=int)\n",
    "print(all_movie_idxs)\n",
    "user_idxs = np.zeros(n_movies, dtype=int) + 1\n",
    "print(user_idxs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = model.predict([user_idxs,all_movie_idxs])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4.7978897, 1.0406588)"
      ]
     },
     "execution_count": 259,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions.max(), predictions.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[3.5254776],\n",
       "       [3.745766 ],\n",
       "       [2.0018735],\n",
       "       ...,\n",
       "       [3.0657752],\n",
       "       [3.2810183],\n",
       "       [3.0640361]], dtype=float32)"
      ]
     },
     "execution_count": 267,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1436, 1619, 1038, ..., 1173, 1626,  850])"
      ]
     },
     "execution_count": 269,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argsort(predictions[:,0])[::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4.7978897], dtype=float32)"
      ]
     },
     "execution_count": 270,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions[1436]"
   ]
  },
  {
   "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}