Keras_LTSM_Index_SP500.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import numpy as np\n",
    "import datetime\n",
    "import itertools\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import library as mio\n",
    "from sklearn.metrics import classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# reduced list only the most correlated\n",
    "indices_list_Complete = [\"^GSPC\",\"SPY\",\"^IXIC\", \"^DJI\", \"^GDAXI\", \"^FTSE\",\"^FCHI\", \"^N225\",\"^HSI\", \"^AXJO\",\"ORB\", \"EUR\",\"AUD\",\"GBP\",\"JPY\", \"SILVER\", \"GOLD\", \"WT1010\"] # reduced list only the most correlated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "start_date = \"2003-01-01\" # Start day of Series\n",
    "end_date = \"2017-01-01\"\t  # Final day of series\n",
    "dates = pd.date_range(start_date, end_date)  # date range as index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_index = mio.get_data(indices_list_Complete, dates)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "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>^GSPC</th>\n",
       "      <th>SPY</th>\n",
       "      <th>^IXIC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GDAXI</th>\n",
       "      <th>^FTSE</th>\n",
       "      <th>^FCHI</th>\n",
       "      <th>^N225</th>\n",
       "      <th>^HSI</th>\n",
       "      <th>^AXJO</th>\n",
       "      <th>ORB</th>\n",
       "      <th>EUR</th>\n",
       "      <th>AUD</th>\n",
       "      <th>GBP</th>\n",
       "      <th>JPY</th>\n",
       "      <th>SILVER</th>\n",
       "      <th>GOLD</th>\n",
       "      <th>WT1010</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-12-23</th>\n",
       "      <td>2263.790039</td>\n",
       "      <td>224.732324</td>\n",
       "      <td>5462.689941</td>\n",
       "      <td>19933.810547</td>\n",
       "      <td>11449.929688</td>\n",
       "      <td>7068.200195</td>\n",
       "      <td>4839.680176</td>\n",
       "      <td>19403.060547</td>\n",
       "      <td>21574.759766</td>\n",
       "      <td>5627.899902</td>\n",
       "      <td>51.74</td>\n",
       "      <td>0.957179</td>\n",
       "      <td>1.394655</td>\n",
       "      <td>0.814821</td>\n",
       "      <td>117.185667</td>\n",
       "      <td>15.74</td>\n",
       "      <td>1131.00</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-27</th>\n",
       "      <td>2268.879883</td>\n",
       "      <td>225.289896</td>\n",
       "      <td>5487.439941</td>\n",
       "      <td>19945.039062</td>\n",
       "      <td>11472.240234</td>\n",
       "      <td>7106.100098</td>\n",
       "      <td>4848.279785</td>\n",
       "      <td>19403.060547</td>\n",
       "      <td>21754.740234</td>\n",
       "      <td>5685.000000</td>\n",
       "      <td>52.25</td>\n",
       "      <td>0.956361</td>\n",
       "      <td>1.391288</td>\n",
       "      <td>0.815262</td>\n",
       "      <td>117.457200</td>\n",
       "      <td>15.85</td>\n",
       "      <td>1139.75</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-28</th>\n",
       "      <td>2249.919922</td>\n",
       "      <td>223.427985</td>\n",
       "      <td>5438.560059</td>\n",
       "      <td>19833.679688</td>\n",
       "      <td>11474.990234</td>\n",
       "      <td>7106.100098</td>\n",
       "      <td>4848.009766</td>\n",
       "      <td>19401.720703</td>\n",
       "      <td>21754.740234</td>\n",
       "      <td>5685.000000</td>\n",
       "      <td>53.39</td>\n",
       "      <td>0.963017</td>\n",
       "      <td>1.393879</td>\n",
       "      <td>0.817868</td>\n",
       "      <td>117.543337</td>\n",
       "      <td>15.85</td>\n",
       "      <td>1139.75</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-29</th>\n",
       "      <td>2249.260010</td>\n",
       "      <td>223.378214</td>\n",
       "      <td>5432.089844</td>\n",
       "      <td>19819.779297</td>\n",
       "      <td>11451.049805</td>\n",
       "      <td>7120.299805</td>\n",
       "      <td>4838.470215</td>\n",
       "      <td>19145.140625</td>\n",
       "      <td>21790.910156</td>\n",
       "      <td>5699.100098</td>\n",
       "      <td>53.46</td>\n",
       "      <td>0.954902</td>\n",
       "      <td>1.387659</td>\n",
       "      <td>0.817942</td>\n",
       "      <td>116.505533</td>\n",
       "      <td>16.06</td>\n",
       "      <td>1146.80</td>\n",
       "      <td>0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-12-30</th>\n",
       "      <td>2238.830078</td>\n",
       "      <td>222.561759</td>\n",
       "      <td>5383.120117</td>\n",
       "      <td>19762.599609</td>\n",
       "      <td>11481.059570</td>\n",
       "      <td>7142.799805</td>\n",
       "      <td>4862.310059</td>\n",
       "      <td>19114.369141</td>\n",
       "      <td>22000.560547</td>\n",
       "      <td>5665.799805</td>\n",
       "      <td>53.30</td>\n",
       "      <td>0.948191</td>\n",
       "      <td>1.385629</td>\n",
       "      <td>0.810056</td>\n",
       "      <td>116.637500</td>\n",
       "      <td>16.24</td>\n",
       "      <td>1159.10</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  ^GSPC         SPY        ^IXIC          ^DJI        ^GDAXI  \\\n",
       "2016-12-23  2263.790039  224.732324  5462.689941  19933.810547  11449.929688   \n",
       "2016-12-27  2268.879883  225.289896  5487.439941  19945.039062  11472.240234   \n",
       "2016-12-28  2249.919922  223.427985  5438.560059  19833.679688  11474.990234   \n",
       "2016-12-29  2249.260010  223.378214  5432.089844  19819.779297  11451.049805   \n",
       "2016-12-30  2238.830078  222.561759  5383.120117  19762.599609  11481.059570   \n",
       "\n",
       "                  ^FTSE        ^FCHI         ^N225          ^HSI        ^AXJO  \\\n",
       "2016-12-23  7068.200195  4839.680176  19403.060547  21574.759766  5627.899902   \n",
       "2016-12-27  7106.100098  4848.279785  19403.060547  21754.740234  5685.000000   \n",
       "2016-12-28  7106.100098  4848.009766  19401.720703  21754.740234  5685.000000   \n",
       "2016-12-29  7120.299805  4838.470215  19145.140625  21790.910156  5699.100098   \n",
       "2016-12-30  7142.799805  4862.310059  19114.369141  22000.560547  5665.799805   \n",
       "\n",
       "              ORB       EUR       AUD       GBP         JPY  SILVER     GOLD  \\\n",
       "2016-12-23  51.74  0.957179  1.394655  0.814821  117.185667   15.74  1131.00   \n",
       "2016-12-27  52.25  0.956361  1.391288  0.815262  117.457200   15.85  1139.75   \n",
       "2016-12-28  53.39  0.963017  1.393879  0.817868  117.543337   15.85  1139.75   \n",
       "2016-12-29  53.46  0.954902  1.387659  0.817942  116.505533   16.06  1146.80   \n",
       "2016-12-30  53.30  0.948191  1.385629  0.810056  116.637500   16.24  1159.10   \n",
       "\n",
       "            WT1010  \n",
       "2016-12-23    0.25  \n",
       "2016-12-27    0.21  \n",
       "2016-12-28    0.19  \n",
       "2016-12-29    0.17  \n",
       "2016-12-30    0.19  "
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_index.fillna(method='bfill', inplace=True)\n",
    "df_index.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "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>^GSPC</th>\n",
       "      <th>SPY</th>\n",
       "      <th>^IXIC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GDAXI</th>\n",
       "      <th>^FTSE</th>\n",
       "      <th>^FCHI</th>\n",
       "      <th>^N225</th>\n",
       "      <th>^HSI</th>\n",
       "      <th>^AXJO</th>\n",
       "      <th>ORB</th>\n",
       "      <th>EUR</th>\n",
       "      <th>AUD</th>\n",
       "      <th>GBP</th>\n",
       "      <th>JPY</th>\n",
       "      <th>SILVER</th>\n",
       "      <th>GOLD</th>\n",
       "      <th>WT1010</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "      <td>3525.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1412.595838</td>\n",
       "      <td>124.247170</td>\n",
       "      <td>2887.485241</td>\n",
       "      <td>12676.749688</td>\n",
       "      <td>6792.847443</td>\n",
       "      <td>5650.849134</td>\n",
       "      <td>4130.124744</td>\n",
       "      <td>12944.451194</td>\n",
       "      <td>19392.780442</td>\n",
       "      <td>4716.324737</td>\n",
       "      <td>70.492837</td>\n",
       "      <td>0.784911</td>\n",
       "      <td>1.218466</td>\n",
       "      <td>0.605921</td>\n",
       "      <td>103.269039</td>\n",
       "      <td>16.829385</td>\n",
       "      <td>993.989504</td>\n",
       "      <td>2.667827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>377.738299</td>\n",
       "      <td>42.716304</td>\n",
       "      <td>1111.738679</td>\n",
       "      <td>3019.752104</td>\n",
       "      <td>2287.728860</td>\n",
       "      <td>879.576705</td>\n",
       "      <td>772.017335</td>\n",
       "      <td>3478.880031</td>\n",
       "      <td>4632.425945</td>\n",
       "      <td>855.962899</td>\n",
       "      <td>29.223703</td>\n",
       "      <td>0.073974</td>\n",
       "      <td>0.180926</td>\n",
       "      <td>0.066218</td>\n",
       "      <td>13.772426</td>\n",
       "      <td>8.778454</td>\n",
       "      <td>427.903735</td>\n",
       "      <td>1.383519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>676.530029</td>\n",
       "      <td>57.440272</td>\n",
       "      <td>1268.640015</td>\n",
       "      <td>6547.049805</td>\n",
       "      <td>2202.959961</td>\n",
       "      <td>3287.000000</td>\n",
       "      <td>2403.040039</td>\n",
       "      <td>7054.979980</td>\n",
       "      <td>8409.009766</td>\n",
       "      <td>2700.399902</td>\n",
       "      <td>22.480000</td>\n",
       "      <td>0.625360</td>\n",
       "      <td>0.906698</td>\n",
       "      <td>0.474330</td>\n",
       "      <td>75.751979</td>\n",
       "      <td>4.370000</td>\n",
       "      <td>319.750000</td>\n",
       "      <td>-0.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1139.810059</td>\n",
       "      <td>93.528227</td>\n",
       "      <td>2094.139893</td>\n",
       "      <td>10444.370117</td>\n",
       "      <td>4998.160156</td>\n",
       "      <td>5053.200195</td>\n",
       "      <td>3580.479980</td>\n",
       "      <td>9844.589844</td>\n",
       "      <td>15413.429688</td>\n",
       "      <td>4169.500000</td>\n",
       "      <td>45.250000</td>\n",
       "      <td>0.734626</td>\n",
       "      <td>1.071435</td>\n",
       "      <td>0.550249</td>\n",
       "      <td>92.727524</td>\n",
       "      <td>10.770000</td>\n",
       "      <td>611.250000</td>\n",
       "      <td>1.520000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1313.010010</td>\n",
       "      <td>112.113904</td>\n",
       "      <td>2480.330078</td>\n",
       "      <td>12105.549805</td>\n",
       "      <td>6534.970215</td>\n",
       "      <td>5808.799805</td>\n",
       "      <td>4038.699951</td>\n",
       "      <td>11944.299805</td>\n",
       "      <td>20672.390625</td>\n",
       "      <td>4807.399902</td>\n",
       "      <td>66.340000</td>\n",
       "      <td>0.772966</td>\n",
       "      <td>1.237935</td>\n",
       "      <td>0.616478</td>\n",
       "      <td>105.553327</td>\n",
       "      <td>15.860000</td>\n",
       "      <td>1081.100000</td>\n",
       "      <td>3.150000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1650.469971</td>\n",
       "      <td>152.589273</td>\n",
       "      <td>3496.429932</td>\n",
       "      <td>15112.190430</td>\n",
       "      <td>8260.480469</td>\n",
       "      <td>6364.700195</td>\n",
       "      <td>4582.830078</td>\n",
       "      <td>16024.849609</td>\n",
       "      <td>22835.820312</td>\n",
       "      <td>5342.399902</td>\n",
       "      <td>101.890000</td>\n",
       "      <td>0.829636</td>\n",
       "      <td>1.341054</td>\n",
       "      <td>0.643910</td>\n",
       "      <td>115.820209</td>\n",
       "      <td>20.020000</td>\n",
       "      <td>1308.250000</td>\n",
       "      <td>3.910000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2271.719971</td>\n",
       "      <td>225.444643</td>\n",
       "      <td>5487.439941</td>\n",
       "      <td>19974.619141</td>\n",
       "      <td>12374.730469</td>\n",
       "      <td>7142.799805</td>\n",
       "      <td>6168.149902</td>\n",
       "      <td>20868.029297</td>\n",
       "      <td>31638.220703</td>\n",
       "      <td>6828.700195</td>\n",
       "      <td>140.730000</td>\n",
       "      <td>0.963904</td>\n",
       "      <td>1.776110</td>\n",
       "      <td>0.822470</td>\n",
       "      <td>125.630900</td>\n",
       "      <td>48.700000</td>\n",
       "      <td>1891.000000</td>\n",
       "      <td>4.690000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             ^GSPC          SPY        ^IXIC          ^DJI        ^GDAXI  \\\n",
       "count  3525.000000  3525.000000  3525.000000   3525.000000   3525.000000   \n",
       "mean   1412.595838   124.247170  2887.485241  12676.749688   6792.847443   \n",
       "std     377.738299    42.716304  1111.738679   3019.752104   2287.728860   \n",
       "min     676.530029    57.440272  1268.640015   6547.049805   2202.959961   \n",
       "25%    1139.810059    93.528227  2094.139893  10444.370117   4998.160156   \n",
       "50%    1313.010010   112.113904  2480.330078  12105.549805   6534.970215   \n",
       "75%    1650.469971   152.589273  3496.429932  15112.190430   8260.480469   \n",
       "max    2271.719971   225.444643  5487.439941  19974.619141  12374.730469   \n",
       "\n",
       "             ^FTSE        ^FCHI         ^N225          ^HSI        ^AXJO  \\\n",
       "count  3525.000000  3525.000000   3525.000000   3525.000000  3525.000000   \n",
       "mean   5650.849134  4130.124744  12944.451194  19392.780442  4716.324737   \n",
       "std     879.576705   772.017335   3478.880031   4632.425945   855.962899   \n",
       "min    3287.000000  2403.040039   7054.979980   8409.009766  2700.399902   \n",
       "25%    5053.200195  3580.479980   9844.589844  15413.429688  4169.500000   \n",
       "50%    5808.799805  4038.699951  11944.299805  20672.390625  4807.399902   \n",
       "75%    6364.700195  4582.830078  16024.849609  22835.820312  5342.399902   \n",
       "max    7142.799805  6168.149902  20868.029297  31638.220703  6828.700195   \n",
       "\n",
       "               ORB          EUR          AUD          GBP          JPY  \\\n",
       "count  3525.000000  3525.000000  3525.000000  3525.000000  3525.000000   \n",
       "mean     70.492837     0.784911     1.218466     0.605921   103.269039   \n",
       "std      29.223703     0.073974     0.180926     0.066218    13.772426   \n",
       "min      22.480000     0.625360     0.906698     0.474330    75.751979   \n",
       "25%      45.250000     0.734626     1.071435     0.550249    92.727524   \n",
       "50%      66.340000     0.772966     1.237935     0.616478   105.553327   \n",
       "75%     101.890000     0.829636     1.341054     0.643910   115.820209   \n",
       "max     140.730000     0.963904     1.776110     0.822470   125.630900   \n",
       "\n",
       "            SILVER         GOLD       WT1010  \n",
       "count  3525.000000  3525.000000  3525.000000  \n",
       "mean     16.829385   993.989504     2.667827  \n",
       "std       8.778454   427.903735     1.383519  \n",
       "min       4.370000   319.750000    -0.200000  \n",
       "25%      10.770000   611.250000     1.520000  \n",
       "50%      15.860000  1081.100000     3.150000  \n",
       "75%      20.020000  1308.250000     3.910000  \n",
       "max      48.700000  1891.000000     4.690000  "
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_index.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import myltsm as ltsm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "diff_values=pd.DataFrame()\n",
    "for t in indices_list_Complete:\n",
    "    diff_values[t] = ltsm.difference(df_index[t].values, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>^GSPC</th>\n",
       "      <th>SPY</th>\n",
       "      <th>^IXIC</th>\n",
       "      <th>^DJI</th>\n",
       "      <th>^GDAXI</th>\n",
       "      <th>^FTSE</th>\n",
       "      <th>^FCHI</th>\n",
       "      <th>^N225</th>\n",
       "      <th>^HSI</th>\n",
       "      <th>^AXJO</th>\n",
       "      <th>ORB</th>\n",
       "      <th>EUR</th>\n",
       "      <th>AUD</th>\n",
       "      <th>GBP</th>\n",
       "      <th>JPY</th>\n",
       "      <th>SILVER</th>\n",
       "      <th>GOLD</th>\n",
       "      <th>WT1010</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3.524000e+03</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "      <td>3524.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.377355</td>\n",
       "      <td>0.043691</td>\n",
       "      <td>1.134583</td>\n",
       "      <td>3.165460</td>\n",
       "      <td>2.376850</td>\n",
       "      <td>0.889132</td>\n",
       "      <td>0.473124</td>\n",
       "      <td>2.951487</td>\n",
       "      <td>3.585426</td>\n",
       "      <td>0.748638</td>\n",
       "      <td>0.006598</td>\n",
       "      <td>-0.000004</td>\n",
       "      <td>-0.000111</td>\n",
       "      <td>5.224858e-05</td>\n",
       "      <td>-0.000842</td>\n",
       "      <td>0.003263</td>\n",
       "      <td>0.231810</td>\n",
       "      <td>-0.001144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>14.714290</td>\n",
       "      <td>1.287708</td>\n",
       "      <td>33.936318</td>\n",
       "      <td>125.334113</td>\n",
       "      <td>92.736873</td>\n",
       "      <td>61.096317</td>\n",
       "      <td>54.774266</td>\n",
       "      <td>195.152334</td>\n",
       "      <td>287.877963</td>\n",
       "      <td>50.423728</td>\n",
       "      <td>1.093099</td>\n",
       "      <td>0.004657</td>\n",
       "      <td>0.010250</td>\n",
       "      <td>3.645091e-03</td>\n",
       "      <td>0.581142</td>\n",
       "      <td>0.487125</td>\n",
       "      <td>12.864238</td>\n",
       "      <td>0.045934</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-106.849976</td>\n",
       "      <td>-8.223334</td>\n",
       "      <td>-202.060059</td>\n",
       "      <td>-777.679688</td>\n",
       "      <td>-699.870117</td>\n",
       "      <td>-391.099610</td>\n",
       "      <td>-368.770020</td>\n",
       "      <td>-1720.140625</td>\n",
       "      <td>-3444.238282</td>\n",
       "      <td>-560.500000</td>\n",
       "      <td>-8.400000</td>\n",
       "      <td>-0.029391</td>\n",
       "      <td>-0.077819</td>\n",
       "      <td>-2.233300e-02</td>\n",
       "      <td>-3.512111</td>\n",
       "      <td>-6.680000</td>\n",
       "      <td>-133.500000</td>\n",
       "      <td>-0.260000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-6.239990</td>\n",
       "      <td>-0.512002</td>\n",
       "      <td>-15.062561</td>\n",
       "      <td>-52.393066</td>\n",
       "      <td>-39.349854</td>\n",
       "      <td>-28.899902</td>\n",
       "      <td>-27.219971</td>\n",
       "      <td>-81.107910</td>\n",
       "      <td>-108.000488</td>\n",
       "      <td>-22.100098</td>\n",
       "      <td>-0.560000</td>\n",
       "      <td>-0.002556</td>\n",
       "      <td>-0.005244</td>\n",
       "      <td>-1.852250e-03</td>\n",
       "      <td>-0.303033</td>\n",
       "      <td>-0.140000</td>\n",
       "      <td>-5.000000</td>\n",
       "      <td>-0.030000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.950012</td>\n",
       "      <td>0.084206</td>\n",
       "      <td>2.565185</td>\n",
       "      <td>6.150391</td>\n",
       "      <td>5.384888</td>\n",
       "      <td>1.700196</td>\n",
       "      <td>1.434937</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.600098</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>-0.000093</td>\n",
       "      <td>-0.000500</td>\n",
       "      <td>-1.000000e-06</td>\n",
       "      <td>0.002914</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.699982</td>\n",
       "      <td>0.666775</td>\n",
       "      <td>19.082642</td>\n",
       "      <td>65.329590</td>\n",
       "      <td>47.277465</td>\n",
       "      <td>32.500000</td>\n",
       "      <td>30.105102</td>\n",
       "      <td>101.142578</td>\n",
       "      <td>131.565430</td>\n",
       "      <td>26.699829</td>\n",
       "      <td>0.612500</td>\n",
       "      <td>0.002550</td>\n",
       "      <td>0.004511</td>\n",
       "      <td>1.826500e-03</td>\n",
       "      <td>0.327468</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>5.900000</td>\n",
       "      <td>0.020000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>104.130005</td>\n",
       "      <td>10.749727</td>\n",
       "      <td>194.739990</td>\n",
       "      <td>936.419922</td>\n",
       "      <td>518.140136</td>\n",
       "      <td>431.299805</td>\n",
       "      <td>367.010010</td>\n",
       "      <td>1883.589843</td>\n",
       "      <td>2332.539063</td>\n",
       "      <td>279.899903</td>\n",
       "      <td>7.340000</td>\n",
       "      <td>0.023450</td>\n",
       "      <td>0.124285</td>\n",
       "      <td>5.573000e-02</td>\n",
       "      <td>2.900577</td>\n",
       "      <td>5.340000</td>\n",
       "      <td>80.750000</td>\n",
       "      <td>0.240000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             ^GSPC          SPY        ^IXIC         ^DJI       ^GDAXI  \\\n",
       "count  3524.000000  3524.000000  3524.000000  3524.000000  3524.000000   \n",
       "mean      0.377355     0.043691     1.134583     3.165460     2.376850   \n",
       "std      14.714290     1.287708    33.936318   125.334113    92.736873   \n",
       "min    -106.849976    -8.223334  -202.060059  -777.679688  -699.870117   \n",
       "25%      -6.239990    -0.512002   -15.062561   -52.393066   -39.349854   \n",
       "50%       0.950012     0.084206     2.565185     6.150391     5.384888   \n",
       "75%       7.699982     0.666775    19.082642    65.329590    47.277465   \n",
       "max     104.130005    10.749727   194.739990   936.419922   518.140136   \n",
       "\n",
       "             ^FTSE        ^FCHI        ^N225         ^HSI        ^AXJO  \\\n",
       "count  3524.000000  3524.000000  3524.000000  3524.000000  3524.000000   \n",
       "mean      0.889132     0.473124     2.951487     3.585426     0.748638   \n",
       "std      61.096317    54.774266   195.152334   287.877963    50.423728   \n",
       "min    -391.099610  -368.770020 -1720.140625 -3444.238282  -560.500000   \n",
       "25%     -28.899902   -27.219971   -81.107910  -108.000488   -22.100098   \n",
       "50%       1.700196     1.434937     0.000000     0.000000     1.600098   \n",
       "75%      32.500000    30.105102   101.142578   131.565430    26.699829   \n",
       "max     431.299805   367.010010  1883.589843  2332.539063   279.899903   \n",
       "\n",
       "               ORB          EUR          AUD           GBP          JPY  \\\n",
       "count  3524.000000  3524.000000  3524.000000  3.524000e+03  3524.000000   \n",
       "mean      0.006598    -0.000004    -0.000111  5.224858e-05    -0.000842   \n",
       "std       1.093099     0.004657     0.010250  3.645091e-03     0.581142   \n",
       "min      -8.400000    -0.029391    -0.077819 -2.233300e-02    -3.512111   \n",
       "25%      -0.560000    -0.002556    -0.005244 -1.852250e-03    -0.303033   \n",
       "50%       0.030000    -0.000093    -0.000500 -1.000000e-06     0.002914   \n",
       "75%       0.612500     0.002550     0.004511  1.826500e-03     0.327468   \n",
       "max       7.340000     0.023450     0.124285  5.573000e-02     2.900577   \n",
       "\n",
       "            SILVER         GOLD       WT1010  \n",
       "count  3524.000000  3524.000000  3524.000000  \n",
       "mean      0.003263     0.231810    -0.001144  \n",
       "std       0.487125    12.864238     0.045934  \n",
       "min      -6.680000  -133.500000    -0.260000  \n",
       "25%      -0.140000    -5.000000    -0.030000  \n",
       "50%       0.000000     0.250000     0.000000  \n",
       "75%       0.170000     5.900000     0.020000  \n",
       "max       5.340000    80.750000     0.240000  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "diff_values.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "supervised_values = df_index.values\n",
    "supervised_values = supervised_values.astype('float32')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "scaled = scaler.fit_transform(supervised_values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-1.33329713, -1.30300069, -1.3518002 , ..., -1.37736118,\n",
       "        -1.52343142,  1.12206113],\n",
       "       [-1.33446217, -1.29806268, -1.34979415, ..., -1.37223423,\n",
       "        -1.51922429,  1.18712199],\n",
       "       [-1.28039587, -1.26966941, -1.31899118, ..., -1.35799277,\n",
       "        -1.49094296,  1.21603775],\n",
       "       ..., \n",
       "       [ 2.21699214,  2.3221786 ,  2.29499674, ..., -0.11158279,\n",
       "         0.34068686, -1.79121351],\n",
       "       [ 2.21524477,  2.32101345,  2.28917599, ..., -0.08765729,\n",
       "         0.35716498, -1.80567145],\n",
       "       [ 2.18762946,  2.30189705,  2.24512196, ..., -0.06714959,\n",
       "         0.38591367, -1.79121351]], dtype=float32)"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scaled"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "reframed = ltsm.series_to_supervised(scaled, 1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['var1(t-1)', 'var2(t-1)', 'var3(t-1)', 'var4(t-1)', 'var5(t-1)',\n",
       "       'var6(t-1)', 'var7(t-1)', 'var8(t-1)', 'var9(t-1)', 'var10(t-1)',\n",
       "       'var11(t-1)', 'var12(t-1)', 'var13(t-1)', 'var14(t-1)', 'var15(t-1)',\n",
       "       'var16(t-1)', 'var17(t-1)', 'var18(t-1)', 'var1(t)', 'var2(t)',\n",
       "       'var3(t)', 'var4(t)', 'var5(t)', 'var6(t)', 'var7(t)', 'var8(t)',\n",
       "       'var9(t)', 'var10(t)', 'var11(t)', 'var12(t)', 'var13(t)', 'var14(t)',\n",
       "       'var15(t)', 'var16(t)', 'var17(t)', 'var18(t)'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reframed.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "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>var1(t-1)</th>\n",
       "      <th>var2(t-1)</th>\n",
       "      <th>var3(t-1)</th>\n",
       "      <th>var4(t-1)</th>\n",
       "      <th>var5(t-1)</th>\n",
       "      <th>var6(t-1)</th>\n",
       "      <th>var7(t-1)</th>\n",
       "      <th>var8(t-1)</th>\n",
       "      <th>var9(t-1)</th>\n",
       "      <th>var10(t-1)</th>\n",
       "      <th>...</th>\n",
       "      <th>var9(t)</th>\n",
       "      <th>var10(t)</th>\n",
       "      <th>var11(t)</th>\n",
       "      <th>var12(t)</th>\n",
       "      <th>var13(t)</th>\n",
       "      <th>var14(t)</th>\n",
       "      <th>var15(t)</th>\n",
       "      <th>var16(t)</th>\n",
       "      <th>var17(t)</th>\n",
       "      <th>var18(t)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3520</th>\n",
       "      <td>2.246223</td>\n",
       "      <td>2.345025</td>\n",
       "      <td>2.302967</td>\n",
       "      <td>2.398594</td>\n",
       "      <td>2.038665</td>\n",
       "      <td>1.606514</td>\n",
       "      <td>0.912680</td>\n",
       "      <td>1.863858</td>\n",
       "      <td>0.484354</td>\n",
       "      <td>1.083816</td>\n",
       "      <td>...</td>\n",
       "      <td>0.471090</td>\n",
       "      <td>1.065121</td>\n",
       "      <td>-0.641791</td>\n",
       "      <td>2.329098</td>\n",
       "      <td>0.973954</td>\n",
       "      <td>3.155198</td>\n",
       "      <td>1.010614</td>\n",
       "      <td>-0.124115</td>\n",
       "      <td>0.320235</td>\n",
       "      <td>-1.747840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3521</th>\n",
       "      <td>2.253716</td>\n",
       "      <td>2.352718</td>\n",
       "      <td>2.316705</td>\n",
       "      <td>2.403538</td>\n",
       "      <td>2.035968</td>\n",
       "      <td>1.611631</td>\n",
       "      <td>0.919223</td>\n",
       "      <td>1.856783</td>\n",
       "      <td>0.471090</td>\n",
       "      <td>1.065121</td>\n",
       "      <td>...</td>\n",
       "      <td>0.509947</td>\n",
       "      <td>1.131839</td>\n",
       "      <td>-0.624336</td>\n",
       "      <td>2.318039</td>\n",
       "      <td>0.955342</td>\n",
       "      <td>3.161859</td>\n",
       "      <td>1.030332</td>\n",
       "      <td>-0.111583</td>\n",
       "      <td>0.340687</td>\n",
       "      <td>-1.776756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3522</th>\n",
       "      <td>2.267193</td>\n",
       "      <td>2.365773</td>\n",
       "      <td>2.338970</td>\n",
       "      <td>2.407257</td>\n",
       "      <td>2.045722</td>\n",
       "      <td>1.654725</td>\n",
       "      <td>0.930363</td>\n",
       "      <td>1.856783</td>\n",
       "      <td>0.509947</td>\n",
       "      <td>1.131839</td>\n",
       "      <td>...</td>\n",
       "      <td>0.509947</td>\n",
       "      <td>1.131839</td>\n",
       "      <td>-0.585322</td>\n",
       "      <td>2.408029</td>\n",
       "      <td>0.969665</td>\n",
       "      <td>3.201219</td>\n",
       "      <td>1.036587</td>\n",
       "      <td>-0.111583</td>\n",
       "      <td>0.340687</td>\n",
       "      <td>-1.791214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3523</th>\n",
       "      <td>2.216992</td>\n",
       "      <td>2.322179</td>\n",
       "      <td>2.294997</td>\n",
       "      <td>2.370375</td>\n",
       "      <td>2.046924</td>\n",
       "      <td>1.654725</td>\n",
       "      <td>0.930014</td>\n",
       "      <td>1.856398</td>\n",
       "      <td>0.509947</td>\n",
       "      <td>1.131839</td>\n",
       "      <td>...</td>\n",
       "      <td>0.517757</td>\n",
       "      <td>1.148314</td>\n",
       "      <td>-0.582926</td>\n",
       "      <td>2.298313</td>\n",
       "      <td>0.935281</td>\n",
       "      <td>3.202338</td>\n",
       "      <td>0.961223</td>\n",
       "      <td>-0.087657</td>\n",
       "      <td>0.357165</td>\n",
       "      <td>-1.805671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3524</th>\n",
       "      <td>2.215245</td>\n",
       "      <td>2.321013</td>\n",
       "      <td>2.289176</td>\n",
       "      <td>2.365771</td>\n",
       "      <td>2.036458</td>\n",
       "      <td>1.670872</td>\n",
       "      <td>0.917655</td>\n",
       "      <td>1.782634</td>\n",
       "      <td>0.517757</td>\n",
       "      <td>1.148314</td>\n",
       "      <td>...</td>\n",
       "      <td>0.563020</td>\n",
       "      <td>1.109405</td>\n",
       "      <td>-0.588402</td>\n",
       "      <td>2.207578</td>\n",
       "      <td>0.924060</td>\n",
       "      <td>3.083227</td>\n",
       "      <td>0.970806</td>\n",
       "      <td>-0.067150</td>\n",
       "      <td>0.385914</td>\n",
       "      <td>-1.791214</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 36 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      var1(t-1)  var2(t-1)  var3(t-1)  var4(t-1)  var5(t-1)  var6(t-1)  \\\n",
       "3520   2.246223   2.345025   2.302967   2.398594   2.038665   1.606514   \n",
       "3521   2.253716   2.352718   2.316705   2.403538   2.035968   1.611631   \n",
       "3522   2.267193   2.365773   2.338970   2.407257   2.045722   1.654725   \n",
       "3523   2.216992   2.322179   2.294997   2.370375   2.046924   1.654725   \n",
       "3524   2.215245   2.321013   2.289176   2.365771   2.036458   1.670872   \n",
       "\n",
       "      var7(t-1)  var8(t-1)  var9(t-1)  var10(t-1)    ...      var9(t)  \\\n",
       "3520   0.912680   1.863858   0.484354    1.083816    ...     0.471090   \n",
       "3521   0.919223   1.856783   0.471090    1.065121    ...     0.509947   \n",
       "3522   0.930363   1.856783   0.509947    1.131839    ...     0.509947   \n",
       "3523   0.930014   1.856398   0.509947    1.131839    ...     0.517757   \n",
       "3524   0.917655   1.782634   0.517757    1.148314    ...     0.563020   \n",
       "\n",
       "      var10(t)  var11(t)  var12(t)  var13(t)  var14(t)  var15(t)  var16(t)  \\\n",
       "3520  1.065121 -0.641791  2.329098  0.973954  3.155198  1.010614 -0.124115   \n",
       "3521  1.131839 -0.624336  2.318039  0.955342  3.161859  1.030332 -0.111583   \n",
       "3522  1.131839 -0.585322  2.408029  0.969665  3.201219  1.036587 -0.111583   \n",
       "3523  1.148314 -0.582926  2.298313  0.935281  3.202338  0.961223 -0.087657   \n",
       "3524  1.109405 -0.588402  2.207578  0.924060  3.083227  0.970806 -0.067150   \n",
       "\n",
       "      var17(t)  var18(t)  \n",
       "3520  0.320235 -1.747840  \n",
       "3521  0.340687 -1.776756  \n",
       "3522  0.340687 -1.791214  \n",
       "3523  0.357165 -1.805671  \n",
       "3524  0.385914 -1.791214  \n",
       "\n",
       "[5 rows x 36 columns]"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reframed.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "reframed.drop(reframed.columns[[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35]], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['var1(t-1)', 'var2(t-1)', 'var3(t-1)', 'var4(t-1)', 'var5(t-1)',\n",
       "       'var6(t-1)', 'var7(t-1)', 'var8(t-1)', 'var9(t-1)', 'var10(t-1)',\n",
       "       'var11(t-1)', 'var12(t-1)', 'var13(t-1)', 'var14(t-1)', 'var15(t-1)',\n",
       "       'var16(t-1)', 'var17(t-1)', 'var18(t-1)', 'var1(t)'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reframed.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observations: 3524\n",
      "Training Observations: 2643\n",
      "Testing Observations: 881\n"
     ]
    }
   ],
   "source": [
    "values = reframed.values\n",
    "train_size = int(len(values) * 0.75)\n",
    "train, test = values[0:train_size], values[train_size:len(values)]\n",
    "print('Observations: %d' % (len(values)))\n",
    "print('Training Observations: %d' % (len(train)))\n",
    "print('Testing Observations: %d' % (len(test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2643, 1, 18) (2643,) (881, 1, 18) (881,)\n"
     ]
    }
   ],
   "source": [
    "# split into input and outputs\n",
    "train_X, train_y = train[:, :-1], train[:, -1]\n",
    "test_X, test_y = test[:, :-1], test[:, -1]\n",
    "# reshape input to be 3D [samples, timesteps, features]\n",
    "train_X = train_X.reshape((train_X.shape[0], 1, train_X.shape[1]))\n",
    "test_X = test_X.reshape((test_X.shape[0], 1, test_X.shape[1]))\n",
    "print(train_X.shape, train_y.shape, test_X.shape, test_y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# design network\n",
    "model = Sequential()\n",
    "model.add(LSTM(50, input_shape=(train_X.shape[1], train_X.shape[2])))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mse', optimizer='adam', metrics=['mse', 'mae', 'mape', 'cosine'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 2643 samples, validate on 881 samples\n",
      "Epoch 1/250\n",
      "0s - loss: 0.4523 - mean_squared_error: 0.4523 - mean_absolute_error: 0.5528 - mean_absolute_percentage_error: 257.8170 - cosine_proximity: -3.4014e-01 - val_loss: 1.8499 - val_mean_squared_error: 1.8499 - val_mean_absolute_error: 1.3084 - val_mean_absolute_percentage_error: 85.3384 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 2/250\n",
      "0s - loss: 0.2298 - mean_squared_error: 0.2298 - mean_absolute_error: 0.3914 - mean_absolute_percentage_error: 199.9183 - cosine_proximity: -5.4597e-01 - val_loss: 1.5343 - val_mean_squared_error: 1.5343 - val_mean_absolute_error: 1.1875 - val_mean_absolute_percentage_error: 77.1148 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 3/250\n",
      "0s - loss: 0.1161 - mean_squared_error: 0.1161 - mean_absolute_error: 0.2624 - mean_absolute_percentage_error: 137.3526 - cosine_proximity: -7.2153e-01 - val_loss: 1.3101 - val_mean_squared_error: 1.3101 - val_mean_absolute_error: 1.0946 - val_mean_absolute_percentage_error: 70.8999 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 4/250\n",
      "0s - loss: 0.0619 - mean_squared_error: 0.0619 - mean_absolute_error: 0.1810 - mean_absolute_percentage_error: 82.2852 - cosine_proximity: -9.0238e-01 - val_loss: 1.1648 - val_mean_squared_error: 1.1648 - val_mean_absolute_error: 1.0309 - val_mean_absolute_percentage_error: 66.7116 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 5/250\n",
      "0s - loss: 0.0416 - mean_squared_error: 0.0416 - mean_absolute_error: 0.1600 - mean_absolute_percentage_error: 91.8280 - cosine_proximity: -9.0163e-01 - val_loss: 1.0745 - val_mean_squared_error: 1.0745 - val_mean_absolute_error: 0.9899 - val_mean_absolute_percentage_error: 64.0750 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 6/250\n",
      "0s - loss: 0.0332 - mean_squared_error: 0.0332 - mean_absolute_error: 0.1487 - mean_absolute_percentage_error: 105.9524 - cosine_proximity: -9.1222e-01 - val_loss: 1.0149 - val_mean_squared_error: 1.0149 - val_mean_absolute_error: 0.9619 - val_mean_absolute_percentage_error: 62.2951 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 7/250\n",
      "0s - loss: 0.0272 - mean_squared_error: 0.0272 - mean_absolute_error: 0.1368 - mean_absolute_percentage_error: 114.0580 - cosine_proximity: -9.2281e-01 - val_loss: 0.9687 - val_mean_squared_error: 0.9687 - val_mean_absolute_error: 0.9392 - val_mean_absolute_percentage_error: 60.7964 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 8/250\n",
      "0s - loss: 0.0223 - mean_squared_error: 0.0223 - mean_absolute_error: 0.1251 - mean_absolute_percentage_error: 113.5561 - cosine_proximity: -9.2736e-01 - val_loss: 0.9270 - val_mean_squared_error: 0.9270 - val_mean_absolute_error: 0.9175 - val_mean_absolute_percentage_error: 59.2850 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 9/250\n",
      "0s - loss: 0.0185 - mean_squared_error: 0.0185 - mean_absolute_error: 0.1142 - mean_absolute_percentage_error: 103.6842 - cosine_proximity: -9.3114e-01 - val_loss: 0.8878 - val_mean_squared_error: 0.8878 - val_mean_absolute_error: 0.8959 - val_mean_absolute_percentage_error: 57.7325 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 10/250\n",
      "0s - loss: 0.0155 - mean_squared_error: 0.0155 - mean_absolute_error: 0.1046 - mean_absolute_percentage_error: 89.7111 - cosine_proximity: -9.3341e-01 - val_loss: 0.8519 - val_mean_squared_error: 0.8519 - val_mean_absolute_error: 0.8754 - val_mean_absolute_percentage_error: 56.2281 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 11/250\n",
      "0s - loss: 0.0132 - mean_squared_error: 0.0132 - mean_absolute_error: 0.0963 - mean_absolute_percentage_error: 76.3726 - cosine_proximity: -9.3795e-01 - val_loss: 0.8203 - val_mean_squared_error: 0.8203 - val_mean_absolute_error: 0.8567 - val_mean_absolute_percentage_error: 54.8512 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 12/250\n",
      "0s - loss: 0.0114 - mean_squared_error: 0.0114 - mean_absolute_error: 0.0890 - mean_absolute_percentage_error: 65.4800 - cosine_proximity: -9.4022e-01 - val_loss: 0.7928 - val_mean_squared_error: 0.7928 - val_mean_absolute_error: 0.8401 - val_mean_absolute_percentage_error: 53.6273 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 13/250\n",
      "0s - loss: 0.0099 - mean_squared_error: 0.0099 - mean_absolute_error: 0.0825 - mean_absolute_percentage_error: 56.9330 - cosine_proximity: -9.4779e-01 - val_loss: 0.7689 - val_mean_squared_error: 0.7689 - val_mean_absolute_error: 0.8254 - val_mean_absolute_percentage_error: 52.5445 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 14/250\n",
      "0s - loss: 0.0086 - mean_squared_error: 0.0086 - mean_absolute_error: 0.0768 - mean_absolute_percentage_error: 50.3485 - cosine_proximity: -9.5006e-01 - val_loss: 0.7480 - val_mean_squared_error: 0.7480 - val_mean_absolute_error: 0.8123 - val_mean_absolute_percentage_error: 51.5826 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 15/250\n",
      "0s - loss: 0.0076 - mean_squared_error: 0.0076 - mean_absolute_error: 0.0720 - mean_absolute_percentage_error: 44.9759 - cosine_proximity: -9.5308e-01 - val_loss: 0.7298 - val_mean_squared_error: 0.7298 - val_mean_absolute_error: 0.8007 - val_mean_absolute_percentage_error: 50.7286 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 16/250\n",
      "0s - loss: 0.0069 - mean_squared_error: 0.0069 - mean_absolute_error: 0.0679 - mean_absolute_percentage_error: 40.5104 - cosine_proximity: -9.5989e-01 - val_loss: 0.7140 - val_mean_squared_error: 0.7140 - val_mean_absolute_error: 0.7904 - val_mean_absolute_percentage_error: 49.9760 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 17/250\n",
      "0s - loss: 0.0063 - mean_squared_error: 0.0063 - mean_absolute_error: 0.0645 - mean_absolute_percentage_error: 37.7210 - cosine_proximity: -9.6141e-01 - val_loss: 0.7002 - val_mean_squared_error: 0.7002 - val_mean_absolute_error: 0.7815 - val_mean_absolute_percentage_error: 49.3184 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 18/250\n",
      "0s - loss: 0.0058 - mean_squared_error: 0.0058 - mean_absolute_error: 0.0617 - mean_absolute_percentage_error: 35.8396 - cosine_proximity: -9.5989e-01 - val_loss: 0.6883 - val_mean_squared_error: 0.6883 - val_mean_absolute_error: 0.7736 - val_mean_absolute_percentage_error: 48.7464 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 19/250\n",
      "0s - loss: 0.0054 - mean_squared_error: 0.0054 - mean_absolute_error: 0.0594 - mean_absolute_percentage_error: 36.6961 - cosine_proximity: -9.5914e-01 - val_loss: 0.6780 - val_mean_squared_error: 0.6780 - val_mean_absolute_error: 0.7668 - val_mean_absolute_percentage_error: 48.2484 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 20/250\n",
      "0s - loss: 0.0051 - mean_squared_error: 0.0051 - mean_absolute_error: 0.0574 - mean_absolute_percentage_error: 37.6509 - cosine_proximity: -9.5838e-01 - val_loss: 0.6688 - val_mean_squared_error: 0.6688 - val_mean_absolute_error: 0.7607 - val_mean_absolute_percentage_error: 47.8127 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 21/250\n",
      "0s - loss: 0.0048 - mean_squared_error: 0.0048 - mean_absolute_error: 0.0556 - mean_absolute_percentage_error: 38.3923 - cosine_proximity: -9.5687e-01 - val_loss: 0.6607 - val_mean_squared_error: 0.6607 - val_mean_absolute_error: 0.7554 - val_mean_absolute_percentage_error: 47.4287 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 22/250\n",
      "0s - loss: 0.0046 - mean_squared_error: 0.0046 - mean_absolute_error: 0.0541 - mean_absolute_percentage_error: 38.9252 - cosine_proximity: -9.5535e-01 - val_loss: 0.6534 - val_mean_squared_error: 0.6534 - val_mean_absolute_error: 0.7506 - val_mean_absolute_percentage_error: 47.0866 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 23/250\n",
      "0s - loss: 0.0044 - mean_squared_error: 0.0044 - mean_absolute_error: 0.0527 - mean_absolute_percentage_error: 39.2953 - cosine_proximity: -9.5535e-01 - val_loss: 0.6468 - val_mean_squared_error: 0.6468 - val_mean_absolute_error: 0.7462 - val_mean_absolute_percentage_error: 46.7780 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 24/250\n",
      "0s - loss: 0.0042 - mean_squared_error: 0.0042 - mean_absolute_error: 0.0515 - mean_absolute_percentage_error: 39.5234 - cosine_proximity: -9.5460e-01 - val_loss: 0.6406 - val_mean_squared_error: 0.6406 - val_mean_absolute_error: 0.7422 - val_mean_absolute_percentage_error: 46.4957 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 25/250\n",
      "0s - loss: 0.0041 - mean_squared_error: 0.0041 - mean_absolute_error: 0.0504 - mean_absolute_percentage_error: 39.6489 - cosine_proximity: -9.5535e-01 - val_loss: 0.6349 - val_mean_squared_error: 0.6349 - val_mean_absolute_error: 0.7384 - val_mean_absolute_percentage_error: 46.2334 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 26/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0039 - mean_squared_error: 0.0039 - mean_absolute_error: 0.0494 - mean_absolute_percentage_error: 39.7130 - cosine_proximity: -9.5535e-01 - val_loss: 0.6294 - val_mean_squared_error: 0.6294 - val_mean_absolute_error: 0.7348 - val_mean_absolute_percentage_error: 45.9865 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 27/250\n",
      "0s - loss: 0.0038 - mean_squared_error: 0.0038 - mean_absolute_error: 0.0484 - mean_absolute_percentage_error: 39.7209 - cosine_proximity: -9.5535e-01 - val_loss: 0.6241 - val_mean_squared_error: 0.6241 - val_mean_absolute_error: 0.7314 - val_mean_absolute_percentage_error: 45.7513 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 28/250\n",
      "0s - loss: 0.0037 - mean_squared_error: 0.0037 - mean_absolute_error: 0.0476 - mean_absolute_percentage_error: 39.6907 - cosine_proximity: -9.5535e-01 - val_loss: 0.6190 - val_mean_squared_error: 0.6190 - val_mean_absolute_error: 0.7280 - val_mean_absolute_percentage_error: 45.5251 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 29/250\n",
      "0s - loss: 0.0036 - mean_squared_error: 0.0036 - mean_absolute_error: 0.0468 - mean_absolute_percentage_error: 39.6351 - cosine_proximity: -9.5535e-01 - val_loss: 0.6141 - val_mean_squared_error: 0.6141 - val_mean_absolute_error: 0.7248 - val_mean_absolute_percentage_error: 45.3061 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 30/250\n",
      "0s - loss: 0.0035 - mean_squared_error: 0.0035 - mean_absolute_error: 0.0461 - mean_absolute_percentage_error: 39.5608 - cosine_proximity: -9.5611e-01 - val_loss: 0.6092 - val_mean_squared_error: 0.6092 - val_mean_absolute_error: 0.7217 - val_mean_absolute_percentage_error: 45.0929 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 31/250\n",
      "0s - loss: 0.0034 - mean_squared_error: 0.0034 - mean_absolute_error: 0.0454 - mean_absolute_percentage_error: 39.4806 - cosine_proximity: -9.5687e-01 - val_loss: 0.6045 - val_mean_squared_error: 0.6045 - val_mean_absolute_error: 0.7186 - val_mean_absolute_percentage_error: 44.8847 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 32/250\n",
      "0s - loss: 0.0033 - mean_squared_error: 0.0033 - mean_absolute_error: 0.0448 - mean_absolute_percentage_error: 39.4002 - cosine_proximity: -9.5535e-01 - val_loss: 0.5999 - val_mean_squared_error: 0.5999 - val_mean_absolute_error: 0.7156 - val_mean_absolute_percentage_error: 44.6811 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 33/250\n",
      "0s - loss: 0.0032 - mean_squared_error: 0.0032 - mean_absolute_error: 0.0442 - mean_absolute_percentage_error: 39.3169 - cosine_proximity: -9.5535e-01 - val_loss: 0.5954 - val_mean_squared_error: 0.5954 - val_mean_absolute_error: 0.7126 - val_mean_absolute_percentage_error: 44.4818 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 34/250\n",
      "0s - loss: 0.0032 - mean_squared_error: 0.0032 - mean_absolute_error: 0.0437 - mean_absolute_percentage_error: 39.2312 - cosine_proximity: -9.5535e-01 - val_loss: 0.5910 - val_mean_squared_error: 0.5910 - val_mean_absolute_error: 0.7097 - val_mean_absolute_percentage_error: 44.2866 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 35/250\n",
      "0s - loss: 0.0031 - mean_squared_error: 0.0031 - mean_absolute_error: 0.0432 - mean_absolute_percentage_error: 39.1434 - cosine_proximity: -9.5535e-01 - val_loss: 0.5866 - val_mean_squared_error: 0.5866 - val_mean_absolute_error: 0.7069 - val_mean_absolute_percentage_error: 44.0956 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 36/250\n",
      "0s - loss: 0.0031 - mean_squared_error: 0.0031 - mean_absolute_error: 0.0427 - mean_absolute_percentage_error: 39.0579 - cosine_proximity: -9.5535e-01 - val_loss: 0.5824 - val_mean_squared_error: 0.5824 - val_mean_absolute_error: 0.7041 - val_mean_absolute_percentage_error: 43.9087 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 37/250\n",
      "0s - loss: 0.0030 - mean_squared_error: 0.0030 - mean_absolute_error: 0.0423 - mean_absolute_percentage_error: 38.9755 - cosine_proximity: -9.5535e-01 - val_loss: 0.5783 - val_mean_squared_error: 0.5783 - val_mean_absolute_error: 0.7014 - val_mean_absolute_percentage_error: 43.7259 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 38/250\n",
      "0s - loss: 0.0030 - mean_squared_error: 0.0030 - mean_absolute_error: 0.0419 - mean_absolute_percentage_error: 38.8956 - cosine_proximity: -9.5535e-01 - val_loss: 0.5743 - val_mean_squared_error: 0.5743 - val_mean_absolute_error: 0.6987 - val_mean_absolute_percentage_error: 43.5473 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 39/250\n",
      "0s - loss: 0.0029 - mean_squared_error: 0.0029 - mean_absolute_error: 0.0415 - mean_absolute_percentage_error: 38.8195 - cosine_proximity: -9.5535e-01 - val_loss: 0.5704 - val_mean_squared_error: 0.5704 - val_mean_absolute_error: 0.6961 - val_mean_absolute_percentage_error: 43.3727 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 40/250\n",
      "0s - loss: 0.0029 - mean_squared_error: 0.0029 - mean_absolute_error: 0.0412 - mean_absolute_percentage_error: 38.7435 - cosine_proximity: -9.5535e-01 - val_loss: 0.5666 - val_mean_squared_error: 0.5666 - val_mean_absolute_error: 0.6936 - val_mean_absolute_percentage_error: 43.2021 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 41/250\n",
      "0s - loss: 0.0028 - mean_squared_error: 0.0028 - mean_absolute_error: 0.0408 - mean_absolute_percentage_error: 38.6673 - cosine_proximity: -9.5611e-01 - val_loss: 0.5629 - val_mean_squared_error: 0.5629 - val_mean_absolute_error: 0.6911 - val_mean_absolute_percentage_error: 43.0354 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 42/250\n",
      "0s - loss: 0.0028 - mean_squared_error: 0.0028 - mean_absolute_error: 0.0405 - mean_absolute_percentage_error: 38.5911 - cosine_proximity: -9.5611e-01 - val_loss: 0.5592 - val_mean_squared_error: 0.5592 - val_mean_absolute_error: 0.6887 - val_mean_absolute_percentage_error: 42.8724 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 43/250\n",
      "0s - loss: 0.0028 - mean_squared_error: 0.0028 - mean_absolute_error: 0.0403 - mean_absolute_percentage_error: 38.5128 - cosine_proximity: -9.5687e-01 - val_loss: 0.5557 - val_mean_squared_error: 0.5557 - val_mean_absolute_error: 0.6863 - val_mean_absolute_percentage_error: 42.7131 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 44/250\n",
      "0s - loss: 0.0027 - mean_squared_error: 0.0027 - mean_absolute_error: 0.0400 - mean_absolute_percentage_error: 38.4330 - cosine_proximity: -9.5687e-01 - val_loss: 0.5523 - val_mean_squared_error: 0.5523 - val_mean_absolute_error: 0.6840 - val_mean_absolute_percentage_error: 42.5574 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 45/250\n",
      "0s - loss: 0.0027 - mean_squared_error: 0.0027 - mean_absolute_error: 0.0397 - mean_absolute_percentage_error: 38.3523 - cosine_proximity: -9.5687e-01 - val_loss: 0.5489 - val_mean_squared_error: 0.5489 - val_mean_absolute_error: 0.6817 - val_mean_absolute_percentage_error: 42.4049 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 46/250\n",
      "0s - loss: 0.0027 - mean_squared_error: 0.0027 - mean_absolute_error: 0.0395 - mean_absolute_percentage_error: 38.2708 - cosine_proximity: -9.5687e-01 - val_loss: 0.5456 - val_mean_squared_error: 0.5456 - val_mean_absolute_error: 0.6795 - val_mean_absolute_percentage_error: 42.2556 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 47/250\n",
      "0s - loss: 0.0026 - mean_squared_error: 0.0026 - mean_absolute_error: 0.0393 - mean_absolute_percentage_error: 38.1882 - cosine_proximity: -9.5687e-01 - val_loss: 0.5424 - val_mean_squared_error: 0.5424 - val_mean_absolute_error: 0.6773 - val_mean_absolute_percentage_error: 42.1094 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 48/250\n",
      "0s - loss: 0.0026 - mean_squared_error: 0.0026 - mean_absolute_error: 0.0390 - mean_absolute_percentage_error: 38.1044 - cosine_proximity: -9.5687e-01 - val_loss: 0.5393 - val_mean_squared_error: 0.5393 - val_mean_absolute_error: 0.6752 - val_mean_absolute_percentage_error: 41.9660 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 49/250\n",
      "0s - loss: 0.0026 - mean_squared_error: 0.0026 - mean_absolute_error: 0.0388 - mean_absolute_percentage_error: 38.0198 - cosine_proximity: -9.5762e-01 - val_loss: 0.5363 - val_mean_squared_error: 0.5363 - val_mean_absolute_error: 0.6731 - val_mean_absolute_percentage_error: 41.8253 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 50/250\n",
      "0s - loss: 0.0026 - mean_squared_error: 0.0026 - mean_absolute_error: 0.0386 - mean_absolute_percentage_error: 37.9335 - cosine_proximity: -9.5762e-01 - val_loss: 0.5333 - val_mean_squared_error: 0.5333 - val_mean_absolute_error: 0.6710 - val_mean_absolute_percentage_error: 41.6873 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 51/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0025 - mean_squared_error: 0.0025 - mean_absolute_error: 0.0384 - mean_absolute_percentage_error: 37.8464 - cosine_proximity: -9.5762e-01 - val_loss: 0.5304 - val_mean_squared_error: 0.5304 - val_mean_absolute_error: 0.6690 - val_mean_absolute_percentage_error: 41.5517 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 52/250\n",
      "0s - loss: 0.0025 - mean_squared_error: 0.0025 - mean_absolute_error: 0.0383 - mean_absolute_percentage_error: 37.7584 - cosine_proximity: -9.5762e-01 - val_loss: 0.5275 - val_mean_squared_error: 0.5275 - val_mean_absolute_error: 0.6670 - val_mean_absolute_percentage_error: 41.4184 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 53/250\n",
      "0s - loss: 0.0025 - mean_squared_error: 0.0025 - mean_absolute_error: 0.0381 - mean_absolute_percentage_error: 37.6694 - cosine_proximity: -9.5687e-01 - val_loss: 0.5247 - val_mean_squared_error: 0.5247 - val_mean_absolute_error: 0.6651 - val_mean_absolute_percentage_error: 41.2873 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 54/250\n",
      "0s - loss: 0.0025 - mean_squared_error: 0.0025 - mean_absolute_error: 0.0379 - mean_absolute_percentage_error: 37.5796 - cosine_proximity: -9.5762e-01 - val_loss: 0.5220 - val_mean_squared_error: 0.5220 - val_mean_absolute_error: 0.6632 - val_mean_absolute_percentage_error: 41.1583 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 55/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0377 - mean_absolute_percentage_error: 37.4888 - cosine_proximity: -9.5762e-01 - val_loss: 0.5193 - val_mean_squared_error: 0.5193 - val_mean_absolute_error: 0.6613 - val_mean_absolute_percentage_error: 41.0313 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 56/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0376 - mean_absolute_percentage_error: 37.3970 - cosine_proximity: -9.5762e-01 - val_loss: 0.5167 - val_mean_squared_error: 0.5167 - val_mean_absolute_error: 0.6595 - val_mean_absolute_percentage_error: 40.9062 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 57/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0374 - mean_absolute_percentage_error: 37.3049 - cosine_proximity: -9.5762e-01 - val_loss: 0.5141 - val_mean_squared_error: 0.5141 - val_mean_absolute_error: 0.6576 - val_mean_absolute_percentage_error: 40.7829 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 58/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0373 - mean_absolute_percentage_error: 37.2126 - cosine_proximity: -9.5838e-01 - val_loss: 0.5116 - val_mean_squared_error: 0.5116 - val_mean_absolute_error: 0.6559 - val_mean_absolute_percentage_error: 40.6613 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 59/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0371 - mean_absolute_percentage_error: 37.1202 - cosine_proximity: -9.5838e-01 - val_loss: 0.5091 - val_mean_squared_error: 0.5091 - val_mean_absolute_error: 0.6541 - val_mean_absolute_percentage_error: 40.5415 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 60/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0369 - mean_absolute_percentage_error: 37.0287 - cosine_proximity: -9.5838e-01 - val_loss: 0.5066 - val_mean_squared_error: 0.5066 - val_mean_absolute_error: 0.6523 - val_mean_absolute_percentage_error: 40.4232 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 61/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0368 - mean_absolute_percentage_error: 36.9449 - cosine_proximity: -9.5838e-01 - val_loss: 0.5042 - val_mean_squared_error: 0.5042 - val_mean_absolute_error: 0.6506 - val_mean_absolute_percentage_error: 40.3064 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 62/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0367 - mean_absolute_percentage_error: 36.8614 - cosine_proximity: -9.5838e-01 - val_loss: 0.5019 - val_mean_squared_error: 0.5019 - val_mean_absolute_error: 0.6489 - val_mean_absolute_percentage_error: 40.1912 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 63/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0365 - mean_absolute_percentage_error: 36.7778 - cosine_proximity: -9.5838e-01 - val_loss: 0.4995 - val_mean_squared_error: 0.4995 - val_mean_absolute_error: 0.6473 - val_mean_absolute_percentage_error: 40.0773 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 64/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0364 - mean_absolute_percentage_error: 36.7064 - cosine_proximity: -9.5838e-01 - val_loss: 0.4973 - val_mean_squared_error: 0.4973 - val_mean_absolute_error: 0.6456 - val_mean_absolute_percentage_error: 39.9649 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 65/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0362 - mean_absolute_percentage_error: 36.6377 - cosine_proximity: -9.5838e-01 - val_loss: 0.4950 - val_mean_squared_error: 0.4950 - val_mean_absolute_error: 0.6440 - val_mean_absolute_percentage_error: 39.8538 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 66/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0361 - mean_absolute_percentage_error: 36.5696 - cosine_proximity: -9.5838e-01 - val_loss: 0.4928 - val_mean_squared_error: 0.4928 - val_mean_absolute_error: 0.6424 - val_mean_absolute_percentage_error: 39.7440 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 67/250\n",
      "0s - loss: 0.0023 - mean_squared_error: 0.0023 - mean_absolute_error: 0.0360 - mean_absolute_percentage_error: 36.5020 - cosine_proximity: -9.5838e-01 - val_loss: 0.4906 - val_mean_squared_error: 0.4906 - val_mean_absolute_error: 0.6408 - val_mean_absolute_percentage_error: 39.6354 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 68/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0358 - mean_absolute_percentage_error: 36.4350 - cosine_proximity: -9.5838e-01 - val_loss: 0.4884 - val_mean_squared_error: 0.4884 - val_mean_absolute_error: 0.6392 - val_mean_absolute_percentage_error: 39.5281 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 69/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0357 - mean_absolute_percentage_error: 36.3695 - cosine_proximity: -9.5838e-01 - val_loss: 0.4863 - val_mean_squared_error: 0.4863 - val_mean_absolute_error: 0.6377 - val_mean_absolute_percentage_error: 39.4219 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 70/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0356 - mean_absolute_percentage_error: 36.3051 - cosine_proximity: -9.5838e-01 - val_loss: 0.4842 - val_mean_squared_error: 0.4842 - val_mean_absolute_error: 0.6361 - val_mean_absolute_percentage_error: 39.3170 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 71/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0354 - mean_absolute_percentage_error: 36.2415 - cosine_proximity: -9.5838e-01 - val_loss: 0.4822 - val_mean_squared_error: 0.4822 - val_mean_absolute_error: 0.6346 - val_mean_absolute_percentage_error: 39.2131 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 72/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0353 - mean_absolute_percentage_error: 36.1787 - cosine_proximity: -9.5914e-01 - val_loss: 0.4802 - val_mean_squared_error: 0.4802 - val_mean_absolute_error: 0.6331 - val_mean_absolute_percentage_error: 39.1103 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 73/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0352 - mean_absolute_percentage_error: 36.1167 - cosine_proximity: -9.5838e-01 - val_loss: 0.4782 - val_mean_squared_error: 0.4782 - val_mean_absolute_error: 0.6316 - val_mean_absolute_percentage_error: 39.0087 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 74/250\n",
      "0s - loss: 0.0022 - mean_squared_error: 0.0022 - mean_absolute_error: 0.0351 - mean_absolute_percentage_error: 36.0554 - cosine_proximity: -9.5838e-01 - val_loss: 0.4762 - val_mean_squared_error: 0.4762 - val_mean_absolute_error: 0.6302 - val_mean_absolute_percentage_error: 38.9080 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 75/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0350 - mean_absolute_percentage_error: 35.9947 - cosine_proximity: -9.5838e-01 - val_loss: 0.4742 - val_mean_squared_error: 0.4742 - val_mean_absolute_error: 0.6287 - val_mean_absolute_percentage_error: 38.8084 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 76/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0348 - mean_absolute_percentage_error: 35.9346 - cosine_proximity: -9.5838e-01 - val_loss: 0.4723 - val_mean_squared_error: 0.4723 - val_mean_absolute_error: 0.6273 - val_mean_absolute_percentage_error: 38.7098 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 77/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0347 - mean_absolute_percentage_error: 35.8753 - cosine_proximity: -9.5838e-01 - val_loss: 0.4704 - val_mean_squared_error: 0.4704 - val_mean_absolute_error: 0.6259 - val_mean_absolute_percentage_error: 38.6122 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 78/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0346 - mean_absolute_percentage_error: 35.8223 - cosine_proximity: -9.5914e-01 - val_loss: 0.4685 - val_mean_squared_error: 0.4685 - val_mean_absolute_error: 0.6245 - val_mean_absolute_percentage_error: 38.5155 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 79/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0345 - mean_absolute_percentage_error: 35.7740 - cosine_proximity: -9.5989e-01 - val_loss: 0.4667 - val_mean_squared_error: 0.4667 - val_mean_absolute_error: 0.6231 - val_mean_absolute_percentage_error: 38.4197 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 80/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0344 - mean_absolute_percentage_error: 35.7307 - cosine_proximity: -9.5914e-01 - val_loss: 0.4648 - val_mean_squared_error: 0.4648 - val_mean_absolute_error: 0.6217 - val_mean_absolute_percentage_error: 38.3248 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 81/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0343 - mean_absolute_percentage_error: 35.6879 - cosine_proximity: -9.5914e-01 - val_loss: 0.4630 - val_mean_squared_error: 0.4630 - val_mean_absolute_error: 0.6203 - val_mean_absolute_percentage_error: 38.2308 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 82/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0342 - mean_absolute_percentage_error: 35.6457 - cosine_proximity: -9.5914e-01 - val_loss: 0.4613 - val_mean_squared_error: 0.4613 - val_mean_absolute_error: 0.6190 - val_mean_absolute_percentage_error: 38.1377 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 83/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0341 - mean_absolute_percentage_error: 35.6041 - cosine_proximity: -9.5989e-01 - val_loss: 0.4595 - val_mean_squared_error: 0.4595 - val_mean_absolute_error: 0.6176 - val_mean_absolute_percentage_error: 38.0455 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 84/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0340 - mean_absolute_percentage_error: 35.5630 - cosine_proximity: -9.5989e-01 - val_loss: 0.4577 - val_mean_squared_error: 0.4577 - val_mean_absolute_error: 0.6163 - val_mean_absolute_percentage_error: 37.9540 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 85/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0339 - mean_absolute_percentage_error: 35.5223 - cosine_proximity: -9.5989e-01 - val_loss: 0.4560 - val_mean_squared_error: 0.4560 - val_mean_absolute_error: 0.6150 - val_mean_absolute_percentage_error: 37.8634 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 86/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0338 - mean_absolute_percentage_error: 35.4819 - cosine_proximity: -9.6065e-01 - val_loss: 0.4543 - val_mean_squared_error: 0.4543 - val_mean_absolute_error: 0.6137 - val_mean_absolute_percentage_error: 37.7735 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 87/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0337 - mean_absolute_percentage_error: 35.4417 - cosine_proximity: -9.5989e-01 - val_loss: 0.4526 - val_mean_squared_error: 0.4526 - val_mean_absolute_error: 0.6124 - val_mean_absolute_percentage_error: 37.6844 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 88/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0336 - mean_absolute_percentage_error: 35.4017 - cosine_proximity: -9.6065e-01 - val_loss: 0.4509 - val_mean_squared_error: 0.4509 - val_mean_absolute_error: 0.6111 - val_mean_absolute_percentage_error: 37.5961 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 89/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0335 - mean_absolute_percentage_error: 35.3620 - cosine_proximity: -9.6065e-01 - val_loss: 0.4493 - val_mean_squared_error: 0.4493 - val_mean_absolute_error: 0.6098 - val_mean_absolute_percentage_error: 37.5085 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 90/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0334 - mean_absolute_percentage_error: 35.3231 - cosine_proximity: -9.6065e-01 - val_loss: 0.4476 - val_mean_squared_error: 0.4476 - val_mean_absolute_error: 0.6086 - val_mean_absolute_percentage_error: 37.4217 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 91/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0333 - mean_absolute_percentage_error: 35.2852 - cosine_proximity: -9.6065e-01 - val_loss: 0.4460 - val_mean_squared_error: 0.4460 - val_mean_absolute_error: 0.6073 - val_mean_absolute_percentage_error: 37.3355 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 92/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0332 - mean_absolute_percentage_error: 35.2475 - cosine_proximity: -9.6065e-01 - val_loss: 0.4444 - val_mean_squared_error: 0.4444 - val_mean_absolute_error: 0.6061 - val_mean_absolute_percentage_error: 37.2501 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 93/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0331 - mean_absolute_percentage_error: 35.2099 - cosine_proximity: -9.6065e-01 - val_loss: 0.4428 - val_mean_squared_error: 0.4428 - val_mean_absolute_error: 0.6048 - val_mean_absolute_percentage_error: 37.1653 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 94/250\n",
      "0s - loss: 0.0020 - mean_squared_error: 0.0020 - mean_absolute_error: 0.0330 - mean_absolute_percentage_error: 35.1724 - cosine_proximity: -9.5989e-01 - val_loss: 0.4412 - val_mean_squared_error: 0.4412 - val_mean_absolute_error: 0.6036 - val_mean_absolute_percentage_error: 37.0811 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 95/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0329 - mean_absolute_percentage_error: 35.1348 - cosine_proximity: -9.5989e-01 - val_loss: 0.4397 - val_mean_squared_error: 0.4397 - val_mean_absolute_error: 0.6024 - val_mean_absolute_percentage_error: 36.9977 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 96/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0328 - mean_absolute_percentage_error: 35.0975 - cosine_proximity: -9.6065e-01 - val_loss: 0.4381 - val_mean_squared_error: 0.4381 - val_mean_absolute_error: 0.6012 - val_mean_absolute_percentage_error: 36.9150 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 97/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0328 - mean_absolute_percentage_error: 35.0602 - cosine_proximity: -9.6065e-01 - val_loss: 0.4366 - val_mean_squared_error: 0.4366 - val_mean_absolute_error: 0.6000 - val_mean_absolute_percentage_error: 36.8331 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 98/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0327 - mean_absolute_percentage_error: 35.0232 - cosine_proximity: -9.6065e-01 - val_loss: 0.4351 - val_mean_squared_error: 0.4351 - val_mean_absolute_error: 0.5988 - val_mean_absolute_percentage_error: 36.7517 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 99/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0326 - mean_absolute_percentage_error: 34.9865 - cosine_proximity: -9.6065e-01 - val_loss: 0.4336 - val_mean_squared_error: 0.4336 - val_mean_absolute_error: 0.5976 - val_mean_absolute_percentage_error: 36.6710 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 100/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0325 - mean_absolute_percentage_error: 34.9495 - cosine_proximity: -9.6065e-01 - val_loss: 0.4321 - val_mean_squared_error: 0.4321 - val_mean_absolute_error: 0.5965 - val_mean_absolute_percentage_error: 36.5909 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 101/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0324 - mean_absolute_percentage_error: 34.9123 - cosine_proximity: -9.6065e-01 - val_loss: 0.4306 - val_mean_squared_error: 0.4306 - val_mean_absolute_error: 0.5953 - val_mean_absolute_percentage_error: 36.5113 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 102/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0324 - mean_absolute_percentage_error: 34.8753 - cosine_proximity: -9.6065e-01 - val_loss: 0.4291 - val_mean_squared_error: 0.4291 - val_mean_absolute_error: 0.5941 - val_mean_absolute_percentage_error: 36.4323 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 103/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0323 - mean_absolute_percentage_error: 34.8381 - cosine_proximity: -9.6065e-01 - val_loss: 0.4276 - val_mean_squared_error: 0.4276 - val_mean_absolute_error: 0.5930 - val_mean_absolute_percentage_error: 36.3538 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 104/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0322 - mean_absolute_percentage_error: 34.8006 - cosine_proximity: -9.6065e-01 - val_loss: 0.4262 - val_mean_squared_error: 0.4262 - val_mean_absolute_error: 0.5918 - val_mean_absolute_percentage_error: 36.2759 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 105/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0321 - mean_absolute_percentage_error: 34.7628 - cosine_proximity: -9.6065e-01 - val_loss: 0.4248 - val_mean_squared_error: 0.4248 - val_mean_absolute_error: 0.5907 - val_mean_absolute_percentage_error: 36.1986 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 106/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0320 - mean_absolute_percentage_error: 34.7248 - cosine_proximity: -9.6065e-01 - val_loss: 0.4233 - val_mean_squared_error: 0.4233 - val_mean_absolute_error: 0.5896 - val_mean_absolute_percentage_error: 36.1217 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 107/250\n",
      "0s - loss: 0.0019 - mean_squared_error: 0.0019 - mean_absolute_error: 0.0320 - mean_absolute_percentage_error: 34.6867 - cosine_proximity: -9.6065e-01 - val_loss: 0.4219 - val_mean_squared_error: 0.4219 - val_mean_absolute_error: 0.5885 - val_mean_absolute_percentage_error: 36.0454 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 108/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0319 - mean_absolute_percentage_error: 34.6483 - cosine_proximity: -9.6065e-01 - val_loss: 0.4205 - val_mean_squared_error: 0.4205 - val_mean_absolute_error: 0.5874 - val_mean_absolute_percentage_error: 35.9696 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 109/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0318 - mean_absolute_percentage_error: 34.6097 - cosine_proximity: -9.6065e-01 - val_loss: 0.4191 - val_mean_squared_error: 0.4191 - val_mean_absolute_error: 0.5863 - val_mean_absolute_percentage_error: 35.8943 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 110/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0317 - mean_absolute_percentage_error: 34.5717 - cosine_proximity: -9.6141e-01 - val_loss: 0.4177 - val_mean_squared_error: 0.4177 - val_mean_absolute_error: 0.5852 - val_mean_absolute_percentage_error: 35.8196 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 111/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0317 - mean_absolute_percentage_error: 34.5334 - cosine_proximity: -9.6141e-01 - val_loss: 0.4164 - val_mean_squared_error: 0.4164 - val_mean_absolute_error: 0.5841 - val_mean_absolute_percentage_error: 35.7453 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 112/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0316 - mean_absolute_percentage_error: 34.4954 - cosine_proximity: -9.6141e-01 - val_loss: 0.4150 - val_mean_squared_error: 0.4150 - val_mean_absolute_error: 0.5830 - val_mean_absolute_percentage_error: 35.6714 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 113/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0315 - mean_absolute_percentage_error: 34.4577 - cosine_proximity: -9.6141e-01 - val_loss: 0.4136 - val_mean_squared_error: 0.4136 - val_mean_absolute_error: 0.5819 - val_mean_absolute_percentage_error: 35.5981 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 114/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0315 - mean_absolute_percentage_error: 34.4199 - cosine_proximity: -9.6141e-01 - val_loss: 0.4123 - val_mean_squared_error: 0.4123 - val_mean_absolute_error: 0.5808 - val_mean_absolute_percentage_error: 35.5252 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 115/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0314 - mean_absolute_percentage_error: 34.3838 - cosine_proximity: -9.6141e-01 - val_loss: 0.4110 - val_mean_squared_error: 0.4110 - val_mean_absolute_error: 0.5798 - val_mean_absolute_percentage_error: 35.4529 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 116/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0313 - mean_absolute_percentage_error: 34.3480 - cosine_proximity: -9.6141e-01 - val_loss: 0.4096 - val_mean_squared_error: 0.4096 - val_mean_absolute_error: 0.5787 - val_mean_absolute_percentage_error: 35.3809 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 117/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0313 - mean_absolute_percentage_error: 34.3118 - cosine_proximity: -9.6065e-01 - val_loss: 0.4083 - val_mean_squared_error: 0.4083 - val_mean_absolute_error: 0.5777 - val_mean_absolute_percentage_error: 35.3095 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 118/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0312 - mean_absolute_percentage_error: 34.2750 - cosine_proximity: -9.6065e-01 - val_loss: 0.4070 - val_mean_squared_error: 0.4070 - val_mean_absolute_error: 0.5766 - val_mean_absolute_percentage_error: 35.2384 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 119/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0312 - mean_absolute_percentage_error: 34.2377 - cosine_proximity: -9.6065e-01 - val_loss: 0.4057 - val_mean_squared_error: 0.4057 - val_mean_absolute_error: 0.5756 - val_mean_absolute_percentage_error: 35.1679 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 120/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0311 - mean_absolute_percentage_error: 34.2000 - cosine_proximity: -9.6141e-01 - val_loss: 0.4044 - val_mean_squared_error: 0.4044 - val_mean_absolute_error: 0.5745 - val_mean_absolute_percentage_error: 35.0978 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 121/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0310 - mean_absolute_percentage_error: 34.1621 - cosine_proximity: -9.6216e-01 - val_loss: 0.4031 - val_mean_squared_error: 0.4031 - val_mean_absolute_error: 0.5735 - val_mean_absolute_percentage_error: 35.0281 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 122/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0310 - mean_absolute_percentage_error: 34.1237 - cosine_proximity: -9.6216e-01 - val_loss: 0.4018 - val_mean_squared_error: 0.4018 - val_mean_absolute_error: 0.5725 - val_mean_absolute_percentage_error: 34.9589 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 123/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0309 - mean_absolute_percentage_error: 34.0846 - cosine_proximity: -9.6216e-01 - val_loss: 0.4006 - val_mean_squared_error: 0.4006 - val_mean_absolute_error: 0.5715 - val_mean_absolute_percentage_error: 34.8900 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 124/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0309 - mean_absolute_percentage_error: 34.0453 - cosine_proximity: -9.6216e-01 - val_loss: 0.3993 - val_mean_squared_error: 0.3993 - val_mean_absolute_error: 0.5704 - val_mean_absolute_percentage_error: 34.8217 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 125/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0308 - mean_absolute_percentage_error: 34.0087 - cosine_proximity: -9.6216e-01 - val_loss: 0.3981 - val_mean_squared_error: 0.3981 - val_mean_absolute_error: 0.5694 - val_mean_absolute_percentage_error: 34.7537 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 126/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0307 - mean_absolute_percentage_error: 33.9751 - cosine_proximity: -9.6216e-01 - val_loss: 0.3968 - val_mean_squared_error: 0.3968 - val_mean_absolute_error: 0.5684 - val_mean_absolute_percentage_error: 34.6862 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 127/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0307 - mean_absolute_percentage_error: 33.9408 - cosine_proximity: -9.6292e-01 - val_loss: 0.3956 - val_mean_squared_error: 0.3956 - val_mean_absolute_error: 0.5674 - val_mean_absolute_percentage_error: 34.6191 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 128/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0306 - mean_absolute_percentage_error: 33.9067 - cosine_proximity: -9.6292e-01 - val_loss: 0.3944 - val_mean_squared_error: 0.3944 - val_mean_absolute_error: 0.5664 - val_mean_absolute_percentage_error: 34.5524 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 129/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0306 - mean_absolute_percentage_error: 33.8720 - cosine_proximity: -9.6368e-01 - val_loss: 0.3931 - val_mean_squared_error: 0.3931 - val_mean_absolute_error: 0.5655 - val_mean_absolute_percentage_error: 34.4862 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 130/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0305 - mean_absolute_percentage_error: 33.8368 - cosine_proximity: -9.6368e-01 - val_loss: 0.3919 - val_mean_squared_error: 0.3919 - val_mean_absolute_error: 0.5645 - val_mean_absolute_percentage_error: 34.4203 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 131/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0305 - mean_absolute_percentage_error: 33.8006 - cosine_proximity: -9.6368e-01 - val_loss: 0.3907 - val_mean_squared_error: 0.3907 - val_mean_absolute_error: 0.5635 - val_mean_absolute_percentage_error: 34.3549 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 132/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0304 - mean_absolute_percentage_error: 33.7639 - cosine_proximity: -9.6368e-01 - val_loss: 0.3895 - val_mean_squared_error: 0.3895 - val_mean_absolute_error: 0.5625 - val_mean_absolute_percentage_error: 34.2898 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 133/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0304 - mean_absolute_percentage_error: 33.7279 - cosine_proximity: -9.6368e-01 - val_loss: 0.3883 - val_mean_squared_error: 0.3883 - val_mean_absolute_error: 0.5616 - val_mean_absolute_percentage_error: 34.2252 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 134/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0303 - mean_absolute_percentage_error: 33.6917 - cosine_proximity: -9.6368e-01 - val_loss: 0.3872 - val_mean_squared_error: 0.3872 - val_mean_absolute_error: 0.5606 - val_mean_absolute_percentage_error: 34.1610 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 135/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0303 - mean_absolute_percentage_error: 33.6545 - cosine_proximity: -9.6368e-01 - val_loss: 0.3860 - val_mean_squared_error: 0.3860 - val_mean_absolute_error: 0.5597 - val_mean_absolute_percentage_error: 34.0973 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 136/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0302 - mean_absolute_percentage_error: 33.6170 - cosine_proximity: -9.6368e-01 - val_loss: 0.3848 - val_mean_squared_error: 0.3848 - val_mean_absolute_error: 0.5587 - val_mean_absolute_percentage_error: 34.0340 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 137/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0302 - mean_absolute_percentage_error: 33.5783 - cosine_proximity: -9.6368e-01 - val_loss: 0.3837 - val_mean_squared_error: 0.3837 - val_mean_absolute_error: 0.5578 - val_mean_absolute_percentage_error: 33.9712 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 138/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0302 - mean_absolute_percentage_error: 33.5393 - cosine_proximity: -9.6292e-01 - val_loss: 0.3825 - val_mean_squared_error: 0.3825 - val_mean_absolute_error: 0.5568 - val_mean_absolute_percentage_error: 33.9087 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 139/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0301 - mean_absolute_percentage_error: 33.4991 - cosine_proximity: -9.6292e-01 - val_loss: 0.3814 - val_mean_squared_error: 0.3814 - val_mean_absolute_error: 0.5559 - val_mean_absolute_percentage_error: 33.8467 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 140/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0301 - mean_absolute_percentage_error: 33.4591 - cosine_proximity: -9.6368e-01 - val_loss: 0.3802 - val_mean_squared_error: 0.3802 - val_mean_absolute_error: 0.5550 - val_mean_absolute_percentage_error: 33.7850 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 141/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0300 - mean_absolute_percentage_error: 33.4184 - cosine_proximity: -9.6368e-01 - val_loss: 0.3791 - val_mean_squared_error: 0.3791 - val_mean_absolute_error: 0.5541 - val_mean_absolute_percentage_error: 33.7237 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 142/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0300 - mean_absolute_percentage_error: 33.3781 - cosine_proximity: -9.6368e-01 - val_loss: 0.3780 - val_mean_squared_error: 0.3780 - val_mean_absolute_error: 0.5532 - val_mean_absolute_percentage_error: 33.6628 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 143/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0299 - mean_absolute_percentage_error: 33.3358 - cosine_proximity: -9.6368e-01 - val_loss: 0.3769 - val_mean_squared_error: 0.3769 - val_mean_absolute_error: 0.5522 - val_mean_absolute_percentage_error: 33.6024 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 144/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0299 - mean_absolute_percentage_error: 33.2947 - cosine_proximity: -9.6368e-01 - val_loss: 0.3758 - val_mean_squared_error: 0.3758 - val_mean_absolute_error: 0.5513 - val_mean_absolute_percentage_error: 33.5422 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 145/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0298 - mean_absolute_percentage_error: 33.2504 - cosine_proximity: -9.6368e-01 - val_loss: 0.3747 - val_mean_squared_error: 0.3747 - val_mean_absolute_error: 0.5504 - val_mean_absolute_percentage_error: 33.4826 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 146/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0298 - mean_absolute_percentage_error: 33.2089 - cosine_proximity: -9.6368e-01 - val_loss: 0.3736 - val_mean_squared_error: 0.3736 - val_mean_absolute_error: 0.5495 - val_mean_absolute_percentage_error: 33.4231 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 147/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0298 - mean_absolute_percentage_error: 33.1622 - cosine_proximity: -9.6368e-01 - val_loss: 0.3725 - val_mean_squared_error: 0.3725 - val_mean_absolute_error: 0.5487 - val_mean_absolute_percentage_error: 33.3644 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 148/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0297 - mean_absolute_percentage_error: 33.1213 - cosine_proximity: -9.6368e-01 - val_loss: 0.3714 - val_mean_squared_error: 0.3714 - val_mean_absolute_error: 0.5478 - val_mean_absolute_percentage_error: 33.3055 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 149/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0297 - mean_absolute_percentage_error: 33.0705 - cosine_proximity: -9.6368e-01 - val_loss: 0.3703 - val_mean_squared_error: 0.3703 - val_mean_absolute_error: 0.5469 - val_mean_absolute_percentage_error: 33.2479 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 150/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0296 - mean_absolute_percentage_error: 33.0323 - cosine_proximity: -9.6368e-01 - val_loss: 0.3693 - val_mean_squared_error: 0.3693 - val_mean_absolute_error: 0.5460 - val_mean_absolute_percentage_error: 33.1893 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 151/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0296 - mean_absolute_percentage_error: 32.9747 - cosine_proximity: -9.6368e-01 - val_loss: 0.3682 - val_mean_squared_error: 0.3682 - val_mean_absolute_error: 0.5452 - val_mean_absolute_percentage_error: 33.1331 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 152/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0296 - mean_absolute_percentage_error: 32.9438 - cosine_proximity: -9.6368e-01 - val_loss: 0.3672 - val_mean_squared_error: 0.3672 - val_mean_absolute_error: 0.5443 - val_mean_absolute_percentage_error: 33.0746 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 153/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0295 - mean_absolute_percentage_error: 32.8725 - cosine_proximity: -9.6368e-01 - val_loss: 0.3661 - val_mean_squared_error: 0.3661 - val_mean_absolute_error: 0.5435 - val_mean_absolute_percentage_error: 33.0204 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 154/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0295 - mean_absolute_percentage_error: 32.8590 - cosine_proximity: -9.6443e-01 - val_loss: 0.3651 - val_mean_squared_error: 0.3651 - val_mean_absolute_error: 0.5426 - val_mean_absolute_percentage_error: 32.9609 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 155/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0295 - mean_absolute_percentage_error: 32.7582 - cosine_proximity: -9.6443e-01 - val_loss: 0.3641 - val_mean_squared_error: 0.3641 - val_mean_absolute_error: 0.5418 - val_mean_absolute_percentage_error: 32.9101 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 156/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 32.7866 - cosine_proximity: -9.6443e-01 - val_loss: 0.3630 - val_mean_squared_error: 0.3630 - val_mean_absolute_error: 0.5409 - val_mean_absolute_percentage_error: 32.8474 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 157/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 32.6186 - cosine_proximity: -9.6443e-01 - val_loss: 0.3621 - val_mean_squared_error: 0.3621 - val_mean_absolute_error: 0.5402 - val_mean_absolute_percentage_error: 32.8033 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 158/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 32.7478 - cosine_proximity: -9.6443e-01 - val_loss: 0.3609 - val_mean_squared_error: 0.3609 - val_mean_absolute_error: 0.5391 - val_mean_absolute_percentage_error: 32.7323 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 159/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 32.4240 - cosine_proximity: -9.6519e-01 - val_loss: 0.3602 - val_mean_squared_error: 0.3602 - val_mean_absolute_error: 0.5386 - val_mean_absolute_percentage_error: 32.7030 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 160/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 32.7981 - cosine_proximity: -9.6443e-01 - val_loss: 0.3588 - val_mean_squared_error: 0.3588 - val_mean_absolute_error: 0.5373 - val_mean_absolute_percentage_error: 32.6109 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 161/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 32.0976 - cosine_proximity: -9.6519e-01 - val_loss: 0.3585 - val_mean_squared_error: 0.3585 - val_mean_absolute_error: 0.5372 - val_mean_absolute_percentage_error: 32.6161 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 162/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0292 - mean_absolute_percentage_error: 33.0843 - cosine_proximity: -9.6443e-01 - val_loss: 0.3564 - val_mean_squared_error: 0.3564 - val_mean_absolute_error: 0.5353 - val_mean_absolute_percentage_error: 32.4717 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 163/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 31.4702 - cosine_proximity: -9.6443e-01 - val_loss: 0.3571 - val_mean_squared_error: 0.3571 - val_mean_absolute_error: 0.5363 - val_mean_absolute_percentage_error: 32.5611 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 164/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 34.1716 - cosine_proximity: -9.6216e-01 - val_loss: 0.3535 - val_mean_squared_error: 0.3535 - val_mean_absolute_error: 0.5326 - val_mean_absolute_percentage_error: 32.2840 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 165/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0299 - mean_absolute_percentage_error: 30.2573 - cosine_proximity: -9.6746e-01 - val_loss: 0.3567 - val_mean_squared_error: 0.3567 - val_mean_absolute_error: 0.5363 - val_mean_absolute_percentage_error: 32.5852 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 166/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0311 - mean_absolute_percentage_error: 38.2064 - cosine_proximity: -9.5914e-01 - val_loss: 0.3491 - val_mean_squared_error: 0.3491 - val_mean_absolute_error: 0.5283 - val_mean_absolute_percentage_error: 31.9667 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 167/250\n",
      "0s - loss: 0.0021 - mean_squared_error: 0.0021 - mean_absolute_error: 0.0339 - mean_absolute_percentage_error: 32.1731 - cosine_proximity: -9.6065e-01 - val_loss: 0.3585 - val_mean_squared_error: 0.3585 - val_mean_absolute_error: 0.5388 - val_mean_absolute_percentage_error: 32.8126 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 168/250\n",
      "0s - loss: 0.0028 - mean_squared_error: 0.0028 - mean_absolute_error: 0.0399 - mean_absolute_percentage_error: 54.9234 - cosine_proximity: -9.4552e-01 - val_loss: 0.3401 - val_mean_squared_error: 0.3401 - val_mean_absolute_error: 0.5192 - val_mean_absolute_percentage_error: 31.2888 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 169/250\n",
      "0s - loss: 0.0045 - mean_squared_error: 0.0045 - mean_absolute_error: 0.0524 - mean_absolute_percentage_error: 65.2920 - cosine_proximity: -9.3719e-01 - val_loss: 0.3657 - val_mean_squared_error: 0.3657 - val_mean_absolute_error: 0.5477 - val_mean_absolute_percentage_error: 33.5626 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 170/250\n",
      "0s - loss: 0.0081 - mean_squared_error: 0.0081 - mean_absolute_error: 0.0698 - mean_absolute_percentage_error: 102.4741 - cosine_proximity: -9.0919e-01 - val_loss: 0.3186 - val_mean_squared_error: 0.3186 - val_mean_absolute_error: 0.4972 - val_mean_absolute_percentage_error: 29.7192 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 171/250\n",
      "0s - loss: 0.0147 - mean_squared_error: 0.0147 - mean_absolute_error: 0.0949 - mean_absolute_percentage_error: 123.1233 - cosine_proximity: -9.2509e-01 - val_loss: 0.3820 - val_mean_squared_error: 0.3820 - val_mean_absolute_error: 0.5675 - val_mean_absolute_percentage_error: 35.2878 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 172/250\n",
      "0s - loss: 0.0208 - mean_squared_error: 0.0208 - mean_absolute_error: 0.1088 - mean_absolute_percentage_error: 137.3698 - cosine_proximity: -8.8120e-01 - val_loss: 0.2803 - val_mean_squared_error: 0.2803 - val_mean_absolute_error: 0.4611 - val_mean_absolute_percentage_error: 27.4844 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 173/250\n",
      "0s - loss: 0.0227 - mean_squared_error: 0.0227 - mean_absolute_error: 0.1140 - mean_absolute_percentage_error: 113.4499 - cosine_proximity: -9.2887e-01 - val_loss: 0.3607 - val_mean_squared_error: 0.3607 - val_mean_absolute_error: 0.5517 - val_mean_absolute_percentage_error: 34.3559 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 174/250\n",
      "0s - loss: 0.0167 - mean_squared_error: 0.0167 - mean_absolute_error: 0.0924 - mean_absolute_percentage_error: 76.1327 - cosine_proximity: -9.3492e-01 - val_loss: 0.2740 - val_mean_squared_error: 0.2740 - val_mean_absolute_error: 0.4576 - val_mean_absolute_percentage_error: 27.2530 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 175/250\n",
      "0s - loss: 0.0066 - mean_squared_error: 0.0066 - mean_absolute_error: 0.0620 - mean_absolute_percentage_error: 39.3943 - cosine_proximity: -9.6065e-01 - val_loss: 0.3034 - val_mean_squared_error: 0.3034 - val_mean_absolute_error: 0.4927 - val_mean_absolute_percentage_error: 29.8692 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 176/250\n",
      "0s - loss: 0.0024 - mean_squared_error: 0.0024 - mean_absolute_error: 0.0355 - mean_absolute_percentage_error: 31.1155 - cosine_proximity: -9.6670e-01 - val_loss: 0.3036 - val_mean_squared_error: 0.3036 - val_mean_absolute_error: 0.4916 - val_mean_absolute_percentage_error: 29.7140 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 177/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0307 - mean_absolute_percentage_error: 34.5792 - cosine_proximity: -9.5838e-01 - val_loss: 0.3048 - val_mean_squared_error: 0.3048 - val_mean_absolute_error: 0.4920 - val_mean_absolute_percentage_error: 29.7012 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 178/250\n",
      "0s - loss: 0.0018 - mean_squared_error: 0.0018 - mean_absolute_error: 0.0315 - mean_absolute_percentage_error: 29.8486 - cosine_proximity: -9.6746e-01 - val_loss: 0.3150 - val_mean_squared_error: 0.3150 - val_mean_absolute_error: 0.5031 - val_mean_absolute_percentage_error: 30.5353 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 179/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0301 - mean_absolute_percentage_error: 31.0776 - cosine_proximity: -9.6292e-01 - val_loss: 0.3129 - val_mean_squared_error: 0.3129 - val_mean_absolute_error: 0.5003 - val_mean_absolute_percentage_error: 30.2933 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 180/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0301 - mean_absolute_percentage_error: 30.9244 - cosine_proximity: -9.6368e-01 - val_loss: 0.3147 - val_mean_squared_error: 0.3147 - val_mean_absolute_error: 0.5022 - val_mean_absolute_percentage_error: 30.4333 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 181/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0298 - mean_absolute_percentage_error: 30.4662 - cosine_proximity: -9.6443e-01 - val_loss: 0.3148 - val_mean_squared_error: 0.3148 - val_mean_absolute_error: 0.5022 - val_mean_absolute_percentage_error: 30.4334 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 182/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0297 - mean_absolute_percentage_error: 30.8641 - cosine_proximity: -9.6368e-01 - val_loss: 0.3143 - val_mean_squared_error: 0.3143 - val_mean_absolute_error: 0.5015 - val_mean_absolute_percentage_error: 30.3769 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 183/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0297 - mean_absolute_percentage_error: 30.6699 - cosine_proximity: -9.6443e-01 - val_loss: 0.3148 - val_mean_squared_error: 0.3148 - val_mean_absolute_error: 0.5020 - val_mean_absolute_percentage_error: 30.4084 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 184/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0296 - mean_absolute_percentage_error: 30.7468 - cosine_proximity: -9.6368e-01 - val_loss: 0.3148 - val_mean_squared_error: 0.3148 - val_mean_absolute_error: 0.5018 - val_mean_absolute_percentage_error: 30.3899 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 185/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0296 - mean_absolute_percentage_error: 30.7580 - cosine_proximity: -9.6368e-01 - val_loss: 0.3151 - val_mean_squared_error: 0.3151 - val_mean_absolute_error: 0.5020 - val_mean_absolute_percentage_error: 30.4027 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 186/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0295 - mean_absolute_percentage_error: 30.7497 - cosine_proximity: -9.6519e-01 - val_loss: 0.3154 - val_mean_squared_error: 0.3154 - val_mean_absolute_error: 0.5022 - val_mean_absolute_percentage_error: 30.4106 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 187/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0295 - mean_absolute_percentage_error: 30.7798 - cosine_proximity: -9.6595e-01 - val_loss: 0.3156 - val_mean_squared_error: 0.3156 - val_mean_absolute_error: 0.5024 - val_mean_absolute_percentage_error: 30.4162 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 188/250\n",
      "0s - loss: 0.0017 - mean_squared_error: 0.0017 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 30.7773 - cosine_proximity: -9.6595e-01 - val_loss: 0.3159 - val_mean_squared_error: 0.3159 - val_mean_absolute_error: 0.5025 - val_mean_absolute_percentage_error: 30.4250 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 189/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 30.7928 - cosine_proximity: -9.6519e-01 - val_loss: 0.3161 - val_mean_squared_error: 0.3161 - val_mean_absolute_error: 0.5026 - val_mean_absolute_percentage_error: 30.4291 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 190/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0294 - mean_absolute_percentage_error: 30.7985 - cosine_proximity: -9.6595e-01 - val_loss: 0.3162 - val_mean_squared_error: 0.3162 - val_mean_absolute_error: 0.5027 - val_mean_absolute_percentage_error: 30.4338 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 191/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 30.8055 - cosine_proximity: -9.6595e-01 - val_loss: 0.3164 - val_mean_squared_error: 0.3164 - val_mean_absolute_error: 0.5028 - val_mean_absolute_percentage_error: 30.4362 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 192/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 30.8124 - cosine_proximity: -9.6595e-01 - val_loss: 0.3165 - val_mean_squared_error: 0.3165 - val_mean_absolute_error: 0.5029 - val_mean_absolute_percentage_error: 30.4376 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 193/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0293 - mean_absolute_percentage_error: 30.8169 - cosine_proximity: -9.6595e-01 - val_loss: 0.3166 - val_mean_squared_error: 0.3166 - val_mean_absolute_error: 0.5029 - val_mean_absolute_percentage_error: 30.4379 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 194/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0292 - mean_absolute_percentage_error: 30.8212 - cosine_proximity: -9.6595e-01 - val_loss: 0.3167 - val_mean_squared_error: 0.3167 - val_mean_absolute_error: 0.5029 - val_mean_absolute_percentage_error: 30.4369 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 195/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0292 - mean_absolute_percentage_error: 30.8235 - cosine_proximity: -9.6595e-01 - val_loss: 0.3168 - val_mean_squared_error: 0.3168 - val_mean_absolute_error: 0.5030 - val_mean_absolute_percentage_error: 30.4350 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 196/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0291 - mean_absolute_percentage_error: 30.8247 - cosine_proximity: -9.6595e-01 - val_loss: 0.3168 - val_mean_squared_error: 0.3168 - val_mean_absolute_error: 0.5029 - val_mean_absolute_percentage_error: 30.4319 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 197/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0291 - mean_absolute_percentage_error: 30.8241 - cosine_proximity: -9.6595e-01 - val_loss: 0.3168 - val_mean_squared_error: 0.3168 - val_mean_absolute_error: 0.5029 - val_mean_absolute_percentage_error: 30.4279 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 198/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0291 - mean_absolute_percentage_error: 30.8222 - cosine_proximity: -9.6595e-01 - val_loss: 0.3168 - val_mean_squared_error: 0.3168 - val_mean_absolute_error: 0.5029 - val_mean_absolute_percentage_error: 30.4230 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 199/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0290 - mean_absolute_percentage_error: 30.8188 - cosine_proximity: -9.6595e-01 - val_loss: 0.3168 - val_mean_squared_error: 0.3168 - val_mean_absolute_error: 0.5028 - val_mean_absolute_percentage_error: 30.4171 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 200/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0290 - mean_absolute_percentage_error: 30.8141 - cosine_proximity: -9.6595e-01 - val_loss: 0.3168 - val_mean_squared_error: 0.3168 - val_mean_absolute_error: 0.5027 - val_mean_absolute_percentage_error: 30.4104 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 201/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0290 - mean_absolute_percentage_error: 30.8080 - cosine_proximity: -9.6595e-01 - val_loss: 0.3167 - val_mean_squared_error: 0.3167 - val_mean_absolute_error: 0.5026 - val_mean_absolute_percentage_error: 30.4029 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 202/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0290 - mean_absolute_percentage_error: 30.8005 - cosine_proximity: -9.6595e-01 - val_loss: 0.3167 - val_mean_squared_error: 0.3167 - val_mean_absolute_error: 0.5025 - val_mean_absolute_percentage_error: 30.3946 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 203/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0289 - mean_absolute_percentage_error: 30.7916 - cosine_proximity: -9.6595e-01 - val_loss: 0.3166 - val_mean_squared_error: 0.3166 - val_mean_absolute_error: 0.5024 - val_mean_absolute_percentage_error: 30.3857 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 204/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0289 - mean_absolute_percentage_error: 30.7812 - cosine_proximity: -9.6595e-01 - val_loss: 0.3165 - val_mean_squared_error: 0.3165 - val_mean_absolute_error: 0.5023 - val_mean_absolute_percentage_error: 30.3760 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 205/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0289 - mean_absolute_percentage_error: 30.7695 - cosine_proximity: -9.6595e-01 - val_loss: 0.3164 - val_mean_squared_error: 0.3164 - val_mean_absolute_error: 0.5022 - val_mean_absolute_percentage_error: 30.3657 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 206/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0288 - mean_absolute_percentage_error: 30.7564 - cosine_proximity: -9.6595e-01 - val_loss: 0.3163 - val_mean_squared_error: 0.3163 - val_mean_absolute_error: 0.5020 - val_mean_absolute_percentage_error: 30.3547 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 207/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0288 - mean_absolute_percentage_error: 30.7419 - cosine_proximity: -9.6595e-01 - val_loss: 0.3161 - val_mean_squared_error: 0.3161 - val_mean_absolute_error: 0.5019 - val_mean_absolute_percentage_error: 30.3432 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 208/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0288 - mean_absolute_percentage_error: 30.7261 - cosine_proximity: -9.6595e-01 - val_loss: 0.3160 - val_mean_squared_error: 0.3160 - val_mean_absolute_error: 0.5017 - val_mean_absolute_percentage_error: 30.3311 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 209/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0288 - mean_absolute_percentage_error: 30.7090 - cosine_proximity: -9.6595e-01 - val_loss: 0.3158 - val_mean_squared_error: 0.3158 - val_mean_absolute_error: 0.5015 - val_mean_absolute_percentage_error: 30.3185 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 210/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0287 - mean_absolute_percentage_error: 30.6906 - cosine_proximity: -9.6670e-01 - val_loss: 0.3156 - val_mean_squared_error: 0.3156 - val_mean_absolute_error: 0.5014 - val_mean_absolute_percentage_error: 30.3053 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 211/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0287 - mean_absolute_percentage_error: 30.6710 - cosine_proximity: -9.6670e-01 - val_loss: 0.3154 - val_mean_squared_error: 0.3154 - val_mean_absolute_error: 0.5012 - val_mean_absolute_percentage_error: 30.2917 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 212/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0287 - mean_absolute_percentage_error: 30.6501 - cosine_proximity: -9.6670e-01 - val_loss: 0.3152 - val_mean_squared_error: 0.3152 - val_mean_absolute_error: 0.5010 - val_mean_absolute_percentage_error: 30.2777 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 213/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0287 - mean_absolute_percentage_error: 30.6281 - cosine_proximity: -9.6670e-01 - val_loss: 0.3150 - val_mean_squared_error: 0.3150 - val_mean_absolute_error: 0.5008 - val_mean_absolute_percentage_error: 30.2632 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 214/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0286 - mean_absolute_percentage_error: 30.6048 - cosine_proximity: -9.6670e-01 - val_loss: 0.3148 - val_mean_squared_error: 0.3148 - val_mean_absolute_error: 0.5006 - val_mean_absolute_percentage_error: 30.2483 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 215/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0286 - mean_absolute_percentage_error: 30.5805 - cosine_proximity: -9.6670e-01 - val_loss: 0.3146 - val_mean_squared_error: 0.3146 - val_mean_absolute_error: 0.5003 - val_mean_absolute_percentage_error: 30.2330 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 216/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0286 - mean_absolute_percentage_error: 30.5551 - cosine_proximity: -9.6670e-01 - val_loss: 0.3144 - val_mean_squared_error: 0.3144 - val_mean_absolute_error: 0.5001 - val_mean_absolute_percentage_error: 30.2173 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 217/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0286 - mean_absolute_percentage_error: 30.5287 - cosine_proximity: -9.6670e-01 - val_loss: 0.3141 - val_mean_squared_error: 0.3141 - val_mean_absolute_error: 0.4999 - val_mean_absolute_percentage_error: 30.2013 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 218/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0285 - mean_absolute_percentage_error: 30.5012 - cosine_proximity: -9.6670e-01 - val_loss: 0.3139 - val_mean_squared_error: 0.3139 - val_mean_absolute_error: 0.4996 - val_mean_absolute_percentage_error: 30.1850 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 219/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0285 - mean_absolute_percentage_error: 30.4728 - cosine_proximity: -9.6670e-01 - val_loss: 0.3136 - val_mean_squared_error: 0.3136 - val_mean_absolute_error: 0.4994 - val_mean_absolute_percentage_error: 30.1684 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 220/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0285 - mean_absolute_percentage_error: 30.4433 - cosine_proximity: -9.6595e-01 - val_loss: 0.3134 - val_mean_squared_error: 0.3134 - val_mean_absolute_error: 0.4992 - val_mean_absolute_percentage_error: 30.1515 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 221/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0285 - mean_absolute_percentage_error: 30.4130 - cosine_proximity: -9.6595e-01 - val_loss: 0.3131 - val_mean_squared_error: 0.3131 - val_mean_absolute_error: 0.4989 - val_mean_absolute_percentage_error: 30.1342 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 222/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0284 - mean_absolute_percentage_error: 30.3818 - cosine_proximity: -9.6595e-01 - val_loss: 0.3128 - val_mean_squared_error: 0.3128 - val_mean_absolute_error: 0.4986 - val_mean_absolute_percentage_error: 30.1167 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 223/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0284 - mean_absolute_percentage_error: 30.3498 - cosine_proximity: -9.6595e-01 - val_loss: 0.3126 - val_mean_squared_error: 0.3126 - val_mean_absolute_error: 0.4984 - val_mean_absolute_percentage_error: 30.0989 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 224/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0284 - mean_absolute_percentage_error: 30.3169 - cosine_proximity: -9.6595e-01 - val_loss: 0.3123 - val_mean_squared_error: 0.3123 - val_mean_absolute_error: 0.4981 - val_mean_absolute_percentage_error: 30.0809 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 225/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0284 - mean_absolute_percentage_error: 30.2833 - cosine_proximity: -9.6595e-01 - val_loss: 0.3120 - val_mean_squared_error: 0.3120 - val_mean_absolute_error: 0.4978 - val_mean_absolute_percentage_error: 30.0626 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 226/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0283 - mean_absolute_percentage_error: 30.2488 - cosine_proximity: -9.6595e-01 - val_loss: 0.3117 - val_mean_squared_error: 0.3117 - val_mean_absolute_error: 0.4976 - val_mean_absolute_percentage_error: 30.0442 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 227/250\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0283 - mean_absolute_percentage_error: 30.2137 - cosine_proximity: -9.6595e-01 - val_loss: 0.3114 - val_mean_squared_error: 0.3114 - val_mean_absolute_error: 0.4973 - val_mean_absolute_percentage_error: 30.0254 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 228/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0283 - mean_absolute_percentage_error: 30.1778 - cosine_proximity: -9.6595e-01 - val_loss: 0.3111 - val_mean_squared_error: 0.3111 - val_mean_absolute_error: 0.4970 - val_mean_absolute_percentage_error: 30.0065 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 229/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0283 - mean_absolute_percentage_error: 30.1413 - cosine_proximity: -9.6595e-01 - val_loss: 0.3108 - val_mean_squared_error: 0.3108 - val_mean_absolute_error: 0.4967 - val_mean_absolute_percentage_error: 29.9874 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 230/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0283 - mean_absolute_percentage_error: 30.1040 - cosine_proximity: -9.6595e-01 - val_loss: 0.3105 - val_mean_squared_error: 0.3105 - val_mean_absolute_error: 0.4964 - val_mean_absolute_percentage_error: 29.9681 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 231/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0282 - mean_absolute_percentage_error: 30.0664 - cosine_proximity: -9.6595e-01 - val_loss: 0.3102 - val_mean_squared_error: 0.3102 - val_mean_absolute_error: 0.4961 - val_mean_absolute_percentage_error: 29.9486 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 232/250\n",
      "0s - loss: 0.0016 - mean_squared_error: 0.0016 - mean_absolute_error: 0.0282 - mean_absolute_percentage_error: 30.0279 - cosine_proximity: -9.6595e-01 - val_loss: 0.3098 - val_mean_squared_error: 0.3098 - val_mean_absolute_error: 0.4958 - val_mean_absolute_percentage_error: 29.9290 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 233/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0282 - mean_absolute_percentage_error: 29.9893 - cosine_proximity: -9.6595e-01 - val_loss: 0.3095 - val_mean_squared_error: 0.3095 - val_mean_absolute_error: 0.4955 - val_mean_absolute_percentage_error: 29.9092 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 234/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0282 - mean_absolute_percentage_error: 29.9496 - cosine_proximity: -9.6595e-01 - val_loss: 0.3092 - val_mean_squared_error: 0.3092 - val_mean_absolute_error: 0.4952 - val_mean_absolute_percentage_error: 29.8893 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 235/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0281 - mean_absolute_percentage_error: 29.9101 - cosine_proximity: -9.6519e-01 - val_loss: 0.3089 - val_mean_squared_error: 0.3089 - val_mean_absolute_error: 0.4949 - val_mean_absolute_percentage_error: 29.8691 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 236/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0281 - mean_absolute_percentage_error: 29.8693 - cosine_proximity: -9.6595e-01 - val_loss: 0.3085 - val_mean_squared_error: 0.3085 - val_mean_absolute_error: 0.4946 - val_mean_absolute_percentage_error: 29.8490 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 237/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0281 - mean_absolute_percentage_error: 29.8291 - cosine_proximity: -9.6595e-01 - val_loss: 0.3082 - val_mean_squared_error: 0.3082 - val_mean_absolute_error: 0.4943 - val_mean_absolute_percentage_error: 29.8286 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 238/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0281 - mean_absolute_percentage_error: 29.7871 - cosine_proximity: -9.6595e-01 - val_loss: 0.3078 - val_mean_squared_error: 0.3078 - val_mean_absolute_error: 0.4940 - val_mean_absolute_percentage_error: 29.8082 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 239/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0281 - mean_absolute_percentage_error: 29.7464 - cosine_proximity: -9.6595e-01 - val_loss: 0.3075 - val_mean_squared_error: 0.3075 - val_mean_absolute_error: 0.4937 - val_mean_absolute_percentage_error: 29.7875 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 240/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0280 - mean_absolute_percentage_error: 29.7032 - cosine_proximity: -9.6595e-01 - val_loss: 0.3072 - val_mean_squared_error: 0.3072 - val_mean_absolute_error: 0.4934 - val_mean_absolute_percentage_error: 29.7670 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 241/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0280 - mean_absolute_percentage_error: 29.6625 - cosine_proximity: -9.6595e-01 - val_loss: 0.3068 - val_mean_squared_error: 0.3068 - val_mean_absolute_error: 0.4931 - val_mean_absolute_percentage_error: 29.7461 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 242/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0280 - mean_absolute_percentage_error: 29.6179 - cosine_proximity: -9.6595e-01 - val_loss: 0.3065 - val_mean_squared_error: 0.3065 - val_mean_absolute_error: 0.4928 - val_mean_absolute_percentage_error: 29.7255 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 243/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0280 - mean_absolute_percentage_error: 29.5778 - cosine_proximity: -9.6595e-01 - val_loss: 0.3061 - val_mean_squared_error: 0.3061 - val_mean_absolute_error: 0.4924 - val_mean_absolute_percentage_error: 29.7042 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 244/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0280 - mean_absolute_percentage_error: 29.5309 - cosine_proximity: -9.6595e-01 - val_loss: 0.3058 - val_mean_squared_error: 0.3058 - val_mean_absolute_error: 0.4921 - val_mean_absolute_percentage_error: 29.6836 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 245/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0280 - mean_absolute_percentage_error: 29.4924 - cosine_proximity: -9.6595e-01 - val_loss: 0.3054 - val_mean_squared_error: 0.3054 - val_mean_absolute_error: 0.4918 - val_mean_absolute_percentage_error: 29.6620 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 246/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0279 - mean_absolute_percentage_error: 29.4419 - cosine_proximity: -9.6595e-01 - val_loss: 0.3051 - val_mean_squared_error: 0.3051 - val_mean_absolute_error: 0.4915 - val_mean_absolute_percentage_error: 29.6416 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 247/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0279 - mean_absolute_percentage_error: 29.4072 - cosine_proximity: -9.6595e-01 - val_loss: 0.3047 - val_mean_squared_error: 0.3047 - val_mean_absolute_error: 0.4911 - val_mean_absolute_percentage_error: 29.6194 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 248/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0279 - mean_absolute_percentage_error: 29.3500 - cosine_proximity: -9.6595e-01 - val_loss: 0.3044 - val_mean_squared_error: 0.3044 - val_mean_absolute_error: 0.4908 - val_mean_absolute_percentage_error: 29.5996 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 249/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0279 - mean_absolute_percentage_error: 29.3239 - cosine_proximity: -9.6595e-01 - val_loss: 0.3040 - val_mean_squared_error: 0.3040 - val_mean_absolute_error: 0.4905 - val_mean_absolute_percentage_error: 29.5764 - val_cosine_proximity: -1.0000e+00\n",
      "Epoch 250/250\n",
      "0s - loss: 0.0015 - mean_squared_error: 0.0015 - mean_absolute_error: 0.0279 - mean_absolute_percentage_error: 29.2532 - cosine_proximity: -9.6595e-01 - val_loss: 0.3036 - val_mean_squared_error: 0.3036 - val_mean_absolute_error: 0.4902 - val_mean_absolute_percentage_error: 29.5578 - val_cosine_proximity: -1.0000e+00\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(train_X, train_y, epochs=250, batch_size=252, validation_data=(test_X, test_y), verbose=2, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD8CAYAAABw1c+bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niEiBUeIXESkwSvwiIgVGiV9EpMD8f3llrzPwBBEaAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x203b70893c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "# plot history\n",
    "pyplot.plot(history.history['loss'], label='train')\n",
    "pyplot.plot(history.history['val_loss'], label='test')\n",
    "pyplot.legend()\n",
    "pyplot.show() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x203b77c73c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot metrics\n",
    "pyplot.plot(history.history['mean_squared_error'], label='MSE')\n",
    "pyplot.plot(history.history['mean_absolute_error'],label='MAE')\n",
    "#pyplot.plot(history.history['mean_absolute_percentage_error'], label='MAPE')\n",
    "#pyplot.plot(history.history['cosine_proximity'],label='CP')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# make a prediction\n",
    "yhat = model.predict(test_X)\n",
    "test_X = test_X.reshape((test_X.shape[0], test_X.shape[2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from numpy import concatenate\n",
    "# invert scaling for forecast\n",
    "inv_yhat = concatenate((yhat, test_X[:, 1:]), axis=1)\n",
    "inv_yhat = scaler.inverse_transform(inv_yhat)\n",
    "inv_yhat = inv_yhat[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# invert scaling for actual\n",
    "test_y = test_y.reshape((len(test_y), 1))\n",
    "inv_y = concatenate((test_y, test_X[:, 1:]), axis=1)\n",
    "inv_y = scaler.inverse_transform(inv_y)\n",
    "inv_y = inv_y[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test RMSE: 208.120\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD8CAYAAABw1c+bAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nY0yG+X/3NbS20lREywAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x203b70d28d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from math import sqrt\n",
    "from sklearn.metrics import mean_squared_error\n",
    "# calculate RMSE\n",
    "rmse = sqrt(mean_squared_error(inv_y, inv_yhat))\n",
    "print('Test RMSE: %.3f' % rmse)\n",
    "pyplot.plot(test_y, label='Values')\n",
    "pyplot.plot(yhat, label='Prediction')\n",
    "pyplot.legend()\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}