notebooks/5.1 Multivariate time-series to images.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib as mpl\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "import dask.dataframe as dd\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers.core import Dense, Activation, Dropout, Flatten\n",
    "from keras.layers import TimeDistributed\n",
    "from keras.layers.recurrent import LSTM\n",
    "from keras.layers import Dense, Conv1D, MaxPool2D, Flatten, Dropout, GRU, CuDNNLSTM\n",
    "from keras.callbacks import EarlyStopping, TensorBoard, ModelCheckpoint\n",
    "from keras.optimizers import Adam, SGD, Nadam\n",
    "from time import time\n",
    "from livelossplot import PlotLossesKeras\n",
    "from keras.layers.advanced_activations import LeakyReLU, PReLU\n",
    "import tensorflow as tf\n",
    "from keras.utils.training_utils import multi_gpu_model\n",
    "from tensorflow.python.client import device_lib\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from keijzer import *\n",
    "\n",
    "%matplotlib inline\n",
    "%config InlineBackend.print_figure_kwargs={'facecolor' : \"w\"} # Make sure the axis background of plots is white, this is usefull for the black theme in JupyterLab\n",
    "sns.set()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'D:\\\\Brian\\\\JupyterLab\\\\Multivariate-time-series-models-in-Keras\\\\notebooks'"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%pwd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Select the data to use"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = _\n",
    "path = path[:-10] # removes '\\\\notebooks' from the path string\n",
    "\n",
    "df = pd.read_csv(path+\"\\\\data\\\\house_data_processed.csv\", delimiter='\\t', parse_dates=['datetime'])\n",
    "df = df.set_index(['datetime']) \n",
    "\n",
    "magnitude = 1 # Take this from the 1. EDA & Feauture engineering notebook. It's the factor where gasPower has been scaled with to the power 10."
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "columns_to_category = ['hour', 'dayofweek', 'season']\n",
    "data[columns_to_category] = data[columns_to_category].astype('category') # change datetypes to category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>ePower</th>\n",
       "      <th>FF</th>\n",
       "      <th>RG</th>\n",
       "      <th>T</th>\n",
       "      <th>gasPower</th>\n",
       "      <th>gasPower_std</th>\n",
       "      <th>hour</th>\n",
       "      <th>dayofweek</th>\n",
       "      <th>season</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datetime</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-02-28 12:00:00</th>\n",
       "      <td>308.321701</td>\n",
       "      <td>9.067500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.183333</td>\n",
       "      <td>6.115723</td>\n",
       "      <td>0.085626</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-28 13:00:00</th>\n",
       "      <td>263.656273</td>\n",
       "      <td>8.684999</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.300000</td>\n",
       "      <td>5.238647</td>\n",
       "      <td>0.318250</td>\n",
       "      <td>13</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-28 14:00:00</th>\n",
       "      <td>181.913345</td>\n",
       "      <td>8.296667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.283333</td>\n",
       "      <td>2.451172</td>\n",
       "      <td>0.041087</td>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-28 15:00:00</th>\n",
       "      <td>378.613129</td>\n",
       "      <td>6.918334</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.516667</td>\n",
       "      <td>3.065186</td>\n",
       "      <td>0.214238</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-02-28 16:00:00</th>\n",
       "      <td>332.392082</td>\n",
       "      <td>6.583333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.433333</td>\n",
       "      <td>8.120117</td>\n",
       "      <td>0.014156</td>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         ePower        FF   RG         T  gasPower  \\\n",
       "datetime                                                             \n",
       "2017-02-28 12:00:00  308.321701  9.067500  0.0  5.183333  6.115723   \n",
       "2017-02-28 13:00:00  263.656273  8.684999  0.0  5.300000  5.238647   \n",
       "2017-02-28 14:00:00  181.913345  8.296667  0.0  5.283333  2.451172   \n",
       "2017-02-28 15:00:00  378.613129  6.918334  0.0  5.516667  3.065186   \n",
       "2017-02-28 16:00:00  332.392082  6.583333  0.0  5.433333  8.120117   \n",
       "\n",
       "                     gasPower_std  hour  dayofweek  season  \n",
       "datetime                                                    \n",
       "2017-02-28 12:00:00      0.085626    12          1       1  \n",
       "2017-02-28 13:00:00      0.318250    13          1       1  \n",
       "2017-02-28 14:00:00      0.041087    14          1       1  \n",
       "2017-02-28 15:00:00      0.214238    15          1       1  \n",
       "2017-02-28 16:00:00      0.014156    16          1       1  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = df.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# To categorical"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice how we do not one-hot encode the categorical variables this time.  \n",
    "By experiment this proved to give a higher accuracy.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns_to_category = ['hour', 'dayofweek', 'season']\n",
    "data[columns_to_category] = data[columns_to_category].astype('category') # change datetypes to category"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# To RNN format\n",
    "Or better said, convert the data to smaller images. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\program files\\python36\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype object was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "c:\\program files\\python36\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype object was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "c:\\program files\\python36\\lib\\site-packages\\sklearn\\utils\\validation.py:595: DataConversionWarning: Data with input dtype object was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((4326, 120, 8), (1786, 120, 8))"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "look_back = 5*24 # D -> 5, H -> 5*24\n",
    "num_features = data.shape[1] - 1\n",
    "output_dim = 1\n",
    "train_size = 0.7 # This acctually is the train size, oops.\n",
    "target_column = 'gasPower'\n",
    "\n",
    "X_train, y_train, X_test, y_test = df_to_cnn_rnn_format(df=data, train_size=train_size, look_back=look_back, target_column=target_column, scale_X=True)\n",
    "X_train.shape, X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a test dataframe, just to check the distribution of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "t = pd.DataFrame(X_train[0, :, :])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>120.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.491970</td>\n",
       "      <td>0.269442</td>\n",
       "      <td>0.051576</td>\n",
       "      <td>-1.364024</td>\n",
       "      <td>0.958272</td>\n",
       "      <td>0.000456</td>\n",
       "      <td>-0.079342</td>\n",
       "      <td>-1.065323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.325411</td>\n",
       "      <td>0.970611</td>\n",
       "      <td>0.779587</td>\n",
       "      <td>0.359802</td>\n",
       "      <td>1.324700</td>\n",
       "      <td>1.005724</td>\n",
       "      <td>1.176478</td>\n",
       "      <td>0.365995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-1.311545</td>\n",
       "      <td>-1.494454</td>\n",
       "      <td>-0.169610</td>\n",
       "      <td>-2.220182</td>\n",
       "      <td>-0.858663</td>\n",
       "      <td>-1.663402</td>\n",
       "      <td>-1.499534</td>\n",
       "      <td>-2.653999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.511567</td>\n",
       "      <td>-0.526232</td>\n",
       "      <td>-0.169610</td>\n",
       "      <td>-1.634634</td>\n",
       "      <td>0.039015</td>\n",
       "      <td>-0.831473</td>\n",
       "      <td>-1.001221</td>\n",
       "      <td>-0.981709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.100158</td>\n",
       "      <td>0.267104</td>\n",
       "      <td>-0.169610</td>\n",
       "      <td>-1.430338</td>\n",
       "      <td>0.739244</td>\n",
       "      <td>0.000456</td>\n",
       "      <td>-0.502908</td>\n",
       "      <td>-0.981709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.875840</td>\n",
       "      <td>0.792881</td>\n",
       "      <td>-0.169610</td>\n",
       "      <td>-1.052125</td>\n",
       "      <td>1.707386</td>\n",
       "      <td>0.832384</td>\n",
       "      <td>0.992031</td>\n",
       "      <td>-0.981709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>3.969831</td>\n",
       "      <td>2.576487</td>\n",
       "      <td>4.844544</td>\n",
       "      <td>-0.549359</td>\n",
       "      <td>6.729274</td>\n",
       "      <td>1.664313</td>\n",
       "      <td>1.490343</td>\n",
       "      <td>-0.981709</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                0           1           2           3           4           5  \\\n",
       "count  120.000000  120.000000  120.000000  120.000000  120.000000  120.000000   \n",
       "mean     0.491970    0.269442    0.051576   -1.364024    0.958272    0.000456   \n",
       "std      1.325411    0.970611    0.779587    0.359802    1.324700    1.005724   \n",
       "min     -1.311545   -1.494454   -0.169610   -2.220182   -0.858663   -1.663402   \n",
       "25%     -0.511567   -0.526232   -0.169610   -1.634634    0.039015   -0.831473   \n",
       "50%      0.100158    0.267104   -0.169610   -1.430338    0.739244    0.000456   \n",
       "75%      0.875840    0.792881   -0.169610   -1.052125    1.707386    0.832384   \n",
       "max      3.969831    2.576487    4.844544   -0.549359    6.729274    1.664313   \n",
       "\n",
       "                6           7  \n",
       "count  120.000000  120.000000  \n",
       "mean    -0.079342   -1.065323  \n",
       "std      1.176478    0.365995  \n",
       "min     -1.499534   -2.653999  \n",
       "25%     -1.001221   -0.981709  \n",
       "50%     -0.502908   -0.981709  \n",
       "75%      0.992031   -0.981709  \n",
       "max      1.490343   -0.981709  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t.describe()"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "sns.pairplot(pd.DataFrame(X_train[0, :, :]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# These are the \"images\" to work with"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "samples = [0, 500, 1000, 1500, 2000]\n",
    "\n",
    "#plt.figure(figsize=(25,25))\n",
    "\n",
    "for i, sample in enumerate(samples):\n",
    "    i += 1\n",
    "    plt.subplot(1,len(samples),i)\n",
    "    plt.imshow(X_train[sample])\n",
    "\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1800 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "samples = [2500, 3000, 3500, 4000, 4300]\n",
    "\n",
    "plt.figure(figsize=(25,25))\n",
    "\n",
    "for i, sample in enumerate(samples):\n",
    "    i += 1\n",
    "    plt.subplot(1,len(samples),i)\n",
    "    plt.imshow(X_train[sample])\n",
    "\n",
    "\n",
    "plt.tight_layout()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make the image wider"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10, 5)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = X_train[0][:10, :5]\n",
    "a.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x29b9c36c978>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(a)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 50)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x29b9c36a2e8>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 'stretch' the columns out by 10 pixels each. \n",
    "# i.e. copy the columns 10 times and add them next to each other. \n",
    "q = np.repeat(a[:,:], 10, axis=1)\n",
    "print(q.shape)\n",
    "plt.imshow(q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(50, 50)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x29b9c2aca90>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "s = np.repeat(q[:,:], 5, axis=0)\n",
    "print(s.shape)\n",
    "plt.imshow(s)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Apply this to the original 120 by 14 arrays"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transform_image(array, heigth, width):\n",
    "    \"\"\"\n",
    "    Array = numpy array\n",
    "    Heigth is the amount of pisels that a single pixel will be stretched out with in the height dimension.\n",
    "    Width is the amount of pixels that a single pixel will be stretched out with in the width dimension.\n",
    "    \"\"\"\n",
    "    array = array.copy()\n",
    "    #print('Original shape: ', array.shape)\n",
    "    array = np.repeat(array[:,:], width, axis=1)\n",
    "    array = np.repeat(array[:,:], heigth, axis=0)\n",
    "    #print('Transformed shape: ', array.shape)\n",
    "    \n",
    "    return array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x29b9c258080>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Test the function. Plot the first 50 rows and columns.\n",
    "plt.imshow(transform_image(X_train[1000, :, :], heigth=5, width=10)[:50, :50]) # Take"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# What's left now is applying the above transform to all ... amount of 'images' from the dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DONE\n"
     ]
    }
   ],
   "source": [
    "reshaped = []\n",
    "for i in range(X_train.shape[0]):\n",
    "    reshaped.append(transform_image(X_train[i, :, :], heigth=5, width=10))\n",
    "    \n",
    "print('DONE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4326, 600, 80)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(reshaped).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 4326 images of (600,80), where 600 is the height and width 80.  \n",
    "In the actual CNN notebooks this 'stretched out' image is not used.  \n",
    "They might be used for transfer learning however, to get something close to the 256 by 256 size which most ConvNets require.  "
   ]
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}