Plot_LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras import applications\n",
    "from keras.models import Sequential\n",
    "from keras.models import Model\n",
    "from keras.layers import Dropout, Flatten, Dense, Activation, Reshape, LeakyReLU\n",
    "from keras.callbacks import CSVLogger\n",
    "import tensorflow as tf\n",
    "from scipy.ndimage import imread\n",
    "import numpy as np\n",
    "import random\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Conv1D, MaxPooling1D\n",
    "from keras import backend as K\n",
    "import keras\n",
    "from keras.callbacks import CSVLogger, ModelCheckpoint\n",
    "from keras.backend.tensorflow_backend import set_session\n",
    "from keras import optimizers\n",
    "import h5py\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "import os\n",
    "import pandas as pd\n",
    "# import matplotlib\n",
    "import h5py\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import \n",
    "with h5py.File(''.join(['bitcoin2015to2017_close.h5']), 'r') as hf:\n",
    "    datas = hf['inputs'].value\n",
    "    labels = hf['outputs'].value\n",
    "    input_times = hf['input_times'].value\n",
    "    output_times = hf['output_times'].value\n",
    "    original_inputs = hf['original_inputs'].value\n",
    "    original_outputs = hf['original_outputs'].value\n",
    "    original_datas = hf['original_datas'].value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "os.environ['CUDA_DEVICE_ORDER'] = 'PCI_BUS_ID'\n",
    "os.environ['CUDA_VISIBLE_DEVICES'] = '0'\n",
    "os.environ['TF_CPP_MIN_LOG_LEVEL']='2'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth = True\n",
    "set_session(tf.Session(config=config))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "scaler=MinMaxScaler()\n",
    "#split training validation\n",
    "training_size = int(0.8* datas.shape[0])\n",
    "training_datas = datas[:training_size,:,:]\n",
    "training_labels = labels[:training_size,:,0]\n",
    "validation_datas = datas[training_size:,:,:]\n",
    "validation_labels = labels[training_size:,:,0]\n",
    "validation_original_outputs = original_outputs[training_size:,:,:]\n",
    "validation_original_inputs = original_inputs[training_size:,:,:]\n",
    "validation_input_times = input_times[training_size:,:,:]\n",
    "validation_output_times = output_times[training_size:,:,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(221, 272, 1)"
      ]
     },
     "execution_count": 252,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ground_true = np.append(validation_original_inputs,validation_original_outputs, axis=1)\n",
    "ground_true.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(221, 272, 1)"
      ]
     },
     "execution_count": 253,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ground_true_times = np.append(validation_input_times,validation_output_times, axis=1)\n",
    "ground_true_times.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "step_size = datas.shape[1]\n",
    "batch_size= 8\n",
    "nb_features = datas.shape[2]\n",
    "epochs = 1\n",
    "output_size=16\n",
    "units= 50\n",
    "second_units=30"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(LSTM(units=units,activation='tanh', input_shape=(step_size,nb_features),return_sequences=False))\n",
    "#model.add(LSTM(units=second_units,activation='relu',return_sequences=False))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(output_size))\n",
    "model.add(LeakyReLU())\n",
    "model.load_weights('weights/bitcoin2015to2017_close_LSTM_1_tanh_leaky_-44-0.00004.hdf5')\n",
    "model.compile(loss='mse', optimizer='adam')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(300628, 1)"
      ]
     },
     "execution_count": 256,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "original_datas.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 221, 16)\n",
      "(3536,)\n"
     ]
    }
   ],
   "source": [
    "predicted = model.predict(validation_datas)\n",
    "predicted_inverted = []\n",
    "\n",
    "# In[7]:\n",
    "# we only care about the 0 axis, close price data\n",
    "\n",
    "scaler.fit(original_datas[:,0].reshape(-1,1))\n",
    "predicted_inverted.append(scaler.inverse_transform(predicted))\n",
    "print np.array(predicted_inverted).shape\n",
    "#get only the close data\n",
    "ground_true = ground_true[:,:,0].reshape(-1)\n",
    "ground_true_times = ground_true_times.reshape(-1)\n",
    "ground_true_times = pd.to_datetime(ground_true_times, unit='s')\n",
    "# since we are appending in the first dimension\n",
    "predicted_inverted = np.array(predicted_inverted)[0,:,:].reshape(-1)\n",
    "print np.array(predicted_inverted).shape\n",
    "validation_output_times = pd.to_datetime(validation_output_times.reshape(-1), unit='s')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60112, 2)"
      ]
     },
     "execution_count": 258,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ground_true_df = pd.DataFrame()\n",
    "ground_true_df['times'] = ground_true_times\n",
    "ground_true_df['value'] = ground_true\n",
    "ground_true_df.set_index('times').reset_index()\n",
    "ground_true_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3536, 2)"
      ]
     },
     "execution_count": 259,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction_df = pd.DataFrame()\n",
    "prediction_df['times'] = validation_output_times\n",
    "prediction_df['value'] = predicted_inverted\n",
    "prediction_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>times</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-06-04 13:50:00</td>\n",
       "      <td>2489.749268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-06-04 13:55:00</td>\n",
       "      <td>2478.548584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-06-04 14:00:00</td>\n",
       "      <td>2465.910400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-06-04 14:05:00</td>\n",
       "      <td>2467.365967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-06-04 14:10:00</td>\n",
       "      <td>2480.132812</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                times        value\n",
       "0 2017-06-04 13:50:00  2489.749268\n",
       "1 2017-06-04 13:55:00  2478.548584\n",
       "2 2017-06-04 14:00:00  2465.910400\n",
       "3 2017-06-04 14:05:00  2467.365967\n",
       "4 2017-06-04 14:10:00  2480.132812"
      ]
     },
     "execution_count": 260,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "prediction_df = prediction_df.loc[(prediction_df[\"times\"].dt.year == 2017 )&(prediction_df[\"times\"].dt.month > 7 ),: ]\n",
    "ground_true_df = ground_true_df.loc[(ground_true_df[\"times\"].dt.year == 2017 )&(ground_true_df[\"times\"].dt.month > 7 ),:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJQAAAJCCAYAAACWHZ1NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcbc49e75d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.plot(ground_true_df.times,ground_true_df.value, label = 'Actual')\n",
    "# plt.plot(prediction_df.times,prediction_df.value, label = 'Predicted')\n",
    "plt.plot(prediction_df.times,prediction_df.value,'ro', label='Predicted')\n",
    "plt.legend(loc='upper left')\n",
    "plt.savefig('result/bitcoin2015to2017_close_LSTM_1_tanh_leaky_result.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15363.599112060345"
      ]
     },
     "execution_count": 263,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "mean_squared_error(validation_original_outputs[:,:,0].reshape(-1),predicted_inverted)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}