5-13-2018 - Generic sine.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting a generic sinusoidal waveform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/bipe26194/miniconda3/envs/tensorflow/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n",
      "Using TensorFlow backend.\n",
      "/home/bipe26194/miniconda3/envs/tensorflow/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: compiletime version 3.5 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.6\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import keras\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Activation, Dropout, LSTM, CuDNNLSTM, Bidirectional\n",
    "\n",
    "import random\n",
    "from random import seed, random\n",
    "from sklearn.utils import shuffle\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of generic samples: 200\n"
     ]
    }
   ],
   "source": [
    "# np.random.seed(1337)  # for reproducibility\n",
    "\n",
    "# Simulate data\n",
    "cycle = 20\n",
    "step = 0.05\n",
    "noise = 0.00 # standard deviation from 0\n",
    "decaying_coef = 0.003\n",
    "start_coef = 1\n",
    "stop_coef = 3\n",
    "step_coef = 0.01\n",
    "\n",
    "data = []\n",
    "\n",
    "for coeff in np.arange(start_coef, stop_coef, step_coef):\n",
    "    end_pt = 2 * np.pi * cycle\n",
    "    t = np.arange(0, end_pt, step)\n",
    "    a = np.sin(t)\n",
    "    a_noisy = a + np.random.normal(0, noise, a.size)\n",
    "    a_decaying = a_noisy * np.exp(-coeff * decaying_coef * t)\n",
    "    data.append(a_decaying)\n",
    "\n",
    "print(\"Number of generic samples:\", len(data))\n",
    "# Note: If noise is not modulated by the exponential function, it will overwhelm the signal near the end of the data sequence\n",
    "# which is where test data is picked from. We can modulate it in simulated environment but in experiments, we may have to discard\n",
    "# data near the end of the motion."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f31033d9048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize data\n",
    "n_plot = int(a.size)\n",
    "\n",
    "_, axarr = plt.subplots(3, sharex=True, sharey=True)\n",
    "axarr[0].plot(a[:n_plot])\n",
    "axarr[0].set_title('Population')\n",
    "axarr[1].plot(a_decaying[:n_plot])\n",
    "axarr[1].set_title('Observation')\n",
    "axarr[2].plot(a[:n_plot], label=\"Population\")\n",
    "axarr[2].plot(a_decaying[:n_plot], label=\"Observation\")\n",
    "axarr[2].set_title('Population and Observation')\n",
    "axarr[2].legend()\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# hyperparameters\n",
    "DATA_DIM = 1\n",
    "N_PREV = 50 # number of previous steps used to predict the next step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _load_data(data, n_prev=N_PREV, is_shuffled=False):\n",
    "    docX, docY = [], []\n",
    "    for series in data:\n",
    "        for j in range(len(series)-n_prev):\n",
    "            docX.append(series[j:j+n_prev])\n",
    "            docY.append(series[j+n_prev])   \n",
    "    alsX = np.array(docX)\n",
    "    alsY = np.array(docY)\n",
    "    if is_shuffled: \n",
    "        alsX, alsY = shuffle(alsX, alsY)\n",
    "\n",
    "    return alsX.reshape(alsX.shape[0], alsX.shape[1], DATA_DIM), alsY.reshape(alsY.shape[0], DATA_DIM)\n",
    "\n",
    "def train_test_split(data):  \n",
    "    \"\"\"\n",
    "    This just splits data to training and testing parts\n",
    "    \"\"\"\n",
    "    test_id = np.random.randint(0, len(data))\n",
    "    data_train = data[:test_id] + data[test_id+1:]\n",
    "    np.random.shuffle(data_train)\n",
    "    X_train, y_train = _load_data(data_train, is_shuffled=True)\n",
    "    X_test, y_test = _load_data([data[test_id]], is_shuffled=False)\n",
    "    \n",
    "    return (X_train, y_train), (X_test, y_test)\n",
    "\n",
    "# retrieve data\n",
    "(X_train, y_train), (X_test, y_test) = train_test_split(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# define model structure\n",
    "model = Sequential()\n",
    "model.add(Bidirectional(CuDNNLSTM(units=100, return_sequences=True), input_shape=(None, DATA_DIM)))\n",
    "model.add(Bidirectional(CuDNNLSTM(units=50, return_sequences=False)))\n",
    "# model.add(Bidirectional(CuDNNLSTM(units=20, return_sequences=False)))\n",
    "# model.add(Bidirectional(CuDNNLSTM(units=10, return_sequences=False)))\n",
    "# model.add(Dense(units=10, activation=\"linear\"))\n",
    "model.add(Dense(units=1, activation=\"linear\"))\n",
    "model.compile(loss=\"mean_squared_error\", optimizer=\"adam\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#n_prev=50, units=10, epochs=50: decent accuracy half the waveform"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 392268 samples, validate on 98068 samples\n",
      "Epoch 1/10\n",
      "  5312/392268 [..............................] - ETA: 4:52 - loss: 0.0113"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-395ddad2f2d5>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0;31m# batch_size should be appropriate to your memory size\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;31m# number of epochs should be higher for real world problems\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0mhistory\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m64\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalidation_split\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras/models.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)\u001b[0m\n\u001b[1;32m   1000\u001b[0m                               \u001b[0minitial_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitial_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1001\u001b[0m                               \u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1002\u001b[0;31m                               validation_steps=validation_steps)\n\u001b[0m\u001b[1;32m   1003\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1004\u001b[0m     def evaluate(self, x=None, y=None,\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)\u001b[0m\n\u001b[1;32m   1703\u001b[0m                               \u001b[0minitial_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitial_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1704\u001b[0m                               \u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1705\u001b[0;31m                               validation_steps=validation_steps)\n\u001b[0m\u001b[1;32m   1706\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1707\u001b[0m     def evaluate(self, x=None, y=None,\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras/engine/training.py\u001b[0m in \u001b[0;36m_fit_loop\u001b[0;34m(self, f, ins, out_labels, batch_size, epochs, verbose, callbacks, val_f, val_ins, shuffle, callback_metrics, initial_epoch, steps_per_epoch, validation_steps)\u001b[0m\n\u001b[1;32m   1234\u001b[0m                         \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtoarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1235\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1236\u001b[0;31m                     \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1237\u001b[0m                     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1238\u001b[0m                         \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   2480\u001b[0m         \u001b[0msession\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_session\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2481\u001b[0m         updated = session.run(fetches=fetches, feed_dict=feed_dict,\n\u001b[0;32m-> 2482\u001b[0;31m                               **self.session_kwargs)\n\u001b[0m\u001b[1;32m   2483\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mupdated\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2484\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36mrun\u001b[0;34m(self, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m    887\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    888\u001b[0m       result = self._run(None, fetches, feed_dict, options_ptr,\n\u001b[0;32m--> 889\u001b[0;31m                          run_metadata_ptr)\n\u001b[0m\u001b[1;32m    890\u001b[0m       \u001b[0;32mif\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    891\u001b[0m         \u001b[0mproto_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf_session\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTF_GetBuffer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrun_metadata_ptr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_run\u001b[0;34m(self, handle, fetches, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m   1118\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfinal_fetches\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mfinal_targets\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mhandle\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mfeed_dict_tensor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1119\u001b[0m       results = self._do_run(handle, final_targets, final_fetches,\n\u001b[0;32m-> 1120\u001b[0;31m                              feed_dict_tensor, options, run_metadata)\n\u001b[0m\u001b[1;32m   1121\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1122\u001b[0m       \u001b[0mresults\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_run\u001b[0;34m(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)\u001b[0m\n\u001b[1;32m   1315\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mhandle\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1316\u001b[0m       return self._do_call(_run_fn, self._session, feeds, fetches, targets,\n\u001b[0;32m-> 1317\u001b[0;31m                            options, run_metadata)\n\u001b[0m\u001b[1;32m   1318\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1319\u001b[0m       \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_do_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0m_prun_fn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_session\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeeds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetches\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_do_call\u001b[0;34m(self, fn, *args)\u001b[0m\n\u001b[1;32m   1321\u001b[0m   \u001b[0;32mdef\u001b[0m \u001b[0m_do_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1322\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1323\u001b[0;31m       \u001b[0;32mreturn\u001b[0m \u001b[0mfn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1324\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0merrors\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mOpError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1325\u001b[0m       \u001b[0mmessage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcompat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mas_text\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmessage\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/miniconda3/envs/tensorflow/lib/python3.6/site-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m_run_fn\u001b[0;34m(session, feed_dict, fetch_list, target_list, options, run_metadata)\u001b[0m\n\u001b[1;32m   1300\u001b[0m           return tf_session.TF_Run(session, options,\n\u001b[1;32m   1301\u001b[0m                                    \u001b[0mfeed_dict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetch_list\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtarget_list\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1302\u001b[0;31m                                    status, run_metadata)\n\u001b[0m\u001b[1;32m   1303\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1304\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_prun_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msession\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeed_dict\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfetch_list\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "## and now train the model\n",
    "# batch_size should be appropriate to your memory size\n",
    "# number of epochs should be higher for real world problems\n",
    "history = model.fit(X_train, y_train, batch_size=64, epochs=10, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(history.history.keys())\n",
    "print(len(history.history['val_loss']))\n",
    "# summarize history for loss\n",
    "plt.plot(history.history['loss'][-20:])\n",
    "plt.plot(history.history['val_loss'][-20:])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'test'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# evaluate model fit\n",
    "score = model.evaluate(X_test, y_test)\n",
    "print('Test score:', score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# perform approximation\n",
    "test_prediction = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# perform modelling\n",
    "X_model = X_test[0]\n",
    "modelled_data = []\n",
    "dims = 1, X_model.shape[0], X_model.shape[1]\n",
    "for i in range(len(y_test)):\n",
    "    prev_steps = X_model.reshape(*dims)\n",
    "    next_step = model.predict(prev_steps)\n",
    "    modelled_data.append(list(next_step[0]))\n",
    "    X_model = np.append(X_model[1:], next_step)\n",
    "# evaluate\n",
    "if len(modelled_data)==len(y_test):\n",
    "    print(np.mean((modelled_data-y_test)**2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(y_test[:], label='test_data')\n",
    "plt.plot(modelled_data[:], label='test_model')\n",
    "plt.title('Test Observation and Prediction')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(y_test[:], label='test_data')\n",
    "plt.plot(test_prediction[:], label='test_pred')\n",
    "plt.title('Test Observation and Prediction')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_prediction = model.predict(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(y_train[:], label='train_data')\n",
    "plt.plot(train_prediction[:], label='train_pred')\n",
    "plt.title('Train Observation and Prediction')\n",
    "plt.legend()"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}