Time Series predictions with TF.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Time Series Forecasting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dans ce tutoriel, nous allons essayer de nous familiariser avec les concepts de bases liés à la prédiction des séries temporelles.\n",
    "On va survoler beaucoup de concepts qui à eux seuls, pris individuellement, pourraient constituer un atelier à part entière."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pré-requis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* quelques notions de Python\n",
    "* **pandas**: librairie super utile pour l'analyse de données en Python\n",
    "* **matplotlib**: outil de représentation de données (un peu velu, mais bon)\n",
    "* **numpy**: librairie de manipulation des données\n",
    "* **tensorflow**: parce que bon, on est là pour en parler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "\n",
    "mpl.rcParams['figure.figsize'] = (8, 6)\n",
    "mpl.rcParams['axes.grid'] = False"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous allons travailler avec les relevés environnementaux d'une localité donnée"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/jena_climate_2009_2016.csv.zip\n",
      "13574144/13568290 [==============================] - 10s 1us/step\n"
     ]
    }
   ],
   "source": [
    "zip_path = tf.keras.utils.get_file(\n",
    "    origin='https://storage.googleapis.com/tensorflow/tf-keras-datasets/jena_climate_2009_2016.csv.zip',\n",
    "    fname='jena_climate_2009_2016.csv.zip',\n",
    "    extract=True)\n",
    "csv_path, _ = os.path.splitext(zip_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/elia/.keras/datasets/jena_climate_2009_2016.csv.zip\n"
     ]
    }
   ],
   "source": [
    "print(zip_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(zip_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Regardons un peu à quoi ressemblent les données"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>Date Time</th>\n",
       "      <th>p (mbar)</th>\n",
       "      <th>T (degC)</th>\n",
       "      <th>Tpot (K)</th>\n",
       "      <th>Tdew (degC)</th>\n",
       "      <th>rh (%)</th>\n",
       "      <th>VPmax (mbar)</th>\n",
       "      <th>VPact (mbar)</th>\n",
       "      <th>VPdef (mbar)</th>\n",
       "      <th>sh (g/kg)</th>\n",
       "      <th>H2OC (mmol/mol)</th>\n",
       "      <th>rho (g/m**3)</th>\n",
       "      <th>wv (m/s)</th>\n",
       "      <th>max. wv (m/s)</th>\n",
       "      <th>wd (deg)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>01.01.2009 00:10:00</td>\n",
       "      <td>996.52</td>\n",
       "      <td>-8.02</td>\n",
       "      <td>265.40</td>\n",
       "      <td>-8.90</td>\n",
       "      <td>93.3</td>\n",
       "      <td>3.33</td>\n",
       "      <td>3.11</td>\n",
       "      <td>0.22</td>\n",
       "      <td>1.94</td>\n",
       "      <td>3.12</td>\n",
       "      <td>1307.75</td>\n",
       "      <td>1.03</td>\n",
       "      <td>1.75</td>\n",
       "      <td>152.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>01.01.2009 00:20:00</td>\n",
       "      <td>996.57</td>\n",
       "      <td>-8.41</td>\n",
       "      <td>265.01</td>\n",
       "      <td>-9.28</td>\n",
       "      <td>93.4</td>\n",
       "      <td>3.23</td>\n",
       "      <td>3.02</td>\n",
       "      <td>0.21</td>\n",
       "      <td>1.89</td>\n",
       "      <td>3.03</td>\n",
       "      <td>1309.80</td>\n",
       "      <td>0.72</td>\n",
       "      <td>1.50</td>\n",
       "      <td>136.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>01.01.2009 00:30:00</td>\n",
       "      <td>996.53</td>\n",
       "      <td>-8.51</td>\n",
       "      <td>264.91</td>\n",
       "      <td>-9.31</td>\n",
       "      <td>93.9</td>\n",
       "      <td>3.21</td>\n",
       "      <td>3.01</td>\n",
       "      <td>0.20</td>\n",
       "      <td>1.88</td>\n",
       "      <td>3.02</td>\n",
       "      <td>1310.24</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.63</td>\n",
       "      <td>171.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>01.01.2009 00:40:00</td>\n",
       "      <td>996.51</td>\n",
       "      <td>-8.31</td>\n",
       "      <td>265.12</td>\n",
       "      <td>-9.07</td>\n",
       "      <td>94.2</td>\n",
       "      <td>3.26</td>\n",
       "      <td>3.07</td>\n",
       "      <td>0.19</td>\n",
       "      <td>1.92</td>\n",
       "      <td>3.08</td>\n",
       "      <td>1309.19</td>\n",
       "      <td>0.34</td>\n",
       "      <td>0.50</td>\n",
       "      <td>198.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>01.01.2009 00:50:00</td>\n",
       "      <td>996.51</td>\n",
       "      <td>-8.27</td>\n",
       "      <td>265.15</td>\n",
       "      <td>-9.04</td>\n",
       "      <td>94.1</td>\n",
       "      <td>3.27</td>\n",
       "      <td>3.08</td>\n",
       "      <td>0.19</td>\n",
       "      <td>1.92</td>\n",
       "      <td>3.09</td>\n",
       "      <td>1309.00</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.63</td>\n",
       "      <td>214.3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Date Time  p (mbar)  T (degC)  Tpot (K)  Tdew (degC)  rh (%)  \\\n",
       "0  01.01.2009 00:10:00    996.52     -8.02    265.40        -8.90    93.3   \n",
       "1  01.01.2009 00:20:00    996.57     -8.41    265.01        -9.28    93.4   \n",
       "2  01.01.2009 00:30:00    996.53     -8.51    264.91        -9.31    93.9   \n",
       "3  01.01.2009 00:40:00    996.51     -8.31    265.12        -9.07    94.2   \n",
       "4  01.01.2009 00:50:00    996.51     -8.27    265.15        -9.04    94.1   \n",
       "\n",
       "   VPmax (mbar)  VPact (mbar)  VPdef (mbar)  sh (g/kg)  H2OC (mmol/mol)  \\\n",
       "0          3.33          3.11          0.22       1.94             3.12   \n",
       "1          3.23          3.02          0.21       1.89             3.03   \n",
       "2          3.21          3.01          0.20       1.88             3.02   \n",
       "3          3.26          3.07          0.19       1.92             3.08   \n",
       "4          3.27          3.08          0.19       1.92             3.09   \n",
       "\n",
       "   rho (g/m**3)  wv (m/s)  max. wv (m/s)  wd (deg)  \n",
       "0       1307.75      1.03           1.75     152.3  \n",
       "1       1309.80      0.72           1.50     136.1  \n",
       "2       1310.24      0.19           0.63     171.6  \n",
       "3       1309.19      0.34           0.50     198.0  \n",
       "4       1309.00      0.32           0.63     214.3  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(420551, 15)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Donc, nous disposons d'une série temporelle, dont les observations sont espacées de 10mn.\n",
    "Pour 1h on a donc 6 observations, pour une journée 144, ...\n",
    "\n",
    "Maintenant, disons que si nous nous positionnons à un certain moment sur notre ligne de temps, nous voudrions prédire les 6 prochaines heures, donc les 36 prochaines observations.\n",
    "Pour faire ce genre de prédiction, nous allons par exemple nous appuyer sur les 5 derniers jours, donc les 720 observations précédant le moment auquel nous voulons établir la prédiction.\n",
    "\n",
    "Le bout de code suivant nous aide précisément à récupérer la fenêtre de temps indispensable pour établir notre prédiction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data_window(dataset, start_index, end_index, history_size, target_size):\n",
    "    \"\"\"\n",
    "    Fonction retournant une fenêtre de données,\n",
    "    utilisées pour prédire l'avenir et les valeurs\n",
    "    correspondantes à prédire\n",
    "    \n",
    "    :start_index: et :end_index: délimitent le set de données\n",
    "    à utiliser\n",
    "    :history_size: indique la quantité de données antérieures\n",
    "    que nous voudrions considérer pour établir la prédiction\n",
    "    \n",
    "    :target_size: indique la 'quantité' de futur qu'on aimerait prédire\n",
    "    \"\"\"\n",
    "    data = []\n",
    "    labels = []\n",
    "\n",
    "    start_index = start_index + history_size\n",
    "    if end_index is None:\n",
    "        end_index = len(dataset) - target_size\n",
    "\n",
    "    for i in range(start_index, end_index):\n",
    "        indices = range(i-history_size, i)\n",
    "        data.append(np.reshape(dataset[indices], (history_size, 1)))\n",
    "        labels.append(dataset[i+target_size])\n",
    "    return np.array(data), np.array(labels)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dans la suite nous allons tenter une approche où nous utiliserons les 300000 premières observations comme ensemble de test, et le reste comme ensemble de validation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "TRAIN_DATA_SIZE = 300000"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Avant de continuer, notons que l'algorithme va générer des valeurs initiales aléatoires pour les paramètres à estimer. La génération de nombres aléatoires n'est pas vraiment aléatoire sur un ordinateur ! Et il est important de pouvoir reproduire nos tests. Pour cela, on paramètre *le seed* qui est un chiffre qui permet d'initialiser le générateur de chiffres aléatoires ! Sur cette base, l'ordinateur générera des nombres aléatoires mais on pourra reproduire et retrouver ces chiffres par la suite (en théorie). On pourra regénérer en effet exactement les mêmes séquences de chiffres aléatoires."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "tf.random.set_seed(13)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Allez, commençons les prédictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dans nos données, essayons d'extraire les températures pour faire des prédictions sur leurs valeurs futures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date Time\n",
       "01.01.2009 00:10:00   -8.02\n",
       "01.01.2009 00:20:00   -8.41\n",
       "01.01.2009 00:30:00   -8.51\n",
       "01.01.2009 00:40:00   -8.31\n",
       "01.01.2009 00:50:00   -8.27\n",
       "Name: T (degC), dtype: float64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = df['T (degC)']\n",
    "data.index = df['Date Time']\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([<matplotlib.axes._subplots.AxesSubplot object at 0x13e14a8d0>],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.plot(subplots=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Maintenant, nous allons normaliser les données de test: en ML la qualité des données est essentielle pour obtenir des modèles prédictifs performants. Pour éviter de traiter des données erronées et améliorer la performance du modèle, il faut impérativement analyser les données, détecter les anomalies le plus tôt possible et déterminer les étapes de prétraitement et de nettoyage appropriées.\n",
    "\n",
    "La normalisation, grosso modo, réduit les valeurs à un intervalle donné afin de les considérer toutes dans le même ordre de grandeur.\n",
    "\n",
    "Une des fonctions de normalisation fréquente est la normalisation par le test Z : mettre les données à l’échelle en fonction de la moyenne et de l’écart type en divisant la différence entre les données et la moyenne par l’écart type"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.values\n",
    "train_data_mean = data[:TRAIN_DATA_SIZE].mean()\n",
    "train_data_std = data[:TRAIN_DATA_SIZE].std()\n",
    "\n",
    "data = (data - train_data_mean) / train_data_std"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Créeons maintenant nos différents sets de données d'apprentissage et de prédiction.\n",
    "Mettons qu'à un instant t, nous allons utiliser les 20 dernières observations pour prédire la prochaine (t+1).\n",
    "\n",
    "Utilisons notre fonction de tout à l'heure:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "past = 20\n",
    "target = 0\n",
    "\n",
    "x_train, y_train = data_window(\n",
    "    data,\n",
    "    0,\n",
    "    TRAIN_DATA_SIZE,\n",
    "    past,\n",
    "    target\n",
    "    )\n",
    "x_validation, y_validation = data_window(\n",
    "    data,\n",
    "    TRAIN_DATA_SIZE,\n",
    "    None,\n",
    "    past,\n",
    "    target\n",
    "    )\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nous avons donc maintenant nos fenêtres de données, ainsi que les valeurs à prédire; nous en avons pour chacun de nos set de données: un d'entraînement et un de validation.\n",
    "Regardons un peu à quoi ça ressemble: les informations passées sont en bleu, et la prédiction est en rouge."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_time_steps(length):\n",
    "  time_steps = []\n",
    "  for i in range(-length, 0, 1):\n",
    "    time_steps.append(i)\n",
    "  return time_steps\n",
    "\n",
    "def show_plot(plot_data, delta, title):\n",
    "  labels = ['Historique', 'Vraie valeur', 'Valeur prédite']\n",
    "  marker = ['.-', 'rx', 'go']\n",
    "  time_steps = create_time_steps(plot_data[0].shape[0])\n",
    "  if delta:\n",
    "    future = delta\n",
    "  else:\n",
    "    future = 0\n",
    "\n",
    "  plt.title(title)\n",
    "  for i, x in enumerate(plot_data):\n",
    "    if i:\n",
    "      plt.plot(future, plot_data[i], marker[i], markersize=10,\n",
    "               label=labels[i])\n",
    "    else:\n",
    "      plt.plot(time_steps, plot_data[i].flatten(), marker[i], label=labels[i])\n",
    "  plt.legend()\n",
    "  plt.xlim([time_steps[0], (future+5)*2])\n",
    "  plt.xlabel('Time-Step')\n",
    "  return plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'matplotlib.pyplot' from '/Users/elia/Kode/gdg/.tensorflow/lib/python3.7/site-packages/matplotlib/pyplot.py'>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_plot([x_train[0], y_train[0]], 0, 'Un échantillon')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On peut se rendre compte que la prédiction basée sur la simple moyenne n'est pas très précise. Voyons si avec un réseau de neurones récurrents on peut mieux faire."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train for 200 steps, validate for 50 steps\n",
      "Epoch 1/10\n",
      "200/200 [==============================] - 6s 28ms/step - loss: 0.4075 - val_loss: 0.1351\n",
      "Epoch 2/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.1118 - val_loss: 0.0359\n",
      "Epoch 3/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0489 - val_loss: 0.0290\n",
      "Epoch 4/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0443 - val_loss: 0.0258\n",
      "Epoch 5/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0299 - val_loss: 0.0235\n",
      "Epoch 6/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0317 - val_loss: 0.0224\n",
      "Epoch 7/10\n",
      "200/200 [==============================] - 3s 15ms/step - loss: 0.0286 - val_loss: 0.0207\n",
      "Epoch 8/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0263 - val_loss: 0.0197\n",
      "Epoch 9/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0253 - val_loss: 0.0181\n",
      "Epoch 10/10\n",
      "200/200 [==============================] - 3s 16ms/step - loss: 0.0227 - val_loss: 0.0174\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x13b599210>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BATCH_SIZE = 256\n",
    "BUFFER_SIZE = 10000\n",
    "\n",
    "training = tf.data.Dataset.from_tensor_slices((x_train, y_train))\n",
    "training = training.cache().shuffle(BUFFER_SIZE).batch(BATCH_SIZE).repeat()\n",
    "\n",
    "validation = tf.data.Dataset.from_tensor_slices((x_validation, y_validation))\n",
    "validation = validation.batch(BATCH_SIZE).repeat()\n",
    "\n",
    "\n",
    "simple_lstm_model = tf.keras.models.Sequential([\n",
    "    tf.keras.layers.LSTM(8, input_shape=x_train.shape[-2:]),\n",
    "    tf.keras.layers.Dense(1)\n",
    "])\n",
    "\n",
    "simple_lstm_model.compile(optimizer='adam', loss='mae')\n",
    "\n",
    "simple_lstm_model.fit(\n",
    "    training,\n",
    "    epochs=10,\n",
    "    steps_per_epoch=200,\n",
    "    validation_data=validation,\n",
    "    validation_steps=50\n",
    "    )\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for x, y in validation.take(4):\n",
    "  plot = show_plot([x[0].numpy(), y[0].numpy(),\n",
    "                    simple_lstm_model.predict(x)[0]], 0, 'Simple LSTM model')\n",
    "  plot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![True ML](https://imgs.xkcd.com/comics/machine_learning.png \"True ML\")"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}