fixed-time-horizon-prediction/Exp_1.2_Bondville_2hour.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import datetime\n",
    "import glob\n",
    "import os.path\n",
    "from pandas.compat import StringIO"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### NREL Bird Model implementation: for obtaining clear sky GHI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import itertools\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "sns.set_color_codes()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pvlib\n",
    "from pvlib import clearsky, atmosphere\n",
    "from pvlib.location import Location"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "bvl = Location(40.1134,-88.3695, 'US/Central', 217.932, 'Bondville')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "times2009 = pd.DatetimeIndex(start='2009-01-01', end='2010-01-01', freq='1min',\n",
    "                        tz=bvl.tz)   # 12 months of 2009 - For testing\n",
    "times2010and2011 = pd.DatetimeIndex(start='2010-01-01', end='2012-01-01', freq='1min',\n",
    "                        tz=bvl.tz)   # 24 months of 2010 and 2011 - For training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_2009 = bvl.get_clearsky(times2009) \n",
    "cs_2010and2011 = bvl.get_clearsky(times2010and2011) # ineichen with climatology table by default\n",
    "#cs_2011 = bvl.get_clearsky(times2011) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_2009.drop(['dni','dhi'],axis=1, inplace=True) #updating the same dataframe by dropping two columns\n",
    "cs_2010and2011.drop(['dni','dhi'],axis=1, inplace=True) #updating the same dataframe by dropping two columns\n",
    "#cs_2011.drop(['dni','dhi'],axis=1, inplace=True) #updating the same dataframe by dropping two columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_2009.reset_index(inplace=True)\n",
    "cs_2010and2011.reset_index(inplace=True)\n",
    "#cs_2011.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_2009['index']=cs_2009['index'].apply(lambda x:x.to_datetime())\n",
    "cs_2009['year'] = cs_2009['index'].apply(lambda x:x.year)\n",
    "cs_2009['month'] = cs_2009['index'].apply(lambda x:x.month)\n",
    "cs_2009['day'] = cs_2009['index'].apply(lambda x:x.day)\n",
    "cs_2009['hour'] = cs_2009['index'].apply(lambda x:x.hour)\n",
    "cs_2009['min'] = cs_2009['index'].apply(lambda x:x.minute)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_2010and2011['index']=cs_2010and2011['index'].apply(lambda x:x.to_datetime())\n",
    "cs_2010and2011['year'] = cs_2010and2011['index'].apply(lambda x:x.year)\n",
    "cs_2010and2011['month'] = cs_2010and2011['index'].apply(lambda x:x.month)\n",
    "cs_2010and2011['day'] = cs_2010and2011['index'].apply(lambda x:x.day)\n",
    "cs_2010and2011['hour'] = cs_2010and2011['index'].apply(lambda x:x.hour)\n",
    "cs_2010and2011['min'] = cs_2010and2011['index'].apply(lambda x:x.minute)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(525601, 7)\n",
      "(1051201, 7)\n"
     ]
    }
   ],
   "source": [
    "print(cs_2009.shape)\n",
    "print(cs_2010and2011.shape)\n",
    "#print(cs_2011.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_2009.drop(cs_2009.index[-1], inplace=True)\n",
    "cs_2010and2011.drop(cs_2010and2011.index[-1], inplace=True)\n",
    "#cs_2011.drop(cs_2011.index[-1], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(525600, 7)\n",
      "(1051200, 7)\n"
     ]
    }
   ],
   "source": [
    "print(cs_2009.shape)\n",
    "print(cs_2010and2011.shape)\n",
    "#print(cs_2011.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>ghi</th>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th>min</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2010-01-01 00:00:00-06:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010-01-01 00:01:00-06:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2010-01-01 00:02:00-06:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2010-01-01 00:03:00-06:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2010-01-01 00:04:00-06:00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2010</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      index  ghi  year  month  day  hour  min\n",
       "0 2010-01-01 00:00:00-06:00  0.0  2010      1    1     0    0\n",
       "1 2010-01-01 00:01:00-06:00  0.0  2010      1    1     0    1\n",
       "2 2010-01-01 00:02:00-06:00  0.0  2010      1    1     0    2\n",
       "3 2010-01-01 00:03:00-06:00  0.0  2010      1    1     0    3\n",
       "4 2010-01-01 00:04:00-06:00  0.0  2010      1    1     0    4"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cs_2010and2011.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import files from each year in a separate dataframe"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "\n",
    "- year            integer\t year, i.e., 1995\n",
    "- jday            integer\t Julian day (1 through 365 [or 366])\n",
    "- month           integer\t number of the month (1-12)\n",
    "- day             integer\t day of the month(1-31)\n",
    "- hour            integer\t hour of the day (0-23)\n",
    "- min             integer\t minute of the hour (0-59)\n",
    "- dt              real\t decimal time (hour.decimalminutes, e.g., 23.5 = 2330)\n",
    "- zen             real\t solar zenith angle (degrees)\n",
    "- dw_solar        real\t downwelling global solar (Watts m^-2)\n",
    "- uw_solar        real\t upwelling global solar (Watts m^-2)\n",
    "- direct_n        real\t direct-normal solar (Watts m^-2)\n",
    "- diffuse         real\t downwelling diffuse solar (Watts m^-2)\n",
    "- dw_ir           real\t downwelling thermal infrared (Watts m^-2)\n",
    "- dw_casetemp     real\t downwelling IR case temp. (K)\n",
    "- dw_dometemp     real\t downwelling IR dome temp. (K)\n",
    "- uw_ir           real\t upwelling thermal infrared (Watts m^-2)\n",
    "- uw_casetemp     real\t upwelling IR case temp. (K)\n",
    "- uw_dometemp     real\t upwelling IR dome temp. (K)\n",
    "- uvb             real\t global UVB (milliWatts m^-2)\n",
    "- par             real\t photosynthetically active radiation (Watts m^-2)\n",
    "- netsolar        real\t net solar (dw_solar - uw_solar) (Watts m^-2)\n",
    "- netir           real\t net infrared (dw_ir - uw_ir) (Watts m^-2)\n",
    "- totalnet        real\t net radiation (netsolar+netir) (Watts m^-2)\n",
    "- temp            real\t 10-meter air temperature (?C)\n",
    "- rh              real\t relative humidity (%)\n",
    "- windspd         real\t wind speed (ms^-1)\n",
    "- winddir         real\t wind direction (degrees, clockwise from north)\n",
    "- pressure        real\t station pressure (mb)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = ['year', 'jday', 'month', 'day','hour','min','dt','zen','dw_solar','dw_solar_QC','uw_solar',\n",
    "       'uw_solar_QC', 'direct_n','direct_n_QC','diffuse', 'diffuse_QC', 'dw_ir', 'dw_ir_QC', 'dw_casetemp',\n",
    "       'dw_casetemp_QC', 'dw_dometemp','dw_dometemp_QC','uw_ir', 'uw_ir_QC', 'uw_casetemp','uw_casetemp_QC',\n",
    "       'uw_dometemp','uw_dometemp_QC','uvb','uvb_QC','par','par_QC','netsolar','netsolar_QC','netir','netir_QC',\n",
    "       'totalnet','totalnet_QC','temp','temp_QC','rh','rh_QC','windspd','windspd_QC','winddir','winddir_QC',\n",
    "       'pressure','pressure_QC']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1050402, 48)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = r'.\\\\data\\\\Bondville\\\\Exp_1_train'\n",
    "all_files = glob.glob(path + \"/*.dat\")\n",
    "all_files.sort()\n",
    "\n",
    "df_big_train = pd.concat([pd.read_csv(f, skipinitialspace = True, quotechar = '\"',skiprows=(2),delimiter=' ', \n",
    "                 index_col=False,header=None, names=cols) for f in all_files],ignore_index=True)\n",
    "df_big_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(524936, 48)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = r'.\\\\data\\\\Bondville\\\\Exp_1_test'\n",
    "all_files = glob.glob(path + \"/*.dat\")\n",
    "all_files.sort()\n",
    "\n",
    "df_big_test = pd.concat((pd.read_csv(f, skipinitialspace = True, quotechar = '\"',skiprows=(2),delimiter=' ', \n",
    "                 index_col=False,header=None, names=cols) for f in all_files),ignore_index=True)\n",
    "df_big_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(19, 48)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_big_test[df_big_test['dw_solar']==-9999.9].shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Merging Clear Sky GHI And the big dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1050402, 50)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train = pd.merge(df_big_train, cs_2010and2011, on=['year','month','day','hour','min'])\n",
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(524936, 50)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test = pd.merge(df_big_test, cs_2009, on=['year','month','day','hour','min'])\n",
    "df_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train.drop(['index'],axis=1, inplace=True) #updating the same dataframe by dropping the index columns from clear sky model\n",
    "df_test.drop(['index'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1050402, 49)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Managing missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Resetting index\n",
    "df_train.reset_index(drop=True, inplace=True)\n",
    "df_test.reset_index(drop=True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dropping rows with two or more -9999.9 values in columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1046527, 49)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Step1: Get indices of all rows with 2 or more -999\n",
    "missing_data_indices = np.where((df_train <=-9999.9).apply(sum, axis=1)>=2)[0]\n",
    "# Step2: Drop those indices\n",
    "df_train.drop(missing_data_indices, axis=0, inplace=True)\n",
    "# Checking that the rows are dropped\n",
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(516826, 49)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "missing_data_indices_test = np.where((df_test <= -9999.9).apply(sum, axis=1)>=2)[0]\n",
    "df_test.drop(missing_data_indices_test, axis=0, inplace=True)\n",
    "df_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# For the rows with only one cell as -9999.9, replacing this cell with the mean of the column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### First resetting index after dropping rows in the previous part of the code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 2nd time - Reseting Index\n",
    "df_train.reset_index(drop=True, inplace=True)\n",
    "df_test.reset_index(drop=True, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "one_miss_train_idx = np.where((df_train <=-9999.9).apply(sum, axis=1)==1)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24739"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(one_miss_train_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1046527, 49)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\s269596\\AppData\\Local\\Continuum\\Anaconda3\\lib\\site-packages\\pandas\\core\\indexing.py:194: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  self._setitem_with_indexer(indexer, value)\n"
     ]
    }
   ],
   "source": [
    "col_names = df_train.columns\n",
    "from collections import defaultdict\n",
    "stats = defaultdict(int)\n",
    "total_single_missing_values = 0\n",
    "for name in col_names:\n",
    "    col_mean = df_train[~(df_train[name] == -9999.9)][name].mean()\n",
    "    missing_indices = np.where((df_train[name] == -9999.9))\n",
    "    stats[name] = len(missing_indices[0])\n",
    "    df_train[name].loc[missing_indices] = col_mean\n",
    "    total_single_missing_values += sum(df_train[name] == -9999.9)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_col_min = df_train.apply(min, axis=0)\n",
    "df_col_max = df_train.apply(max, axis =0)\n",
    "#print(df_col_min, df_col_max)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = np.where((df_train <=-9999.9).apply(sum, axis=1)==1)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# doing the same thing on test dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15271"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "one_miss_test_idx = np.where((df_test <=-9999.9).apply(sum, axis=1)==1)[0]\n",
    "len(one_miss_test_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\s269596\\AppData\\Local\\Continuum\\Anaconda3\\lib\\site-packages\\pandas\\core\\indexing.py:194: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  self._setitem_with_indexer(indexer, value)\n"
     ]
    }
   ],
   "source": [
    "col_names_test = df_test.columns\n",
    "from collections import defaultdict\n",
    "stats_test = defaultdict(int)\n",
    "total_single_missing_values_test = 0\n",
    "for name in col_names_test:\n",
    "    col_mean = df_test[~(df_test[name] == -9999.9)][name].mean()\n",
    "    missing_indices = np.where((df_test[name] == -9999.9))\n",
    "    stats_test[name] = len(missing_indices[0])\n",
    "    df_test[name].loc[missing_indices] = col_mean\n",
    "    total_single_missing_values_test += sum(df_test[name] == -9999.9)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = np.where((df_test <=-9999.9).apply(sum, axis=1)==1)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1046527, 49)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(516826, 49)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Exploratory Data Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "dw_solar_everyday = df_test.groupby(['jday'])['dw_solar'].mean()\n",
    "ghi_everyday = df_test.groupby(['jday'])['ghi'].mean()\n",
    "j_day = df_test['jday'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEgCAYAAACjEpTiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/VEIhwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "\n",
    "axes1 = fig.add_axes([0.1,0.1,0.8,0.8])\n",
    "#axes2 = fig.add_axes([0.1,0.1,0.8,0.8])\n",
    "\n",
    "axes1.scatter(j_day,dw_solar_everyday,label='Observed dw_solar',color='red')\n",
    "axes1.scatter(j_day, ghi_everyday, label='Clear Sky GHI',color='green')\n",
    "\n",
    "axes1.set_xlabel('Days')\n",
    "axes1.set_ylabel('Solar Irradiance (Watts /m^2)')\n",
    "axes1.set_title('Solar Irradiance - Test Year 2009')\n",
    "axes1.legend(loc='best')\n",
    "\n",
    "fig.savefig('Figure2a.png', bbox_inches = 'tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\s269596\\AppData\\Local\\Continuum\\Anaconda3\\lib\\site-packages\\matplotlib\\axes\\_axes.py:6462: UserWarning: The 'normed' kwarg is deprecated, and has been replaced by the 'density' kwarg.\n",
      "  warnings.warn(\"The 'normed' kwarg is deprecated, and has been \"\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbEAAAGpCAYAAAADVQhHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(x=dw_solar_everyday,y=ghi_everyday,kind='reg')\n",
    "#plt.title('observed dw_solar vs clear sky ghi')\n",
    "plt.xlabel('Observed global downwelling solar (Watts/m^2)')\n",
    "plt.ylabel('Clear Sky GHI (Watts/m^2)')\n",
    "plt.savefig('Figure3a.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### making the Kt (clear sky index at time t) column by first removing rows with ghi==0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_train[df_train['ghi']!=0]\n",
    "df_test = df_test[df_test['ghi']!=0]\n",
    "df_train['Kt'] = df_train['dw_solar']/df_train['ghi']\n",
    "df_test['Kt'] = df_test['dw_solar']/df_test['ghi']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train.reset_index(inplace=True)\n",
    "df_test.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test Kt max: 44687.07900300023\n",
      "test Kt min: -869.1104944675631\n",
      "test Kt mean: 2.7894944379552764\n",
      "\n",
      "\n",
      "train Kt max: 13966.236460411343\n",
      "train Kt min: -1176.7698469370364\n",
      "train Kt mean: 2.447186234312092\n"
     ]
    }
   ],
   "source": [
    "print(\"test Kt max: \"+str(df_test['Kt'].max()))\n",
    "print(\"test Kt min: \"+str(df_test['Kt'].min()))\n",
    "print(\"test Kt mean: \"+str(df_test['Kt'].mean()))\n",
    "print(\"\\n\")\n",
    "print(\"train Kt max: \"+str(df_train['Kt'].max()))\n",
    "print(\"train Kt min: \"+str(df_train['Kt'].min()))\n",
    "print(\"train Kt mean: \"+str(df_train['Kt'].mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xaad3c88>]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD3CAYAAAD10FRmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkMJfRMSD/j8yv4JKa9aICwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df_train['Kt'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xaadec50>]"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD3CAYAAAD10FRmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrIKEGdXiAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df_test['Kt'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train= df_train[df_train['Kt']< 5000]\n",
    "df_train= df_train[df_train['Kt']> -1000]\n",
    "df_test= df_test[df_test['Kt']< 5000]\n",
    "df_test= df_test[df_test['Kt']> -1000]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Group the data (train dataframe)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "zen = df_train.groupby(['year','month','day','hour'])['zen'].mean()\n",
    "dw_solar = df_train.groupby(['year','month','day','hour'])['dw_solar'].mean()\n",
    "uw_solar = df_train.groupby(['year','month','day','hour'])['uw_solar'].mean()\n",
    "direct_n = df_train.groupby(['year','month','day','hour'])['direct_n'].mean()\n",
    "diffuse = df_train.groupby(['year','month','day','hour'])['diffuse'].mean()\n",
    "dw_ir = df_train.groupby(['year','month','day','hour'])['dw_ir'].mean()\n",
    "dw_casetemp = df_train.groupby(['year','month','day','hour'])['dw_casetemp'].mean()\n",
    "dw_dometemp = df_train.groupby(['year','month','day','hour'])['dw_dometemp'].mean()\n",
    "uw_ir = df_train.groupby(['year','month','day','hour'])['uw_ir'].mean()\n",
    "uw_casetemp = df_train.groupby(['year','month','day','hour'])['uw_casetemp'].mean()\n",
    "uw_dometemp = df_train.groupby(['year','month','day','hour'])['uw_dometemp'].mean()\n",
    "uvb = df_train.groupby(['year','month','day','hour'])['uvb'].mean()\n",
    "par = df_train.groupby(['year','month','day','hour'])['par'].mean()\n",
    "netsolar = df_train.groupby(['year','month','day','hour'])['netsolar'].mean()\n",
    "netir = df_train.groupby(['year','month','day','hour'])['netir'].mean()\n",
    "totalnet = df_train.groupby(['year','month','day','hour'])['totalnet'].mean()\n",
    "temp = df_train.groupby(['year','month','day','hour'])['temp'].mean()\n",
    "rh = df_train.groupby(['year','month','day','hour'])['rh'].mean()\n",
    "windspd = df_train.groupby(['year','month','day','hour'])['windspd'].mean()\n",
    "winddir = df_train.groupby(['year','month','day','hour'])['winddir'].mean()\n",
    "pressure = df_train.groupby(['year','month','day','hour'])['pressure'].mean()\n",
    "ghi = df_train.groupby(['year','month','day','hour'])['ghi'].mean()\n",
    "Kt = df_train.groupby(['year','month','day','hour'])['Kt'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_new_train = pd.concat([zen,dw_solar,uw_solar,direct_n,diffuse,dw_ir,dw_casetemp,dw_dometemp,uw_ir,uw_casetemp,uw_dometemp,\n",
    "                    uvb,par,netsolar,netir,totalnet,temp,rh,windspd,winddir,pressure,ghi,Kt], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>zen</th>\n",
       "      <th>dw_solar</th>\n",
       "      <th>uw_solar</th>\n",
       "      <th>direct_n</th>\n",
       "      <th>diffuse</th>\n",
       "      <th>dw_ir</th>\n",
       "      <th>dw_casetemp</th>\n",
       "      <th>dw_dometemp</th>\n",
       "      <th>uw_ir</th>\n",
       "      <th>uw_casetemp</th>\n",
       "      <th>...</th>\n",
       "      <th>netsolar</th>\n",
       "      <th>netir</th>\n",
       "      <th>totalnet</th>\n",
       "      <th>temp</th>\n",
       "      <th>rh</th>\n",
       "      <th>windspd</th>\n",
       "      <th>winddir</th>\n",
       "      <th>pressure</th>\n",
       "      <th>ghi</th>\n",
       "      <th>Kt</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">2010.0</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1.0</th>\n",
       "      <th rowspan=\"5\" valign=\"top\">1.0</th>\n",
       "      <th>7.0</th>\n",
       "      <td>152.659762</td>\n",
       "      <td>-4.019048</td>\n",
       "      <td>0.026190</td>\n",
       "      <td>1.554762</td>\n",
       "      <td>-0.202381</td>\n",
       "      <td>182.345238</td>\n",
       "      <td>261.159524</td>\n",
       "      <td>260.650000</td>\n",
       "      <td>253.014286</td>\n",
       "      <td>259.557143</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-70.657143</td>\n",
       "      <td>-70.657143</td>\n",
       "      <td>-13.607143</td>\n",
       "      <td>83.883333</td>\n",
       "      <td>5.145238</td>\n",
       "      <td>293.504762</td>\n",
       "      <td>999.397619</td>\n",
       "      <td>61.060726</td>\n",
       "      <td>-0.069829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0</th>\n",
       "      <td>143.730167</td>\n",
       "      <td>-3.900000</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>1.731667</td>\n",
       "      <td>-0.200000</td>\n",
       "      <td>181.795000</td>\n",
       "      <td>260.780000</td>\n",
       "      <td>260.283333</td>\n",
       "      <td>251.493333</td>\n",
       "      <td>259.196667</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-69.693333</td>\n",
       "      <td>-69.693333</td>\n",
       "      <td>-13.956667</td>\n",
       "      <td>84.893333</td>\n",
       "      <td>4.805000</td>\n",
       "      <td>296.086667</td>\n",
       "      <td>999.803333</td>\n",
       "      <td>166.338212</td>\n",
       "      <td>-0.025197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0</th>\n",
       "      <td>132.476167</td>\n",
       "      <td>-3.825000</td>\n",
       "      <td>0.011667</td>\n",
       "      <td>1.841667</td>\n",
       "      <td>-0.200000</td>\n",
       "      <td>181.033333</td>\n",
       "      <td>260.256667</td>\n",
       "      <td>259.761667</td>\n",
       "      <td>249.923333</td>\n",
       "      <td>258.665000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-68.890000</td>\n",
       "      <td>-68.890000</td>\n",
       "      <td>-14.476667</td>\n",
       "      <td>85.781667</td>\n",
       "      <td>4.898333</td>\n",
       "      <td>295.981667</td>\n",
       "      <td>999.783333</td>\n",
       "      <td>306.455432</td>\n",
       "      <td>-0.012680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <td>121.024000</td>\n",
       "      <td>-3.793333</td>\n",
       "      <td>-0.003333</td>\n",
       "      <td>1.783333</td>\n",
       "      <td>-0.210000</td>\n",
       "      <td>183.566667</td>\n",
       "      <td>259.941667</td>\n",
       "      <td>259.446667</td>\n",
       "      <td>249.775000</td>\n",
       "      <td>258.315000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-66.203333</td>\n",
       "      <td>-66.203333</td>\n",
       "      <td>-14.826667</td>\n",
       "      <td>85.966667</td>\n",
       "      <td>5.083333</td>\n",
       "      <td>298.381667</td>\n",
       "      <td>999.968333</td>\n",
       "      <td>411.028556</td>\n",
       "      <td>-0.009267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11.0</th>\n",
       "      <td>109.724667</td>\n",
       "      <td>-3.703333</td>\n",
       "      <td>0.006667</td>\n",
       "      <td>1.758333</td>\n",
       "      <td>-0.210000</td>\n",
       "      <td>191.770000</td>\n",
       "      <td>259.800000</td>\n",
       "      <td>259.353333</td>\n",
       "      <td>250.373333</td>\n",
       "      <td>258.070000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-58.600000</td>\n",
       "      <td>-58.600000</td>\n",
       "      <td>-15.123333</td>\n",
       "      <td>86.431667</td>\n",
       "      <td>4.973333</td>\n",
       "      <td>299.090000</td>\n",
       "      <td>1000.193333</td>\n",
       "      <td>465.002071</td>\n",
       "      <td>-0.007967</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                              zen  dw_solar  uw_solar  direct_n   diffuse  \\\n",
       "year   month day hour                                                       \n",
       "2010.0 1.0   1.0 7.0   152.659762 -4.019048  0.026190  1.554762 -0.202381   \n",
       "                 8.0   143.730167 -3.900000  0.030000  1.731667 -0.200000   \n",
       "                 9.0   132.476167 -3.825000  0.011667  1.841667 -0.200000   \n",
       "                 10.0  121.024000 -3.793333 -0.003333  1.783333 -0.210000   \n",
       "                 11.0  109.724667 -3.703333  0.006667  1.758333 -0.210000   \n",
       "\n",
       "                            dw_ir  dw_casetemp  dw_dometemp       uw_ir  \\\n",
       "year   month day hour                                                     \n",
       "2010.0 1.0   1.0 7.0   182.345238   261.159524   260.650000  253.014286   \n",
       "                 8.0   181.795000   260.780000   260.283333  251.493333   \n",
       "                 9.0   181.033333   260.256667   259.761667  249.923333   \n",
       "                 10.0  183.566667   259.941667   259.446667  249.775000   \n",
       "                 11.0  191.770000   259.800000   259.353333  250.373333   \n",
       "\n",
       "                       uw_casetemp    ...     netsolar      netir   totalnet  \\\n",
       "year   month day hour                 ...                                      \n",
       "2010.0 1.0   1.0 7.0    259.557143    ...          0.0 -70.657143 -70.657143   \n",
       "                 8.0    259.196667    ...          0.0 -69.693333 -69.693333   \n",
       "                 9.0    258.665000    ...          0.0 -68.890000 -68.890000   \n",
       "                 10.0   258.315000    ...          0.0 -66.203333 -66.203333   \n",
       "                 11.0   258.070000    ...          0.0 -58.600000 -58.600000   \n",
       "\n",
       "                            temp         rh   windspd     winddir  \\\n",
       "year   month day hour                                               \n",
       "2010.0 1.0   1.0 7.0  -13.607143  83.883333  5.145238  293.504762   \n",
       "                 8.0  -13.956667  84.893333  4.805000  296.086667   \n",
       "                 9.0  -14.476667  85.781667  4.898333  295.981667   \n",
       "                 10.0 -14.826667  85.966667  5.083333  298.381667   \n",
       "                 11.0 -15.123333  86.431667  4.973333  299.090000   \n",
       "\n",
       "                          pressure         ghi        Kt  \n",
       "year   month day hour                                     \n",
       "2010.0 1.0   1.0 7.0    999.397619   61.060726 -0.069829  \n",
       "                 8.0    999.803333  166.338212 -0.025197  \n",
       "                 9.0    999.783333  306.455432 -0.012680  \n",
       "                 10.0   999.968333  411.028556 -0.009267  \n",
       "                 11.0  1000.193333  465.002071 -0.007967  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_new_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Groupdata - test dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_zen = df_test.groupby(['month','day','hour'])['zen'].mean()\n",
    "test_dw_solar = df_test.groupby(['month','day','hour'])['dw_solar'].mean()\n",
    "test_uw_solar = df_test.groupby(['month','day','hour'])['uw_solar'].mean()\n",
    "test_direct_n = df_test.groupby(['month','day','hour'])['direct_n'].mean()\n",
    "test_diffuse = df_test.groupby(['month','day','hour'])['diffuse'].mean()\n",
    "test_dw_ir = df_test.groupby(['month','day','hour'])['dw_ir'].mean()\n",
    "test_dw_casetemp = df_test.groupby(['month','day','hour'])['dw_casetemp'].mean()\n",
    "test_dw_dometemp = df_test.groupby(['month','day','hour'])['dw_dometemp'].mean()\n",
    "test_uw_ir = df_test.groupby(['month','day','hour'])['uw_ir'].mean()\n",
    "test_uw_casetemp = df_test.groupby(['month','day','hour'])['uw_casetemp'].mean()\n",
    "test_uw_dometemp = df_test.groupby(['month','day','hour'])['uw_dometemp'].mean()\n",
    "test_uvb = df_test.groupby(['month','day','hour'])['uvb'].mean()\n",
    "test_par = df_test.groupby(['month','day','hour'])['par'].mean()\n",
    "test_netsolar = df_test.groupby(['month','day','hour'])['netsolar'].mean()\n",
    "test_netir = df_test.groupby(['month','day','hour'])['netir'].mean()\n",
    "test_totalnet = df_test.groupby(['month','day','hour'])['totalnet'].mean()\n",
    "test_temp = df_test.groupby(['month','day','hour'])['temp'].mean()\n",
    "test_rh = df_test.groupby(['month','day','hour'])['rh'].mean()\n",
    "test_windspd = df_test.groupby(['month','day','hour'])['windspd'].mean()\n",
    "test_winddir = df_test.groupby(['month','day','hour'])['winddir'].mean()\n",
    "test_pressure = df_test.groupby(['month','day','hour'])['pressure'].mean()\n",
    "test_ghi = df_test.groupby(['month','day','hour'])['ghi'].mean()\n",
    "test_Kt = df_test.groupby(['month','day','hour'])['Kt'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_new_test = pd.concat([test_zen,test_dw_solar,test_uw_solar,test_direct_n,test_diffuse,test_dw_ir,\n",
    "                         test_dw_casetemp,test_dw_dometemp,test_uw_ir,test_uw_casetemp,test_uw_dometemp,\n",
    "                    test_uvb,test_par,test_netsolar,test_netir,test_totalnet,test_temp,test_rh,\n",
    "                         test_windspd,test_winddir,test_pressure,test_ghi,test_Kt], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>zen</th>\n",
       "      <th>dw_solar</th>\n",
       "      <th>uw_solar</th>\n",
       "      <th>direct_n</th>\n",
       "      <th>diffuse</th>\n",
       "      <th>dw_ir</th>\n",
       "      <th>dw_casetemp</th>\n",
       "      <th>dw_dometemp</th>\n",
       "      <th>uw_ir</th>\n",
       "      <th>uw_casetemp</th>\n",
       "      <th>...</th>\n",
       "      <th>netsolar</th>\n",
       "      <th>netir</th>\n",
       "      <th>totalnet</th>\n",
       "      <th>temp</th>\n",
       "      <th>rh</th>\n",
       "      <th>windspd</th>\n",
       "      <th>winddir</th>\n",
       "      <th>pressure</th>\n",
       "      <th>ghi</th>\n",
       "      <th>Kt</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hour</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>150.182143</td>\n",
       "      <td>-4.521429</td>\n",
       "      <td>-0.471429</td>\n",
       "      <td>1.007143</td>\n",
       "      <td>0.421429</td>\n",
       "      <td>216.700000</td>\n",
       "      <td>272.085714</td>\n",
       "      <td>271.642857</td>\n",
       "      <td>286.392857</td>\n",
       "      <td>270.300000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-69.692857</td>\n",
       "      <td>-69.692857</td>\n",
       "      <td>-2.271429</td>\n",
       "      <td>89.750000</td>\n",
       "      <td>2.871429</td>\n",
       "      <td>174.300000</td>\n",
       "      <td>998.471429</td>\n",
       "      <td>37.874247</td>\n",
       "      <td>-0.120932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0</th>\n",
       "      <td>146.446833</td>\n",
       "      <td>-4.301667</td>\n",
       "      <td>-0.345000</td>\n",
       "      <td>0.326667</td>\n",
       "      <td>0.518333</td>\n",
       "      <td>216.170000</td>\n",
       "      <td>271.606667</td>\n",
       "      <td>271.146667</td>\n",
       "      <td>284.053333</td>\n",
       "      <td>269.938333</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-67.881667</td>\n",
       "      <td>-67.881667</td>\n",
       "      <td>-2.773333</td>\n",
       "      <td>92.965000</td>\n",
       "      <td>3.066667</td>\n",
       "      <td>146.796667</td>\n",
       "      <td>997.966667</td>\n",
       "      <td>108.318844</td>\n",
       "      <td>-0.050473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0</th>\n",
       "      <td>137.938333</td>\n",
       "      <td>-4.831667</td>\n",
       "      <td>-0.351667</td>\n",
       "      <td>1.045000</td>\n",
       "      <td>0.490000</td>\n",
       "      <td>216.516667</td>\n",
       "      <td>271.153333</td>\n",
       "      <td>270.733333</td>\n",
       "      <td>288.993333</td>\n",
       "      <td>269.398333</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-72.466667</td>\n",
       "      <td>-72.466667</td>\n",
       "      <td>-3.256667</td>\n",
       "      <td>94.690000</td>\n",
       "      <td>3.873333</td>\n",
       "      <td>152.585000</td>\n",
       "      <td>996.900000</td>\n",
       "      <td>284.228645</td>\n",
       "      <td>-0.017408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0</th>\n",
       "      <td>127.563000</td>\n",
       "      <td>-5.281667</td>\n",
       "      <td>-0.308333</td>\n",
       "      <td>0.951667</td>\n",
       "      <td>0.436667</td>\n",
       "      <td>223.628333</td>\n",
       "      <td>271.841667</td>\n",
       "      <td>271.390000</td>\n",
       "      <td>294.260000</td>\n",
       "      <td>269.860000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-70.631667</td>\n",
       "      <td>-70.631667</td>\n",
       "      <td>-2.865000</td>\n",
       "      <td>91.560000</td>\n",
       "      <td>4.900000</td>\n",
       "      <td>164.426667</td>\n",
       "      <td>995.828333</td>\n",
       "      <td>448.178675</td>\n",
       "      <td>-0.011885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <td>116.367333</td>\n",
       "      <td>-5.035000</td>\n",
       "      <td>-0.243333</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>0.373333</td>\n",
       "      <td>236.190000</td>\n",
       "      <td>272.423333</td>\n",
       "      <td>271.983333</td>\n",
       "      <td>296.791667</td>\n",
       "      <td>270.326667</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-60.611667</td>\n",
       "      <td>-60.611667</td>\n",
       "      <td>-2.485000</td>\n",
       "      <td>89.688333</td>\n",
       "      <td>4.981667</td>\n",
       "      <td>173.578333</td>\n",
       "      <td>995.345000</td>\n",
       "      <td>570.306262</td>\n",
       "      <td>-0.008854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11.0</th>\n",
       "      <td>104.909333</td>\n",
       "      <td>-5.066667</td>\n",
       "      <td>-0.135000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.401667</td>\n",
       "      <td>233.243333</td>\n",
       "      <td>272.581667</td>\n",
       "      <td>272.093333</td>\n",
       "      <td>297.226667</td>\n",
       "      <td>270.480000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-63.985000</td>\n",
       "      <td>-63.985000</td>\n",
       "      <td>-2.386667</td>\n",
       "      <td>92.431667</td>\n",
       "      <td>5.595000</td>\n",
       "      <td>167.585000</td>\n",
       "      <td>994.613333</td>\n",
       "      <td>638.422279</td>\n",
       "      <td>-0.007937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12.0</th>\n",
       "      <td>93.430833</td>\n",
       "      <td>-2.668333</td>\n",
       "      <td>0.580000</td>\n",
       "      <td>1.250000</td>\n",
       "      <td>3.091667</td>\n",
       "      <td>238.826667</td>\n",
       "      <td>272.686667</td>\n",
       "      <td>272.196667</td>\n",
       "      <td>298.433333</td>\n",
       "      <td>270.543333</td>\n",
       "      <td>...</td>\n",
       "      <td>2.351667</td>\n",
       "      <td>-59.603333</td>\n",
       "      <td>-57.251667</td>\n",
       "      <td>-2.320000</td>\n",
       "      <td>92.493333</td>\n",
       "      <td>6.421667</td>\n",
       "      <td>167.895000</td>\n",
       "      <td>993.798333</td>\n",
       "      <td>646.734117</td>\n",
       "      <td>-0.004067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13.0</th>\n",
       "      <td>82.473333</td>\n",
       "      <td>50.488333</td>\n",
       "      <td>14.828333</td>\n",
       "      <td>7.531667</td>\n",
       "      <td>57.045000</td>\n",
       "      <td>245.631667</td>\n",
       "      <td>273.396667</td>\n",
       "      <td>272.983333</td>\n",
       "      <td>302.830000</td>\n",
       "      <td>271.061667</td>\n",
       "      <td>...</td>\n",
       "      <td>43.403333</td>\n",
       "      <td>-57.188333</td>\n",
       "      <td>-13.786667</td>\n",
       "      <td>-1.801667</td>\n",
       "      <td>91.796667</td>\n",
       "      <td>6.791667</td>\n",
       "      <td>167.960000</td>\n",
       "      <td>993.278333</td>\n",
       "      <td>594.568596</td>\n",
       "      <td>0.086837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14.0</th>\n",
       "      <td>72.405000</td>\n",
       "      <td>136.151667</td>\n",
       "      <td>36.918333</td>\n",
       "      <td>13.550000</td>\n",
       "      <td>140.373333</td>\n",
       "      <td>286.001667</td>\n",
       "      <td>275.108333</td>\n",
       "      <td>274.810000</td>\n",
       "      <td>312.523333</td>\n",
       "      <td>272.315000</td>\n",
       "      <td>...</td>\n",
       "      <td>107.461667</td>\n",
       "      <td>-26.526667</td>\n",
       "      <td>80.938333</td>\n",
       "      <td>-0.561667</td>\n",
       "      <td>101.838333</td>\n",
       "      <td>7.736667</td>\n",
       "      <td>175.430000</td>\n",
       "      <td>992.623333</td>\n",
       "      <td>486.293072</td>\n",
       "      <td>0.287014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <td>63.572667</td>\n",
       "      <td>249.433333</td>\n",
       "      <td>63.790000</td>\n",
       "      <td>22.685000</td>\n",
       "      <td>249.351667</td>\n",
       "      <td>287.831667</td>\n",
       "      <td>277.000000</td>\n",
       "      <td>276.723333</td>\n",
       "      <td>322.285000</td>\n",
       "      <td>273.716667</td>\n",
       "      <td>...</td>\n",
       "      <td>195.410000</td>\n",
       "      <td>-34.446667</td>\n",
       "      <td>160.953333</td>\n",
       "      <td>0.821667</td>\n",
       "      <td>109.766667</td>\n",
       "      <td>8.668333</td>\n",
       "      <td>182.706667</td>\n",
       "      <td>991.760000</td>\n",
       "      <td>332.051692</td>\n",
       "      <td>0.774799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16.0</th>\n",
       "      <td>56.723667</td>\n",
       "      <td>288.480000</td>\n",
       "      <td>73.990000</td>\n",
       "      <td>45.445000</td>\n",
       "      <td>272.428333</td>\n",
       "      <td>299.576667</td>\n",
       "      <td>278.595000</td>\n",
       "      <td>278.275000</td>\n",
       "      <td>331.316667</td>\n",
       "      <td>274.995000</td>\n",
       "      <td>...</td>\n",
       "      <td>222.996667</td>\n",
       "      <td>-31.736667</td>\n",
       "      <td>191.260000</td>\n",
       "      <td>1.906667</td>\n",
       "      <td>107.371667</td>\n",
       "      <td>9.136667</td>\n",
       "      <td>179.690000</td>\n",
       "      <td>990.573333</td>\n",
       "      <td>153.597510</td>\n",
       "      <td>2.067046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17.0</th>\n",
       "      <td>53.264194</td>\n",
       "      <td>314.187097</td>\n",
       "      <td>79.038710</td>\n",
       "      <td>4.867742</td>\n",
       "      <td>319.570968</td>\n",
       "      <td>309.864516</td>\n",
       "      <td>279.222581</td>\n",
       "      <td>278.980645</td>\n",
       "      <td>336.116129</td>\n",
       "      <td>275.416129</td>\n",
       "      <td>...</td>\n",
       "      <td>243.445161</td>\n",
       "      <td>-26.254839</td>\n",
       "      <td>217.200000</td>\n",
       "      <td>2.474194</td>\n",
       "      <td>111.645161</td>\n",
       "      <td>9.616129</td>\n",
       "      <td>176.358065</td>\n",
       "      <td>989.632258</td>\n",
       "      <td>45.373147</td>\n",
       "      <td>7.893298</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             zen    dw_solar   uw_solar   direct_n     diffuse       dw_ir  \\\n",
       "hour                                                                         \n",
       "6.0   150.182143   -4.521429  -0.471429   1.007143    0.421429  216.700000   \n",
       "7.0   146.446833   -4.301667  -0.345000   0.326667    0.518333  216.170000   \n",
       "8.0   137.938333   -4.831667  -0.351667   1.045000    0.490000  216.516667   \n",
       "9.0   127.563000   -5.281667  -0.308333   0.951667    0.436667  223.628333   \n",
       "10.0  116.367333   -5.035000  -0.243333   0.833333    0.373333  236.190000   \n",
       "11.0  104.909333   -5.066667  -0.135000   0.480000    0.401667  233.243333   \n",
       "12.0   93.430833   -2.668333   0.580000   1.250000    3.091667  238.826667   \n",
       "13.0   82.473333   50.488333  14.828333   7.531667   57.045000  245.631667   \n",
       "14.0   72.405000  136.151667  36.918333  13.550000  140.373333  286.001667   \n",
       "15.0   63.572667  249.433333  63.790000  22.685000  249.351667  287.831667   \n",
       "16.0   56.723667  288.480000  73.990000  45.445000  272.428333  299.576667   \n",
       "17.0   53.264194  314.187097  79.038710   4.867742  319.570968  309.864516   \n",
       "\n",
       "      dw_casetemp  dw_dometemp       uw_ir  uw_casetemp    ...       netsolar  \\\n",
       "hour                                                       ...                  \n",
       "6.0    272.085714   271.642857  286.392857   270.300000    ...       0.000000   \n",
       "7.0    271.606667   271.146667  284.053333   269.938333    ...       0.000000   \n",
       "8.0    271.153333   270.733333  288.993333   269.398333    ...       0.000000   \n",
       "9.0    271.841667   271.390000  294.260000   269.860000    ...       0.000000   \n",
       "10.0   272.423333   271.983333  296.791667   270.326667    ...       0.000000   \n",
       "11.0   272.581667   272.093333  297.226667   270.480000    ...       0.000000   \n",
       "12.0   272.686667   272.196667  298.433333   270.543333    ...       2.351667   \n",
       "13.0   273.396667   272.983333  302.830000   271.061667    ...      43.403333   \n",
       "14.0   275.108333   274.810000  312.523333   272.315000    ...     107.461667   \n",
       "15.0   277.000000   276.723333  322.285000   273.716667    ...     195.410000   \n",
       "16.0   278.595000   278.275000  331.316667   274.995000    ...     222.996667   \n",
       "17.0   279.222581   278.980645  336.116129   275.416129    ...     243.445161   \n",
       "\n",
       "          netir    totalnet      temp          rh   windspd     winddir  \\\n",
       "hour                                                                      \n",
       "6.0  -69.692857  -69.692857 -2.271429   89.750000  2.871429  174.300000   \n",
       "7.0  -67.881667  -67.881667 -2.773333   92.965000  3.066667  146.796667   \n",
       "8.0  -72.466667  -72.466667 -3.256667   94.690000  3.873333  152.585000   \n",
       "9.0  -70.631667  -70.631667 -2.865000   91.560000  4.900000  164.426667   \n",
       "10.0 -60.611667  -60.611667 -2.485000   89.688333  4.981667  173.578333   \n",
       "11.0 -63.985000  -63.985000 -2.386667   92.431667  5.595000  167.585000   \n",
       "12.0 -59.603333  -57.251667 -2.320000   92.493333  6.421667  167.895000   \n",
       "13.0 -57.188333  -13.786667 -1.801667   91.796667  6.791667  167.960000   \n",
       "14.0 -26.526667   80.938333 -0.561667  101.838333  7.736667  175.430000   \n",
       "15.0 -34.446667  160.953333  0.821667  109.766667  8.668333  182.706667   \n",
       "16.0 -31.736667  191.260000  1.906667  107.371667  9.136667  179.690000   \n",
       "17.0 -26.254839  217.200000  2.474194  111.645161  9.616129  176.358065   \n",
       "\n",
       "        pressure         ghi        Kt  \n",
       "hour                                    \n",
       "6.0   998.471429   37.874247 -0.120932  \n",
       "7.0   997.966667  108.318844 -0.050473  \n",
       "8.0   996.900000  284.228645 -0.017408  \n",
       "9.0   995.828333  448.178675 -0.011885  \n",
       "10.0  995.345000  570.306262 -0.008854  \n",
       "11.0  994.613333  638.422279 -0.007937  \n",
       "12.0  993.798333  646.734117 -0.004067  \n",
       "13.0  993.278333  594.568596  0.086837  \n",
       "14.0  992.623333  486.293072  0.287014  \n",
       "15.0  991.760000  332.051692  0.774799  \n",
       "16.0  990.573333  153.597510  2.067046  \n",
       "17.0  989.632258   45.373147  7.893298  \n",
       "\n",
       "[12 rows x 23 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_new_test.loc[2].xs(17,level='day')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shifting Kt values to make 1 hour ahead forecast"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "levels_index= []\n",
    "for m in df_new_train.index.levels:\n",
    "    levels_index.append(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for i in levels_index[0]:\n",
    "    for j in levels_index[1]:\n",
    "        df_new_train.loc[i].loc[j]['Kt'] = df_new_train.loc[i].loc[j]['Kt'].shift(-2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_new_train = df_new_train[~(df_new_train['Kt'].isnull())]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "levels_index2= []\n",
    "for m in df_new_test.index.levels:\n",
    "    levels_index2.append(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in levels_index2[0]:\n",
    "    for j in levels_index2[1]:\n",
    "        df_new_test.loc[i].loc[j]['Kt'] = df_new_test.loc[i].loc[j]['Kt'].shift(-2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_new_test = df_new_test[~(df_new_test['Kt'].isnull())]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>zen</th>\n",
       "      <th>dw_solar</th>\n",
       "      <th>uw_solar</th>\n",
       "      <th>direct_n</th>\n",
       "      <th>diffuse</th>\n",
       "      <th>dw_ir</th>\n",
       "      <th>dw_casetemp</th>\n",
       "      <th>dw_dometemp</th>\n",
       "      <th>uw_ir</th>\n",
       "      <th>uw_casetemp</th>\n",
       "      <th>...</th>\n",
       "      <th>netsolar</th>\n",
       "      <th>netir</th>\n",
       "      <th>totalnet</th>\n",
       "      <th>temp</th>\n",
       "      <th>rh</th>\n",
       "      <th>windspd</th>\n",
       "      <th>winddir</th>\n",
       "      <th>pressure</th>\n",
       "      <th>ghi</th>\n",
       "      <th>Kt</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>hour</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>18.0</th>\n",
       "      <td>24.3985</td>\n",
       "      <td>903.236667</td>\n",
       "      <td>170.488333</td>\n",
       "      <td>609.081667</td>\n",
       "      <td>346.083333</td>\n",
       "      <td>382.34</td>\n",
       "      <td>304.0</td>\n",
       "      <td>303.905</td>\n",
       "      <td>482.805</td>\n",
       "      <td>300.676667</td>\n",
       "      <td>...</td>\n",
       "      <td>729.176667</td>\n",
       "      <td>-100.458333</td>\n",
       "      <td>628.715</td>\n",
       "      <td>25.17</td>\n",
       "      <td>68.681667</td>\n",
       "      <td>2.18</td>\n",
       "      <td>141.071667</td>\n",
       "      <td>992.59</td>\n",
       "      <td>201.959502</td>\n",
       "      <td>1146.230573</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    zen    dw_solar    uw_solar    direct_n     diffuse  \\\n",
       "month day hour                                                            \n",
       "8.0   5.0 18.0  24.3985  903.236667  170.488333  609.081667  346.083333   \n",
       "\n",
       "                 dw_ir  dw_casetemp  dw_dometemp    uw_ir  uw_casetemp  \\\n",
       "month day hour                                                           \n",
       "8.0   5.0 18.0  382.34        304.0      303.905  482.805   300.676667   \n",
       "\n",
       "                   ...         netsolar       netir  totalnet   temp  \\\n",
       "month day hour     ...                                                 \n",
       "8.0   5.0 18.0     ...       729.176667 -100.458333   628.715  25.17   \n",
       "\n",
       "                       rh  windspd     winddir  pressure         ghi  \\\n",
       "month day hour                                                         \n",
       "8.0   5.0 18.0  68.681667     2.18  141.071667    992.59  201.959502   \n",
       "\n",
       "                         Kt  \n",
       "month day hour               \n",
       "8.0   5.0 18.0  1146.230573  \n",
       "\n",
       "[1 rows x 23 columns]"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_new_test[df_new_test['Kt']==df_new_test['Kt'].max()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Normalize train and test dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_norm = (df_new_train - df_new_train.mean()) / (df_new_train.max() - df_new_train.min())\n",
    "test_norm =  (df_new_test - df_new_test.mean()) / (df_new_test.max() - df_new_test.min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_norm.reset_index(inplace=True,drop=True)\n",
    "test_norm.reset_index(inplace=True,drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Making train and test sets with train_norm and test_norm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "#### finding the gcf (greatest common factor) of train and test dataset's length and chop off the extra rows to make it divisible with the batchsize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\s269596\\AppData\\Local\\Continuum\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: DeprecationWarning: fractions.gcd() is deprecated. Use math.gcd() instead.\n",
      "  \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from fractions import gcd\n",
    "gcd(train_norm.shape[0],test_norm.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "def roundup(x):\n",
    "    return int(math.ceil(x / 100.0)) * 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_lim = roundup(train_norm.shape[0])\n",
    "test_lim = roundup(test_norm.shape[0])\n",
    "\n",
    "train_random = train_norm.sample(train_lim-train_norm.shape[0])\n",
    "test_random = test_norm.sample(test_lim-test_norm.shape[0])\n",
    "\n",
    "train_norm = train_norm.append(train_random)\n",
    "test_norm = test_norm.append(test_random)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1 = train_norm.drop('Kt',axis=1)\n",
    "y1 = train_norm['Kt']\n",
    "\n",
    "X2 = test_norm.drop('Kt',axis=1)\n",
    "y2 = test_norm['Kt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X1_train shape is (9600, 22)\n",
      "y1_train shape is (9600,)\n",
      "X2_test shape is (4100, 22)\n",
      "y2_test shape is (4100,)\n"
     ]
    }
   ],
   "source": [
    "print(\"X1_train shape is {}\".format(X1.shape))\n",
    "print(\"y1_train shape is {}\".format(y1.shape))\n",
    "print(\"X2_test shape is {}\".format(X2.shape))\n",
    "print(\"y2_test shape is {}\".format(y2.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.array(X1)\n",
    "y_train  = np.array(y1)\n",
    "X_test = np.array(X2)\n",
    "y_test = np.array(y2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### start of RNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torchvision.transforms as transforms\n",
    "import torchvision.datasets as dsets\n",
    "from torch.autograd import Variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "class RNNModel(nn.Module):\n",
    "    def __init__(self, input_dim, hidden_dim, layer_dim, output_dim):\n",
    "        super(RNNModel, self).__init__()\n",
    "        #Hidden Dimension\n",
    "        self.hidden_dim = hidden_dim\n",
    "        \n",
    "        # Number of hidden layers\n",
    "        self.layer_dim = layer_dim\n",
    "        \n",
    "        #Building the RNN\n",
    "        self.rnn = nn.RNN(input_dim, hidden_dim, layer_dim, batch_first=True, nonlinearity='relu')\n",
    "        \n",
    "        # Readout layer\n",
    "        self.fc = nn.Linear(hidden_dim, output_dim)\n",
    "        \n",
    "    def forward(self, x):\n",
    "        # Initializing the hidden state with zeros\n",
    "        # (layer_dim, batch_size, hidden_dim)\n",
    "        h0 = Variable(torch.zeros(self.layer_dim, x.size(0), self.hidden_dim))\n",
    "        \n",
    "        #One time step (the last one perhaps?)\n",
    "        out, hn = self.rnn(x, h0)\n",
    "        \n",
    "        # Indexing hidden state of the last time step\n",
    "        # out.size() --> ??\n",
    "        #out[:,-1,:] --> is it going to be 100,100\n",
    "        out = self.fc(out[:,-1,:])\n",
    "        # out.size() --> 100,1\n",
    "        return out\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Instantiating Model Class\n",
    "input_dim = 22\n",
    "hidden_dim = 15\n",
    "layer_dim = 1\n",
    "output_dim = 1\n",
    "batch_size = 100\n",
    "\n",
    "model = RNNModel(input_dim, hidden_dim, layer_dim, output_dim)\n",
    "\n",
    "# Instantiating Loss Class\n",
    "criterion = nn.MSELoss()\n",
    "\n",
    "# Instantiate Optimizer Class\n",
    "learning_rate = 0.001\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)\n",
    "\n",
    "# converting numpy array to torch tensor\n",
    "X_train = torch.from_numpy(X_train)\n",
    "y_train = torch.from_numpy(y_train)\n",
    "X_test = torch.from_numpy(X_test)\n",
    "y_test = torch.from_numpy(y_test)\n",
    "\n",
    "# initializing lists to store losses over epochs:\n",
    "train_loss = []\n",
    "test_loss = []\n",
    "train_iter = []\n",
    "test_iter = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0 Iteration: 0. Train_MSE: 0.003014477202668786. Test_MSE: 0.00043477852058157983\n",
      "Epoch: 2 Iteration: 200. Train_MSE: 0.004386840853840113. Test_MSE: 0.00036709572612368793\n",
      "Epoch: 4 Iteration: 400. Train_MSE: 0.003066215431317687. Test_MSE: 0.0003605415669057718\n",
      "Epoch: 6 Iteration: 600. Train_MSE: 0.0017822093795984983. Test_MSE: 0.0003600953492064432\n",
      "Epoch: 8 Iteration: 800. Train_MSE: 0.00207426305860281. Test_MSE: 0.0003561617319794113\n",
      "Epoch: 10 Iteration: 1000. Train_MSE: 0.002230407204478979. Test_MSE: 0.0003476258395019613\n",
      "Epoch: 12 Iteration: 1200. Train_MSE: 0.002361592836678028. Test_MSE: 0.00034094246726485176\n",
      "Epoch: 14 Iteration: 1400. Train_MSE: 0.002364522311836481. Test_MSE: 0.0003348758525612563\n",
      "Epoch: 16 Iteration: 1600. Train_MSE: 0.0012210470158606768. Test_MSE: 0.00032842937346032385\n",
      "Epoch: 18 Iteration: 1800. Train_MSE: 0.0017839764477685094. Test_MSE: 0.0003223764149041517\n",
      "Epoch: 21 Iteration: 2000. Train_MSE: 0.0020191252697259188. Test_MSE: 0.00031637173431459245\n",
      "Epoch: 23 Iteration: 2200. Train_MSE: 0.0013153713662177324. Test_MSE: 0.0003115950608143553\n",
      "Epoch: 25 Iteration: 2400. Train_MSE: 0.0013169359881430864. Test_MSE: 0.0003073509994829285\n",
      "Epoch: 27 Iteration: 2600. Train_MSE: 0.0039107403717935085. Test_MSE: 0.0003033633253598889\n",
      "Epoch: 29 Iteration: 2800. Train_MSE: 0.0015834323130548. Test_MSE: 0.00029964910398802097\n",
      "Epoch: 31 Iteration: 3000. Train_MSE: 0.0010893784929066896. Test_MSE: 0.0002963299113396649\n",
      "Epoch: 33 Iteration: 3200. Train_MSE: 0.0020225902553647757. Test_MSE: 0.0002929915147784804\n",
      "Epoch: 35 Iteration: 3400. Train_MSE: 0.003434755140915513. Test_MSE: 0.00029003515261632774\n",
      "Epoch: 37 Iteration: 3600. Train_MSE: 0.0008371029980480671. Test_MSE: 0.00028752267902984246\n",
      "Epoch: 40 Iteration: 3800. Train_MSE: 0.0021206773817539215. Test_MSE: 0.0002850626493751435\n",
      "Epoch: 42 Iteration: 4000. Train_MSE: 0.0012005496537312865. Test_MSE: 0.0002827612558821958\n",
      "Epoch: 44 Iteration: 4200. Train_MSE: 0.0015359168173745275. Test_MSE: 0.0002803073083195569\n",
      "Epoch: 46 Iteration: 4400. Train_MSE: 0.0005546695319935679. Test_MSE: 0.0002784519318302464\n",
      "Epoch: 48 Iteration: 4600. Train_MSE: 0.0009585462976247072. Test_MSE: 0.00027651522167492155\n",
      "Epoch: 50 Iteration: 4800. Train_MSE: 0.0009964904747903347. Test_MSE: 0.00027499820274898897\n",
      "Epoch: 52 Iteration: 5000. Train_MSE: 0.0013425083598122. Test_MSE: 0.00027259180682507564\n",
      "Epoch: 54 Iteration: 5200. Train_MSE: 0.000995708629488945. Test_MSE: 0.0002704601081524755\n",
      "Epoch: 56 Iteration: 5400. Train_MSE: 0.0007389651727862656. Test_MSE: 0.00026942210336775696\n",
      "Epoch: 58 Iteration: 5600. Train_MSE: 0.001171767944470048. Test_MSE: 0.00026777696251693785\n",
      "Epoch: 61 Iteration: 5800. Train_MSE: 0.0010103308595716953. Test_MSE: 0.0002663597967807979\n",
      "Epoch: 63 Iteration: 6000. Train_MSE: 0.0007723983144387603. Test_MSE: 0.0002646327144933488\n",
      "Epoch: 65 Iteration: 6200. Train_MSE: 0.0007861359626986086. Test_MSE: 0.0002630047725878294\n",
      "Epoch: 67 Iteration: 6400. Train_MSE: 0.003771392395719886. Test_MSE: 0.0002620114618523\n",
      "Epoch: 69 Iteration: 6600. Train_MSE: 0.0008284391951747239. Test_MSE: 0.0002604444341210526\n",
      "Epoch: 71 Iteration: 6800. Train_MSE: 0.0006420391146093607. Test_MSE: 0.00025912422187159223\n",
      "Epoch: 73 Iteration: 7000. Train_MSE: 0.0014610935468226671. Test_MSE: 0.000257081257460287\n",
      "Epoch: 75 Iteration: 7200. Train_MSE: 0.0035580613184720278. Test_MSE: 0.00025591303723173655\n",
      "Epoch: 77 Iteration: 7400. Train_MSE: 0.0005273711867630482. Test_MSE: 0.0002552852133101534\n",
      "Epoch: 80 Iteration: 7600. Train_MSE: 0.00126516236923635. Test_MSE: 0.00025358848639997326\n",
      "Epoch: 82 Iteration: 7800. Train_MSE: 0.000737281865440309. Test_MSE: 0.0002521401872665316\n",
      "Epoch: 84 Iteration: 8000. Train_MSE: 0.0012609566329047084. Test_MSE: 0.0002505799788570361\n",
      "Epoch: 86 Iteration: 8200. Train_MSE: 0.0004932305891998112. Test_MSE: 0.0002499016398940088\n",
      "Epoch: 88 Iteration: 8400. Train_MSE: 0.0006831980426795781. Test_MSE: 0.00024870162305584275\n",
      "Epoch: 90 Iteration: 8600. Train_MSE: 0.0007130982703529298. Test_MSE: 0.0002475117726561942\n",
      "Epoch: 92 Iteration: 8800. Train_MSE: 0.0009678774513304234. Test_MSE: 0.0002453387929759211\n",
      "Epoch: 94 Iteration: 9000. Train_MSE: 0.000686570827383548. Test_MSE: 0.00024383397762362186\n",
      "Epoch: 96 Iteration: 9200. Train_MSE: 0.0006678100908175111. Test_MSE: 0.0002436534794806128\n",
      "Epoch: 98 Iteration: 9400. Train_MSE: 0.0009251667652279139. Test_MSE: 0.00024230260182376854\n",
      "Epoch: 101 Iteration: 9600. Train_MSE: 0.0007142897229641676. Test_MSE: 0.00024100061352536653\n",
      "Epoch: 103 Iteration: 9800. Train_MSE: 0.000611158786341548. Test_MSE: 0.00023963021440665433\n",
      "Epoch: 105 Iteration: 10000. Train_MSE: 0.000656632415484637. Test_MSE: 0.00023846586394306833\n",
      "Epoch: 107 Iteration: 10200. Train_MSE: 0.0037245897110551596. Test_MSE: 0.00023794328998681316\n",
      "Epoch: 109 Iteration: 10400. Train_MSE: 0.0006046669441275299. Test_MSE: 0.0002364650035104805\n",
      "Epoch: 111 Iteration: 10600. Train_MSE: 0.0004589994205161929. Test_MSE: 0.00023520931701108484\n",
      "Epoch: 113 Iteration: 10800. Train_MSE: 0.0012208562111482024. Test_MSE: 0.00023352615800274504\n",
      "Epoch: 115 Iteration: 11000. Train_MSE: 0.0035637172404676676. Test_MSE: 0.0002327439515207991\n",
      "Epoch: 117 Iteration: 11200. Train_MSE: 0.0004371996328700334. Test_MSE: 0.00023246070501726354\n",
      "Epoch: 120 Iteration: 11400. Train_MSE: 0.0009845560416579247. Test_MSE: 0.0002307954240906552\n",
      "Epoch: 122 Iteration: 11600. Train_MSE: 0.0005664598429575562. Test_MSE: 0.00022957599862543783\n",
      "Epoch: 124 Iteration: 11800. Train_MSE: 0.0011116296518594027. Test_MSE: 0.00022833527102610214\n",
      "Epoch: 126 Iteration: 12000. Train_MSE: 0.00045534915989264846. Test_MSE: 0.00022795062572915521\n",
      "Epoch: 128 Iteration: 12200. Train_MSE: 0.0005512065254151821. Test_MSE: 0.00022695564010116837\n",
      "Epoch: 130 Iteration: 12400. Train_MSE: 0.000584968423936516. Test_MSE: 0.00022586426766086685\n",
      "Epoch: 132 Iteration: 12600. Train_MSE: 0.0007514673634432256. Test_MSE: 0.00022393899579080143\n",
      "Epoch: 134 Iteration: 12800. Train_MSE: 0.0005616085254587233. Test_MSE: 0.0002227812460920609\n",
      "Epoch: 136 Iteration: 13000. Train_MSE: 0.0006174279842525721. Test_MSE: 0.0002228278356550722\n",
      "Epoch: 138 Iteration: 13200. Train_MSE: 0.0007679768605157733. Test_MSE: 0.00022164484693037165\n",
      "Epoch: 141 Iteration: 13400. Train_MSE: 0.0005805891123600304. Test_MSE: 0.00022047073290972427\n",
      "Epoch: 143 Iteration: 13600. Train_MSE: 0.0005186137277632952. Test_MSE: 0.00021933784334669817\n",
      "Epoch: 145 Iteration: 13800. Train_MSE: 0.0005893992492929101. Test_MSE: 0.00021840998447132616\n",
      "Epoch: 147 Iteration: 14000. Train_MSE: 0.0036833800841122866. Test_MSE: 0.00021806250712111492\n",
      "Epoch: 149 Iteration: 14200. Train_MSE: 0.0005035631475038826. Test_MSE: 0.0002167340514396534\n",
      "Epoch: 151 Iteration: 14400. Train_MSE: 0.000357722194166854. Test_MSE: 0.00021564487635199753\n",
      "Epoch: 153 Iteration: 14600. Train_MSE: 0.0010741211008280516. Test_MSE: 0.0002143077223861394\n",
      "Epoch: 155 Iteration: 14800. Train_MSE: 0.00351944868452847. Test_MSE: 0.00021372113785716928\n",
      "Epoch: 157 Iteration: 15000. Train_MSE: 0.0003995517618022859. Test_MSE: 0.00021357224206559506\n",
      "Epoch: 160 Iteration: 15200. Train_MSE: 0.0008379195351153612. Test_MSE: 0.00021206265307053624\n",
      "Epoch: 162 Iteration: 15400. Train_MSE: 0.0004724494065158069. Test_MSE: 0.0002110969103823336\n",
      "Epoch: 164 Iteration: 15600. Train_MSE: 0.0010027706157416105. Test_MSE: 0.00021008862213020994\n",
      "Epoch: 166 Iteration: 15800. Train_MSE: 0.0004165588761679828. Test_MSE: 0.000209818384624129\n",
      "Epoch: 168 Iteration: 16000. Train_MSE: 0.00046378251863643527. Test_MSE: 0.00020897488785129502\n",
      "Epoch: 170 Iteration: 16200. Train_MSE: 0.0005068730097264051. Test_MSE: 0.00020806502949699077\n",
      "Epoch: 172 Iteration: 16400. Train_MSE: 0.0006072197575122118. Test_MSE: 0.00020644161295542112\n",
      "Epoch: 174 Iteration: 16600. Train_MSE: 0.0004919819766655564. Test_MSE: 0.00020555312240894404\n",
      "Epoch: 176 Iteration: 16800. Train_MSE: 0.0005716890445910394. Test_MSE: 0.00020567531624951808\n",
      "Epoch: 178 Iteration: 17000. Train_MSE: 0.0006541252369061112. Test_MSE: 0.00020467811264231355\n",
      "Epoch: 181 Iteration: 17200. Train_MSE: 0.0005024181446060538. Test_MSE: 0.00020369935712940635\n",
      "Epoch: 183 Iteration: 17400. Train_MSE: 0.000454365013865754. Test_MSE: 0.0002027803823417597\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 185 Iteration: 17600. Train_MSE: 0.0005426138523034751. Test_MSE: 0.00020202143248690315\n",
      "Epoch: 187 Iteration: 17800. Train_MSE: 0.003650691593065858. Test_MSE: 0.00020178212351939652\n",
      "Epoch: 189 Iteration: 18000. Train_MSE: 0.000443875091150403. Test_MSE: 0.00020064160818246375\n",
      "Epoch: 191 Iteration: 18200. Train_MSE: 0.0002933972573373467. Test_MSE: 0.00019974744413793987\n",
      "Epoch: 193 Iteration: 18400. Train_MSE: 0.0009730558376759291. Test_MSE: 0.00019870896984905284\n",
      "Epoch: 195 Iteration: 18600. Train_MSE: 0.0034597229678183794. Test_MSE: 0.0001982531404439916\n",
      "Epoch: 197 Iteration: 18800. Train_MSE: 0.00037943021743558347. Test_MSE: 0.00019818822373565462\n",
      "Epoch: 200 Iteration: 19000. Train_MSE: 0.0007358342409133911. Test_MSE: 0.0001968781806197731\n",
      "Epoch: 202 Iteration: 19200. Train_MSE: 0.0004081617807969451. Test_MSE: 0.0001961484743968021\n",
      "Epoch: 204 Iteration: 19400. Train_MSE: 0.0009236788027919829. Test_MSE: 0.00019532634517155267\n",
      "Epoch: 206 Iteration: 19600. Train_MSE: 0.0003835046663880348. Test_MSE: 0.00019512615435086393\n",
      "Epoch: 208 Iteration: 19800. Train_MSE: 0.0003960388421546668. Test_MSE: 0.00019442296272348899\n",
      "Epoch: 210 Iteration: 20000. Train_MSE: 0.0004522150265984237. Test_MSE: 0.00019369839082731944\n",
      "Epoch: 212 Iteration: 20200. Train_MSE: 0.0005070131737738848. Test_MSE: 0.00019234741744810957\n",
      "Epoch: 214 Iteration: 20400. Train_MSE: 0.000446480349637568. Test_MSE: 0.00019166747810994263\n",
      "Epoch: 216 Iteration: 20600. Train_MSE: 0.0005382453091442585. Test_MSE: 0.00019182732367398767\n",
      "Epoch: 218 Iteration: 20800. Train_MSE: 0.0005674015847034752. Test_MSE: 0.00019100526080783833\n",
      "Epoch: 221 Iteration: 21000. Train_MSE: 0.00044992018956691027. Test_MSE: 0.00019020846365874793\n",
      "Epoch: 223 Iteration: 21200. Train_MSE: 0.00040737370727583766. Test_MSE: 0.00018947102330391162\n",
      "Epoch: 225 Iteration: 21400. Train_MSE: 0.0005110602942295372. Test_MSE: 0.00018884563198578942\n",
      "Epoch: 227 Iteration: 21600. Train_MSE: 0.003626501653343439. Test_MSE: 0.0001886883698217261\n",
      "Epoch: 229 Iteration: 21800. Train_MSE: 0.0004031157004646957. Test_MSE: 0.00018773130796782462\n",
      "Epoch: 231 Iteration: 22000. Train_MSE: 0.00025045478832907975. Test_MSE: 0.00018701678977076162\n",
      "Epoch: 233 Iteration: 22200. Train_MSE: 0.0008999721612781286. Test_MSE: 0.00018621977650280425\n",
      "Epoch: 235 Iteration: 22400. Train_MSE: 0.0034056329168379307. Test_MSE: 0.00018586552968551058\n",
      "Epoch: 237 Iteration: 22600. Train_MSE: 0.00036245532101020217. Test_MSE: 0.00018586579066101326\n",
      "Epoch: 240 Iteration: 22800. Train_MSE: 0.0006569685647264123. Test_MSE: 0.00018474680171502071\n",
      "Epoch: 242 Iteration: 23000. Train_MSE: 0.00036102489684708416. Test_MSE: 0.00018420556195609\n",
      "Epoch: 244 Iteration: 23200. Train_MSE: 0.0008665548521094024. Test_MSE: 0.00018353275769883047\n",
      "Epoch: 246 Iteration: 23400. Train_MSE: 0.00035703848698176444. Test_MSE: 0.0001833787011973376\n",
      "Epoch: 248 Iteration: 23600. Train_MSE: 0.00034230490564368665. Test_MSE: 0.00018279267456024084\n",
      "Epoch: 250 Iteration: 23800. Train_MSE: 0.0004116835771128535. Test_MSE: 0.00018222231606571213\n",
      "Epoch: 252 Iteration: 24000. Train_MSE: 0.0004354169359430671. Test_MSE: 0.00018109328600270268\n",
      "Epoch: 254 Iteration: 24200. Train_MSE: 0.0004142009129282087. Test_MSE: 0.0001805668928798454\n",
      "Epoch: 256 Iteration: 24400. Train_MSE: 0.0005131162470206618. Test_MSE: 0.00018074053156060446\n",
      "Epoch: 258 Iteration: 24600. Train_MSE: 0.000499830930493772. Test_MSE: 0.00018006040370754616\n",
      "Epoch: 261 Iteration: 24800. Train_MSE: 0.0004110559821128845. Test_MSE: 0.00017940723748713098\n",
      "Epoch: 263 Iteration: 25000. Train_MSE: 0.0003717654326464981. Test_MSE: 0.00017880828000771542\n",
      "Epoch: 265 Iteration: 25200. Train_MSE: 0.0004893235745839775. Test_MSE: 0.00017827902577683715\n",
      "Epoch: 267 Iteration: 25400. Train_MSE: 0.0036081732250750065. Test_MSE: 0.00017817001539717454\n",
      "Epoch: 269 Iteration: 25600. Train_MSE: 0.0003721550165209919. Test_MSE: 0.00017735713504168376\n",
      "Epoch: 271 Iteration: 25800. Train_MSE: 0.000219332316191867. Test_MSE: 0.0001767750102515658\n",
      "Epoch: 273 Iteration: 26000. Train_MSE: 0.0008454358321614563. Test_MSE: 0.0001761533531071251\n",
      "Epoch: 275 Iteration: 26200. Train_MSE: 0.003358607180416584. Test_MSE: 0.0001758617933646687\n",
      "Epoch: 277 Iteration: 26400. Train_MSE: 0.00034872686956077814. Test_MSE: 0.00017589849097392676\n",
      "Epoch: 280 Iteration: 26600. Train_MSE: 0.0005929035833105445. Test_MSE: 0.00017492799582113173\n",
      "Epoch: 282 Iteration: 26800. Train_MSE: 0.00032316319993697107. Test_MSE: 0.00017451616355996185\n",
      "Epoch: 284 Iteration: 27000. Train_MSE: 0.0008238942245952785. Test_MSE: 0.00017394733705458949\n",
      "Epoch: 286 Iteration: 27200. Train_MSE: 0.00033596769208088517. Test_MSE: 0.00017381237032637655\n",
      "Epoch: 288 Iteration: 27400. Train_MSE: 0.00029811818967573345. Test_MSE: 0.00017331288769210579\n",
      "Epoch: 290 Iteration: 27600. Train_MSE: 0.00037919238093309104. Test_MSE: 0.00017285256001901692\n",
      "Epoch: 292 Iteration: 27800. Train_MSE: 0.00038279383443295956. Test_MSE: 0.00017188995648857136\n",
      "Epoch: 294 Iteration: 28000. Train_MSE: 0.00039006746374070644. Test_MSE: 0.00017147295819091373\n",
      "Epoch: 296 Iteration: 28200. Train_MSE: 0.0004940131912007928. Test_MSE: 0.0001716349281139191\n",
      "Epoch: 298 Iteration: 28400. Train_MSE: 0.0004452375287655741. Test_MSE: 0.0001710586474549802\n",
      "Epoch: 301 Iteration: 28600. Train_MSE: 0.00038039960782043636. Test_MSE: 0.0001705012947207443\n",
      "Epoch: 303 Iteration: 28800. Train_MSE: 0.000344254047377035. Test_MSE: 0.00016999531146503206\n",
      "Epoch: 305 Iteration: 29000. Train_MSE: 0.0004731729859486222. Test_MSE: 0.0001695260291959281\n",
      "Epoch: 307 Iteration: 29200. Train_MSE: 0.0035945477429777384. Test_MSE: 0.00016943849570306573\n",
      "Epoch: 309 Iteration: 29400. Train_MSE: 0.00034710229374468327. Test_MSE: 0.00016873791360555275\n",
      "Epoch: 311 Iteration: 29600. Train_MSE: 0.000195659973542206. Test_MSE: 0.00016825155713810641\n",
      "Epoch: 313 Iteration: 29800. Train_MSE: 0.0008031467441469431. Test_MSE: 0.00016775520678740313\n",
      "Epoch: 315 Iteration: 30000. Train_MSE: 0.0033178480807691813. Test_MSE: 0.00016750295984031008\n",
      "Epoch: 317 Iteration: 30200. Train_MSE: 0.000337200821377337. Test_MSE: 0.0001675538557626257\n",
      "Epoch: 320 Iteration: 30400. Train_MSE: 0.0005388515419326723. Test_MSE: 0.00016669976081966722\n",
      "Epoch: 322 Iteration: 30600. Train_MSE: 0.00029196517425589263. Test_MSE: 0.0001663776068839166\n",
      "Epoch: 324 Iteration: 30800. Train_MSE: 0.000790810736361891. Test_MSE: 0.00016588541777411692\n",
      "Epoch: 326 Iteration: 31000. Train_MSE: 0.00031826517079025507. Test_MSE: 0.00016575488805443392\n",
      "Epoch: 328 Iteration: 31200. Train_MSE: 0.00026177382096648216. Test_MSE: 0.00016532218656856875\n",
      "Epoch: 330 Iteration: 31400. Train_MSE: 0.00035232622758485377. Test_MSE: 0.00016494613607777404\n",
      "Epoch: 332 Iteration: 31600. Train_MSE: 0.0003426145703997463. Test_MSE: 0.00016411338494053853\n",
      "Epoch: 334 Iteration: 31800. Train_MSE: 0.0003710249438881874. Test_MSE: 0.00016377741629026158\n",
      "Epoch: 336 Iteration: 32000. Train_MSE: 0.0004777761932928115. Test_MSE: 0.0001639278953974697\n",
      "Epoch: 338 Iteration: 32200. Train_MSE: 0.0004003439098596573. Test_MSE: 0.00016343757585444318\n",
      "Epoch: 341 Iteration: 32400. Train_MSE: 0.0003556623705662787. Test_MSE: 0.00016295670182242552\n",
      "Epoch: 343 Iteration: 32600. Train_MSE: 0.00032211196958087385. Test_MSE: 0.00016252910424376183\n",
      "Epoch: 345 Iteration: 32800. Train_MSE: 0.00045970961218699813. Test_MSE: 0.00016211166516961827\n",
      "Epoch: 347 Iteration: 33000. Train_MSE: 0.003584229154512286. Test_MSE: 0.00016204719938175017\n",
      "Epoch: 349 Iteration: 33200. Train_MSE: 0.0003259861550759524. Test_MSE: 0.00016144003968957584\n",
      "Epoch: 351 Iteration: 33400. Train_MSE: 0.00017729161481838673. Test_MSE: 0.0001610277506502114\n",
      "Epoch: 353 Iteration: 33600. Train_MSE: 0.0007661987910978496. Test_MSE: 0.00016063018228558223\n",
      "Epoch: 355 Iteration: 33800. Train_MSE: 0.0032823679503053427. Test_MSE: 0.0001604104863996643\n",
      "Epoch: 357 Iteration: 34000. Train_MSE: 0.0003281881217844784. Test_MSE: 0.00016047518462170202\n",
      "Epoch: 360 Iteration: 34200. Train_MSE: 0.0004929825663566589. Test_MSE: 0.00015971650470810714\n",
      "Epoch: 362 Iteration: 34400. Train_MSE: 0.00026624335441738367. Test_MSE: 0.00015946366638034507\n",
      "Epoch: 364 Iteration: 34600. Train_MSE: 0.0007645647856406868. Test_MSE: 0.0001590378240118089\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 366 Iteration: 34800. Train_MSE: 0.0003035461122635752. Test_MSE: 0.00015891483801617887\n",
      "Epoch: 368 Iteration: 35000. Train_MSE: 0.00023158983094617724. Test_MSE: 0.000158540842647926\n",
      "Epoch: 370 Iteration: 35200. Train_MSE: 0.0003292987239547074. Test_MSE: 0.000158228178054087\n",
      "Epoch: 372 Iteration: 35400. Train_MSE: 0.0003114611026830971. Test_MSE: 0.00015749836294957857\n",
      "Epoch: 374 Iteration: 35600. Train_MSE: 0.00035578940878622234. Test_MSE: 0.0001572258823361059\n",
      "Epoch: 376 Iteration: 35800. Train_MSE: 0.00046435330295935273. Test_MSE: 0.0001573655193187588\n",
      "Epoch: 378 Iteration: 36000. Train_MSE: 0.0003626819234341383. Test_MSE: 0.00015694447785142578\n",
      "Epoch: 381 Iteration: 36200. Train_MSE: 0.0003351863124407828. Test_MSE: 0.00015651868663117192\n",
      "Epoch: 383 Iteration: 36400. Train_MSE: 0.0003038130817003548. Test_MSE: 0.00015614888308113542\n",
      "Epoch: 385 Iteration: 36600. Train_MSE: 0.00044844139483757317. Test_MSE: 0.00015576940488916258\n",
      "Epoch: 387 Iteration: 36800. Train_MSE: 0.003575886134058237. Test_MSE: 0.00015571970454594108\n",
      "Epoch: 389 Iteration: 37000. Train_MSE: 0.00030762498499825597. Test_MSE: 0.00015518774040612158\n",
      "Epoch: 391 Iteration: 37200. Train_MSE: 0.00016278514522127807. Test_MSE: 0.00015483188646087318\n",
      "Epoch: 393 Iteration: 37400. Train_MSE: 0.0007347261416725814. Test_MSE: 0.00015450916600348542\n",
      "Epoch: 395 Iteration: 37600. Train_MSE: 0.0032506913412362337. Test_MSE: 0.0001543122451909069\n",
      "Epoch: 397 Iteration: 37800. Train_MSE: 0.0003203360829502344. Test_MSE: 0.00015438475700759355\n",
      "Epoch: 400 Iteration: 38000. Train_MSE: 0.0004541309317573905. Test_MSE: 0.00015370175493611813\n",
      "Epoch: 402 Iteration: 38200. Train_MSE: 0.0002447041042614728. Test_MSE: 0.00015349849547045174\n",
      "Epoch: 404 Iteration: 38400. Train_MSE: 0.0007431184640154243. Test_MSE: 0.00015312093275227957\n",
      "Epoch: 406 Iteration: 38600. Train_MSE: 0.00029122692649252713. Test_MSE: 0.0001529996402797348\n",
      "Epoch: 408 Iteration: 38800. Train_MSE: 0.00020624909666366875. Test_MSE: 0.00015267097178293555\n",
      "Epoch: 410 Iteration: 39000. Train_MSE: 0.00030957465060055256. Test_MSE: 0.0001524059333310527\n",
      "Epoch: 412 Iteration: 39200. Train_MSE: 0.00028596934862434864. Test_MSE: 0.0001517556561907668\n",
      "Epoch: 414 Iteration: 39400. Train_MSE: 0.0003431103832554072. Test_MSE: 0.00015153211178937142\n",
      "Epoch: 416 Iteration: 39600. Train_MSE: 0.0004532876191660762. Test_MSE: 0.0001516591557400711\n",
      "Epoch: 418 Iteration: 39800. Train_MSE: 0.0003304783604107797. Test_MSE: 0.00015129163618125127\n",
      "Epoch: 421 Iteration: 40000. Train_MSE: 0.00031737476820126176. Test_MSE: 0.00015090675415124356\n",
      "Epoch: 423 Iteration: 40200. Train_MSE: 0.00028828310314565897. Test_MSE: 0.00015058028229490572\n",
      "Epoch: 425 Iteration: 40400. Train_MSE: 0.0004387158842291683. Test_MSE: 0.0001502275044392176\n",
      "Epoch: 427 Iteration: 40600. Train_MSE: 0.0035673407837748528. Test_MSE: 0.00015018562934743495\n",
      "Epoch: 429 Iteration: 40800. Train_MSE: 0.00029151610215194523. Test_MSE: 0.00014971317949514303\n",
      "Epoch: 431 Iteration: 41000. Train_MSE: 0.00015096311108209193. Test_MSE: 0.00014939894371979406\n",
      "Epoch: 433 Iteration: 41200. Train_MSE: 0.0007080241339281201. Test_MSE: 0.00014913687209638896\n",
      "Epoch: 435 Iteration: 41400. Train_MSE: 0.0032222685404121876. Test_MSE: 0.00014895813300191371\n",
      "Epoch: 437 Iteration: 41600. Train_MSE: 0.00031323765870183706. Test_MSE: 0.00014903777083623493\n",
      "Epoch: 440 Iteration: 41800. Train_MSE: 0.0004203811986371875. Test_MSE: 0.000148418035193019\n",
      "Epoch: 442 Iteration: 42000. Train_MSE: 0.00022649409947916865. Test_MSE: 0.00014825370048355519\n",
      "Epoch: 444 Iteration: 42200. Train_MSE: 0.0007248754845932126. Test_MSE: 0.0001479177661207985\n",
      "Epoch: 446 Iteration: 42400. Train_MSE: 0.0002813748433254659. Test_MSE: 0.0001477998546113147\n",
      "Epoch: 448 Iteration: 42600. Train_MSE: 0.00018511980306357145. Test_MSE: 0.0001475118645179532\n",
      "Epoch: 450 Iteration: 42800. Train_MSE: 0.0002925302542280406. Test_MSE: 0.00014728463211458417\n",
      "Epoch: 452 Iteration: 43000. Train_MSE: 0.000264914269791916. Test_MSE: 0.00014669903761333546\n",
      "Epoch: 454 Iteration: 43200. Train_MSE: 0.0003324110875837505. Test_MSE: 0.00014651663660056167\n",
      "Epoch: 456 Iteration: 43400. Train_MSE: 0.00044390230323188007. Test_MSE: 0.00014663453588849022\n",
      "Epoch: 458 Iteration: 43600. Train_MSE: 0.0003028341743629426. Test_MSE: 0.00014631790480465338\n",
      "Epoch: 461 Iteration: 43800. Train_MSE: 0.0003017825656570494. Test_MSE: 0.00014596739100135482\n",
      "Epoch: 463 Iteration: 44000. Train_MSE: 0.0002752503496594727. Test_MSE: 0.00014568136073592696\n",
      "Epoch: 465 Iteration: 44200. Train_MSE: 0.00043047082726843655. Test_MSE: 0.00014535783568566654\n",
      "Epoch: 467 Iteration: 44400. Train_MSE: 0.0035606950987130404. Test_MSE: 0.00014533016696614834\n",
      "Epoch: 469 Iteration: 44600. Train_MSE: 0.0002768360136542469. Test_MSE: 0.00014491177734626105\n",
      "Epoch: 471 Iteration: 44800. Train_MSE: 0.00014113768702372909. Test_MSE: 0.00014463229557246918\n",
      "Epoch: 473 Iteration: 45000. Train_MSE: 0.0006854334496892989. Test_MSE: 0.0001444213413948289\n",
      "Epoch: 475 Iteration: 45200. Train_MSE: 0.0031962778884917498. Test_MSE: 0.0001442591839640772\n",
      "Epoch: 477 Iteration: 45400. Train_MSE: 0.0003072976542171091. Test_MSE: 0.00014434468561420983\n",
      "Epoch: 480 Iteration: 45600. Train_MSE: 0.0003906083002220839. Test_MSE: 0.00014377830386527366\n",
      "Epoch: 482 Iteration: 45800. Train_MSE: 0.0002112538495566696. Test_MSE: 0.00014364358823585368\n",
      "Epoch: 484 Iteration: 46000. Train_MSE: 0.0007094035972841084. Test_MSE: 0.00014334320877456468\n",
      "Epoch: 486 Iteration: 46200. Train_MSE: 0.00027374751516617835. Test_MSE: 0.0001432271457529773\n",
      "Epoch: 488 Iteration: 46400. Train_MSE: 0.0001677013497101143. Test_MSE: 0.0001429747695474775\n",
      "Epoch: 490 Iteration: 46600. Train_MSE: 0.0002773737069219351. Test_MSE: 0.00014277903319052053\n",
      "Epoch: 492 Iteration: 46800. Train_MSE: 0.0002474733046256006. Test_MSE: 0.0001422485033244779\n",
      "Epoch: 494 Iteration: 47000. Train_MSE: 0.0003231377631891519. Test_MSE: 0.00014210198542451925\n",
      "Epoch: 496 Iteration: 47200. Train_MSE: 0.0004366719804238528. Test_MSE: 0.00014221380259626967\n",
      "Epoch: 498 Iteration: 47400. Train_MSE: 0.00027910276548936963. Test_MSE: 0.00014194063837945757\n",
      "Epoch: 501 Iteration: 47600. Train_MSE: 0.00028846540953963995. Test_MSE: 0.0001416172322593129\n",
      "Epoch: 503 Iteration: 47800. Train_MSE: 0.00026383931981399655. Test_MSE: 0.00014136344555608138\n",
      "Epoch: 505 Iteration: 48000. Train_MSE: 0.00042321678483858705. Test_MSE: 0.00014106006974080992\n",
      "Epoch: 507 Iteration: 48200. Train_MSE: 0.0035556084476411343. Test_MSE: 0.00014104130606357076\n",
      "Epoch: 509 Iteration: 48400. Train_MSE: 0.00026342360069975257. Test_MSE: 0.00014066688623803482\n",
      "Epoch: 511 Iteration: 48600. Train_MSE: 0.0001329197402810678. Test_MSE: 0.00014041437218952294\n",
      "Epoch: 513 Iteration: 48800. Train_MSE: 0.0006657863268628716. Test_MSE: 0.00014024440423870708\n",
      "Epoch: 515 Iteration: 49000. Train_MSE: 0.00317241158336401. Test_MSE: 0.00014009392393802325\n",
      "Epoch: 517 Iteration: 49200. Train_MSE: 0.000301860534818843. Test_MSE: 0.0001401836638140485\n",
      "Epoch: 520 Iteration: 49400. Train_MSE: 0.0003646856057457626. Test_MSE: 0.00013965920821881053\n",
      "Epoch: 522 Iteration: 49600. Train_MSE: 0.00019809012883342803. Test_MSE: 0.0001395464140364477\n",
      "Epoch: 524 Iteration: 49800. Train_MSE: 0.0006957396981306374. Test_MSE: 0.00013927389538449792\n",
      "Epoch: 526 Iteration: 50000. Train_MSE: 0.0002674441784620285. Test_MSE: 0.00013915807660749055\n",
      "Epoch: 528 Iteration: 50200. Train_MSE: 0.0001533211616333574. Test_MSE: 0.0001389363289880188\n",
      "Epoch: 530 Iteration: 50400. Train_MSE: 0.00026389153208583593. Test_MSE: 0.00013876360573983364\n",
      "Epoch: 532 Iteration: 50600. Train_MSE: 0.0002327112597413361. Test_MSE: 0.00013827586978899136\n",
      "Epoch: 534 Iteration: 50800. Train_MSE: 0.00031498834141530097. Test_MSE: 0.00013815641110388334\n",
      "Epoch: 536 Iteration: 51000. Train_MSE: 0.0004306727205403149. Test_MSE: 0.00013825901419276614\n",
      "Epoch: 538 Iteration: 51200. Train_MSE: 0.0002584360772743821. Test_MSE: 0.00013801975558096983\n",
      "Epoch: 541 Iteration: 51400. Train_MSE: 0.0002766952966339886. Test_MSE: 0.00013771527683851924\n",
      "Epoch: 543 Iteration: 51600. Train_MSE: 0.00025410926900804043. Test_MSE: 0.00013748559600467708\n",
      "Epoch: 545 Iteration: 51800. Train_MSE: 0.00041670529753901064. Test_MSE: 0.0001371982290833027\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 547 Iteration: 52000. Train_MSE: 0.0035528549924492836. Test_MSE: 0.00013718625206656824\n",
      "Epoch: 549 Iteration: 52200. Train_MSE: 0.0002511153288651258. Test_MSE: 0.00013684680918986812\n",
      "Epoch: 551 Iteration: 52400. Train_MSE: 0.00012580069596879184. Test_MSE: 0.00013661389298185787\n",
      "Epoch: 553 Iteration: 52600. Train_MSE: 0.0006487768259830773. Test_MSE: 0.00013647612053275352\n",
      "Epoch: 555 Iteration: 52800. Train_MSE: 0.00314994715154171. Test_MSE: 0.00013633316670149875\n",
      "Epoch: 557 Iteration: 53000. Train_MSE: 0.00029700304730795324. Test_MSE: 0.00013642609706429726\n",
      "Epoch: 560 Iteration: 53200. Train_MSE: 0.0003417095576878637. Test_MSE: 0.00013593747324463416\n",
      "Epoch: 562 Iteration: 53400. Train_MSE: 0.00018661684589460492. Test_MSE: 0.00013584095028115598\n",
      "Epoch: 564 Iteration: 53600. Train_MSE: 0.0006838106201030314. Test_MSE: 0.00013559198090335858\n",
      "Epoch: 566 Iteration: 53800. Train_MSE: 0.000261971348663792. Test_MSE: 0.00013547426759170746\n",
      "Epoch: 568 Iteration: 54000. Train_MSE: 0.00014120429113972932. Test_MSE: 0.00013527650834713203\n",
      "Epoch: 570 Iteration: 54200. Train_MSE: 0.00025168576394207776. Test_MSE: 0.0001351202879998701\n",
      "Epoch: 572 Iteration: 54400. Train_MSE: 0.00022011110559105873. Test_MSE: 0.00013466730082045198\n",
      "Epoch: 574 Iteration: 54600. Train_MSE: 0.0003077841829508543. Test_MSE: 0.0001345702723144387\n",
      "Epoch: 576 Iteration: 54800. Train_MSE: 0.0004255693347658962. Test_MSE: 0.00013466347370242644\n",
      "Epoch: 578 Iteration: 55000. Train_MSE: 0.0002401743840891868. Test_MSE: 0.00013445251004972136\n",
      "Epoch: 581 Iteration: 55200. Train_MSE: 0.00026616762625053525. Test_MSE: 0.00013416009266758989\n",
      "Epoch: 583 Iteration: 55400. Train_MSE: 0.0002455156354699284. Test_MSE: 0.00013394864741381415\n",
      "Epoch: 585 Iteration: 55600. Train_MSE: 0.0004108943394385278. Test_MSE: 0.0001336730230526246\n",
      "Epoch: 587 Iteration: 55800. Train_MSE: 0.00355077744461596. Test_MSE: 0.00013366449525575242\n",
      "Epoch: 589 Iteration: 56000. Train_MSE: 0.0002396632917225361. Test_MSE: 0.00013335255054005416\n",
      "Epoch: 591 Iteration: 56200. Train_MSE: 0.00011967412137892097. Test_MSE: 0.000133132604381804\n",
      "Epoch: 593 Iteration: 56400. Train_MSE: 0.0006338107632473111. Test_MSE: 0.00013302111894295806\n",
      "Epoch: 595 Iteration: 56600. Train_MSE: 0.0031247895676642656. Test_MSE: 0.00013288430417760283\n",
      "Epoch: 597 Iteration: 56800. Train_MSE: 0.0002926049055531621. Test_MSE: 0.0001329780716755258\n",
      "Epoch: 600 Iteration: 57000. Train_MSE: 0.00032092840410768986. Test_MSE: 0.00013251743021134378\n",
      "Epoch: 602 Iteration: 57200. Train_MSE: 0.00017654694966040552. Test_MSE: 0.00013243117737002803\n",
      "Epoch: 604 Iteration: 57400. Train_MSE: 0.000673186790663749. Test_MSE: 0.0001322001130672252\n",
      "Epoch: 606 Iteration: 57600. Train_MSE: 0.0002570324868429452. Test_MSE: 0.00013208034662106594\n",
      "Epoch: 608 Iteration: 57800. Train_MSE: 0.00013098615454509854. Test_MSE: 0.00013190460431237224\n",
      "Epoch: 610 Iteration: 58000. Train_MSE: 0.0002405139384791255. Test_MSE: 0.00013176187908379695\n",
      "Epoch: 612 Iteration: 58200. Train_MSE: 0.00020929791207890958. Test_MSE: 0.00013133820309879125\n",
      "Epoch: 614 Iteration: 58400. Train_MSE: 0.0003012824454344809. Test_MSE: 0.00013126045464262662\n",
      "Epoch: 616 Iteration: 58600. Train_MSE: 0.0004210863262414932. Test_MSE: 0.00013134565862059222\n",
      "Epoch: 618 Iteration: 58800. Train_MSE: 0.00022396797430701554. Test_MSE: 0.0001311595421573376\n",
      "Epoch: 621 Iteration: 59000. Train_MSE: 0.0002564483438618481. Test_MSE: 0.0001308781368302272\n",
      "Epoch: 623 Iteration: 59200. Train_MSE: 0.0002377138880547136. Test_MSE: 0.00013068443008308375\n",
      "Epoch: 625 Iteration: 59400. Train_MSE: 0.0004056946490891278. Test_MSE: 0.00013041973105296675\n",
      "Epoch: 627 Iteration: 59600. Train_MSE: 0.0035492880269885063. Test_MSE: 0.00013041664519959525\n",
      "Epoch: 629 Iteration: 59800. Train_MSE: 0.00022923681535758078. Test_MSE: 0.0001301313911164894\n",
      "Epoch: 631 Iteration: 60000. Train_MSE: 0.00011439430818427354. Test_MSE: 0.00012992261799890886\n",
      "Epoch: 633 Iteration: 60200. Train_MSE: 0.0006204742821864784. Test_MSE: 0.0001298341918717759\n",
      "Epoch: 635 Iteration: 60400. Train_MSE: 0.003101547947153449. Test_MSE: 0.00012970198016886146\n",
      "Epoch: 637 Iteration: 60600. Train_MSE: 0.0002888360759243369. Test_MSE: 0.00012979683207999613\n",
      "Epoch: 640 Iteration: 60800. Train_MSE: 0.0003023566969204694. Test_MSE: 0.00012936137168319497\n",
      "Epoch: 642 Iteration: 61000. Train_MSE: 0.00016772732487879694. Test_MSE: 0.00012928330203685958\n",
      "Epoch: 644 Iteration: 61200. Train_MSE: 0.0006636910839006305. Test_MSE: 0.00012906686226706046\n",
      "Epoch: 646 Iteration: 61400. Train_MSE: 0.0002528247714508325. Test_MSE: 0.00012894384470069512\n",
      "Epoch: 648 Iteration: 61600. Train_MSE: 0.00012259642244316638. Test_MSE: 0.00012878571593391303\n",
      "Epoch: 650 Iteration: 61800. Train_MSE: 0.00023029811563901603. Test_MSE: 0.0001286534694840142\n",
      "Epoch: 652 Iteration: 62000. Train_MSE: 0.00019983842503279448. Test_MSE: 0.00012825460008212343\n",
      "Epoch: 654 Iteration: 62200. Train_MSE: 0.00029542925767600536. Test_MSE: 0.00012819239410025979\n",
      "Epoch: 656 Iteration: 62400. Train_MSE: 0.0004171114123892039. Test_MSE: 0.00012826931373184224\n",
      "Epoch: 658 Iteration: 62600. Train_MSE: 0.0002095146046485752. Test_MSE: 0.00012810453081893872\n",
      "Epoch: 661 Iteration: 62800. Train_MSE: 0.0002475735091138631. Test_MSE: 0.00012783000756787883\n",
      "Epoch: 663 Iteration: 63000. Train_MSE: 0.00023076345678418875. Test_MSE: 0.0001276494170938807\n",
      "Epoch: 665 Iteration: 63200. Train_MSE: 0.00040090526454150677. Test_MSE: 0.00012739254196926056\n",
      "Epoch: 667 Iteration: 63400. Train_MSE: 0.0035480731166899204. Test_MSE: 0.0001273917090471723\n",
      "Epoch: 669 Iteration: 63600. Train_MSE: 0.00021965548512525856. Test_MSE: 0.00012712807137169624\n",
      "Epoch: 671 Iteration: 63800. Train_MSE: 0.00010982465755660087. Test_MSE: 0.00012693052265469587\n",
      "Epoch: 673 Iteration: 64000. Train_MSE: 0.0006083822809159756. Test_MSE: 0.00012686432791934857\n",
      "Epoch: 675 Iteration: 64200. Train_MSE: 0.0030800532549619675. Test_MSE: 0.0001267367684916672\n",
      "Epoch: 677 Iteration: 64400. Train_MSE: 0.00028536460013128817. Test_MSE: 0.0001268333539105731\n",
      "Epoch: 680 Iteration: 64600. Train_MSE: 0.0002855487400665879. Test_MSE: 0.00012642079903516477\n",
      "Epoch: 682 Iteration: 64800. Train_MSE: 0.00015989090024959296. Test_MSE: 0.00012635080421228442\n",
      "Epoch: 684 Iteration: 65000. Train_MSE: 0.0006552139529958367. Test_MSE: 0.0001261509508944073\n",
      "Epoch: 686 Iteration: 65200. Train_MSE: 0.00024925978505052626. Test_MSE: 0.00012602662059082613\n",
      "Epoch: 688 Iteration: 65400. Train_MSE: 0.00011536040983628482. Test_MSE: 0.0001258869905508626\n",
      "Epoch: 690 Iteration: 65600. Train_MSE: 0.0002209785598097369. Test_MSE: 0.00012576497259967346\n",
      "Epoch: 692 Iteration: 65800. Train_MSE: 0.00019148559658788145. Test_MSE: 0.00012538898282519096\n",
      "Epoch: 694 Iteration: 66000. Train_MSE: 0.00029018305940553546. Test_MSE: 0.0001253416274052814\n",
      "Epoch: 696 Iteration: 66200. Train_MSE: 0.00041356158908456564. Test_MSE: 0.00012541197999461527\n",
      "Epoch: 698 Iteration: 66400. Train_MSE: 0.00019666460866574198. Test_MSE: 0.00012526833119059895\n",
      "Epoch: 701 Iteration: 66600. Train_MSE: 0.00023944515851326287. Test_MSE: 0.000125000688119372\n",
      "Epoch: 703 Iteration: 66800. Train_MSE: 0.00022458084276877344. Test_MSE: 0.00012483236163403716\n",
      "Epoch: 705 Iteration: 67000. Train_MSE: 0.0003966779913753271. Test_MSE: 0.00012458331163853548\n",
      "Epoch: 707 Iteration: 67200. Train_MSE: 0.00354692991822958. Test_MSE: 0.00012458663931107366\n",
      "Epoch: 709 Iteration: 67400. Train_MSE: 0.00021085154730826616. Test_MSE: 0.00012434302193504003\n",
      "Epoch: 711 Iteration: 67600. Train_MSE: 0.00010583426046650857. Test_MSE: 0.00012415316581589002\n",
      "Epoch: 713 Iteration: 67800. Train_MSE: 0.0005973658990114927. Test_MSE: 0.00012410513120231223\n",
      "Epoch: 715 Iteration: 68000. Train_MSE: 0.003060045186430216. Test_MSE: 0.00012398010642745744\n",
      "Epoch: 717 Iteration: 68200. Train_MSE: 0.00028208119329065084. Test_MSE: 0.0001240780802721397\n",
      "Epoch: 720 Iteration: 68400. Train_MSE: 0.00027052333462052047. Test_MSE: 0.00012368535771606261\n",
      "Epoch: 722 Iteration: 68600. Train_MSE: 0.00015291049203369766. Test_MSE: 0.00012362115292975518\n",
      "Epoch: 724 Iteration: 68800. Train_MSE: 0.0006475074333138764. Test_MSE: 0.0001234349881703452\n",
      "Epoch: 726 Iteration: 69000. Train_MSE: 0.00024614075664430857. Test_MSE: 0.0001233083496698798\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 728 Iteration: 69200. Train_MSE: 0.00010911999561358243. Test_MSE: 0.00012318548794628697\n",
      "Epoch: 730 Iteration: 69400. Train_MSE: 0.00021242383809294552. Test_MSE: 0.00012307176598583738\n",
      "Epoch: 732 Iteration: 69600. Train_MSE: 0.00018408335745334625. Test_MSE: 0.00012271578821299915\n",
      "Epoch: 734 Iteration: 69800. Train_MSE: 0.00028543570078909397. Test_MSE: 0.00012268175903348827\n",
      "Epoch: 736 Iteration: 70000. Train_MSE: 0.00041037629125639796. Test_MSE: 0.00012274630699712013\n",
      "Epoch: 738 Iteration: 70200. Train_MSE: 0.00018506840569898486. Test_MSE: 0.00012262139872716351\n",
      "Epoch: 741 Iteration: 70400. Train_MSE: 0.0002319767518201843. Test_MSE: 0.00012235919600368745\n",
      "Epoch: 743 Iteration: 70600. Train_MSE: 0.00021902962180320174. Test_MSE: 0.00012220180031096418\n",
      "Epoch: 745 Iteration: 70800. Train_MSE: 0.00039285593084059656. Test_MSE: 0.00012195900595929441\n",
      "Epoch: 747 Iteration: 71000. Train_MSE: 0.0035461466759443283. Test_MSE: 0.00012196577955839673\n",
      "Epoch: 749 Iteration: 71200. Train_MSE: 0.0002026456786552444. Test_MSE: 0.0001217406697159026\n",
      "Epoch: 751 Iteration: 71400. Train_MSE: 0.00010234067303827032. Test_MSE: 0.00012155796604635721\n",
      "Epoch: 753 Iteration: 71600. Train_MSE: 0.0005872743204236031. Test_MSE: 0.00012152692432112162\n",
      "Epoch: 755 Iteration: 71800. Train_MSE: 0.0030414878856390715. Test_MSE: 0.00012140455092293564\n",
      "Epoch: 757 Iteration: 72000. Train_MSE: 0.00027894965023733675. Test_MSE: 0.00012150448927597077\n",
      "Epoch: 760 Iteration: 72200. Train_MSE: 0.0002567544870544225. Test_MSE: 0.00012112965351168849\n",
      "Epoch: 762 Iteration: 72400. Train_MSE: 0.0001465956593165174. Test_MSE: 0.00012107001556195922\n",
      "Epoch: 764 Iteration: 72600. Train_MSE: 0.0006405109306797385. Test_MSE: 0.00012089653616227895\n",
      "Epoch: 766 Iteration: 72800. Train_MSE: 0.0002434463967802003. Test_MSE: 0.00012076820229972277\n",
      "Epoch: 768 Iteration: 73000. Train_MSE: 0.00010355378617532551. Test_MSE: 0.00012066003506617201\n",
      "Epoch: 770 Iteration: 73200. Train_MSE: 0.0002044870489044115. Test_MSE: 0.00012055373860110012\n",
      "Epoch: 772 Iteration: 73400. Train_MSE: 0.00017744921206030995. Test_MSE: 0.00012021568191459761\n",
      "Epoch: 774 Iteration: 73600. Train_MSE: 0.0002813976025208831. Test_MSE: 0.00012019344982566351\n",
      "Epoch: 776 Iteration: 73800. Train_MSE: 0.00040749486652202904. Test_MSE: 0.00012025222414922453\n",
      "Epoch: 778 Iteration: 74000. Train_MSE: 0.00017451017629355192. Test_MSE: 0.00012014424213511957\n",
      "Epoch: 781 Iteration: 74200. Train_MSE: 0.00022505498782265931. Test_MSE: 0.00011988716035727699\n",
      "Epoch: 783 Iteration: 74400. Train_MSE: 0.0002140783763024956. Test_MSE: 0.00011974032695866861\n",
      "Epoch: 785 Iteration: 74600. Train_MSE: 0.000389349355828017. Test_MSE: 0.00011950467071878804\n",
      "Epoch: 787 Iteration: 74800. Train_MSE: 0.0035466060508042574. Test_MSE: 0.00011951530597081452\n",
      "Epoch: 789 Iteration: 75000. Train_MSE: 0.00019498038454912603. Test_MSE: 0.00011930785237115925\n",
      "Epoch: 791 Iteration: 75200. Train_MSE: 9.928031067829579e-05. Test_MSE: 0.00011913057271468533\n",
      "Epoch: 793 Iteration: 75400. Train_MSE: 0.0005780926439911127. Test_MSE: 0.00011911306312196325\n",
      "Epoch: 795 Iteration: 75600. Train_MSE: 0.0030243429355323315. Test_MSE: 0.00011899051300257057\n",
      "Epoch: 797 Iteration: 75800. Train_MSE: 0.0002760248025879264. Test_MSE: 0.00011908919324853207\n",
      "Epoch: 800 Iteration: 76000. Train_MSE: 0.0002441935648676008. Test_MSE: 0.00011872916491552027\n",
      "Epoch: 802 Iteration: 76200. Train_MSE: 0.00014100573025643826. Test_MSE: 0.00011867096469310585\n",
      "Epoch: 804 Iteration: 76400. Train_MSE: 0.0006341918488033116. Test_MSE: 0.00011850502518373494\n",
      "Epoch: 806 Iteration: 76600. Train_MSE: 0.0002411210589343682. Test_MSE: 0.00011837199211893718\n",
      "Epoch: 808 Iteration: 76800. Train_MSE: 9.869461791822687e-05. Test_MSE: 0.0001182755522687983\n",
      "Epoch: 810 Iteration: 77000. Train_MSE: 0.00019713750225491822. Test_MSE: 0.00011817342198791005\n",
      "Epoch: 812 Iteration: 77200. Train_MSE: 0.00017152349755633622. Test_MSE: 0.00011784931083789392\n",
      "Epoch: 814 Iteration: 77400. Train_MSE: 0.00027777624200098217. Test_MSE: 0.00011783431876176886\n",
      "Epoch: 816 Iteration: 77600. Train_MSE: 0.00040492601692676544. Test_MSE: 0.00011788446495414142\n",
      "Epoch: 818 Iteration: 77800. Train_MSE: 0.000165001634741202. Test_MSE: 0.00011778936969119077\n",
      "Epoch: 821 Iteration: 78000. Train_MSE: 0.00021860886772628874. Test_MSE: 0.00011753367584910808\n",
      "Epoch: 823 Iteration: 78200. Train_MSE: 0.00020960815891157836. Test_MSE: 0.00011739170444313643\n",
      "Epoch: 825 Iteration: 78400. Train_MSE: 0.0003861405421048403. Test_MSE: 0.00011715884772136815\n",
      "Epoch: 827 Iteration: 78600. Train_MSE: 0.003547247964888811. Test_MSE: 0.00011717031040505938\n",
      "Epoch: 829 Iteration: 78800. Train_MSE: 0.00018781813560053706. Test_MSE: 0.00011697560302331036\n",
      "Epoch: 831 Iteration: 79000. Train_MSE: 9.65722429100424e-05. Test_MSE: 0.00011680231066636107\n",
      "Epoch: 833 Iteration: 79200. Train_MSE: 0.0005696138250641525. Test_MSE: 0.00011679793710201449\n",
      "Epoch: 835 Iteration: 79400. Train_MSE: 0.0030082636512815952. Test_MSE: 0.00011667701521687379\n",
      "Epoch: 837 Iteration: 79600. Train_MSE: 0.00027329789008945227. Test_MSE: 0.00011677752064245857\n",
      "Epoch: 840 Iteration: 79800. Train_MSE: 0.0002326702669961378. Test_MSE: 0.00011643157253964479\n",
      "Epoch: 842 Iteration: 80000. Train_MSE: 0.00013593827316071838. Test_MSE: 0.00011637559684402338\n",
      "Epoch: 844 Iteration: 80200. Train_MSE: 0.0006283629918470979. Test_MSE: 0.00011621858843092819\n",
      "Epoch: 846 Iteration: 80400. Train_MSE: 0.00023904634872451425. Test_MSE: 0.00011608290898313814\n",
      "Epoch: 848 Iteration: 80600. Train_MSE: 9.451393998460844e-05. Test_MSE: 0.00011599860779823134\n",
      "Epoch: 850 Iteration: 80800. Train_MSE: 0.0001902500371215865. Test_MSE: 0.00011589946067634256\n",
      "Epoch: 852 Iteration: 81000. Train_MSE: 0.00016624586714897305. Test_MSE: 0.00011558842864664793\n",
      "Epoch: 854 Iteration: 81200. Train_MSE: 0.00027441210113465786. Test_MSE: 0.00011558226491105131\n",
      "Epoch: 856 Iteration: 81400. Train_MSE: 0.00040267250733450055. Test_MSE: 0.00011562666269292067\n",
      "Epoch: 858 Iteration: 81600. Train_MSE: 0.00015641491336282343. Test_MSE: 0.00011554511160650287\n",
      "Epoch: 861 Iteration: 81800. Train_MSE: 0.0002127590705640614. Test_MSE: 0.00011529092887690179\n",
      "Epoch: 863 Iteration: 82000. Train_MSE: 0.0002055130898952484. Test_MSE: 0.00011515481459695544\n",
      "Epoch: 865 Iteration: 82200. Train_MSE: 0.0003832065558526665. Test_MSE: 0.00011492541196648499\n",
      "Epoch: 867 Iteration: 82400. Train_MSE: 0.003547960426658392. Test_MSE: 0.00011493877229250919\n",
      "Epoch: 869 Iteration: 82600. Train_MSE: 0.00018113810801878572. Test_MSE: 0.00011475602452147534\n",
      "Epoch: 871 Iteration: 82800. Train_MSE: 9.41490288823843e-05. Test_MSE: 0.00011458514979306245\n",
      "Epoch: 873 Iteration: 83000. Train_MSE: 0.0005617118440568447. Test_MSE: 0.00011459198226669313\n",
      "Epoch: 875 Iteration: 83200. Train_MSE: 0.002993182046338916. Test_MSE: 0.00011447130568039409\n",
      "Epoch: 877 Iteration: 83400. Train_MSE: 0.0002708244719542563. Test_MSE: 0.00011457266773651333\n",
      "Epoch: 880 Iteration: 83600. Train_MSE: 0.0002221025206381455. Test_MSE: 0.00011423964815388064\n",
      "Epoch: 882 Iteration: 83800. Train_MSE: 0.00013135767949279398. Test_MSE: 0.00011418461576639578\n",
      "Epoch: 884 Iteration: 84000. Train_MSE: 0.0006230365252122283. Test_MSE: 0.00011403566370840384\n",
      "Epoch: 886 Iteration: 84200. Train_MSE: 0.000237189611652866. Test_MSE: 0.0001138972042293605\n",
      "Epoch: 888 Iteration: 84400. Train_MSE: 9.086745558306575e-05. Test_MSE: 0.00011382453547089878\n",
      "Epoch: 890 Iteration: 84600. Train_MSE: 0.0001838023163145408. Test_MSE: 0.00011372953084447659\n",
      "Epoch: 892 Iteration: 84800. Train_MSE: 0.0001614797511138022. Test_MSE: 0.00011343109788263138\n",
      "Epoch: 894 Iteration: 85000. Train_MSE: 0.00027130558737553656. Test_MSE: 0.00011343345649722856\n",
      "Epoch: 896 Iteration: 85200. Train_MSE: 0.00040064382483251393. Test_MSE: 0.00011347240329595945\n",
      "Epoch: 898 Iteration: 85400. Train_MSE: 0.00014854369510430843. Test_MSE: 0.00011340406051337026\n",
      "Epoch: 901 Iteration: 85600. Train_MSE: 0.00020726598449982703. Test_MSE: 0.00011315200079724762\n",
      "Epoch: 903 Iteration: 85800. Train_MSE: 0.0002017860097112134. Test_MSE: 0.00011302200155959241\n",
      "Epoch: 905 Iteration: 86000. Train_MSE: 0.0003803481813520193. Test_MSE: 0.00011279691577921548\n",
      "Epoch: 907 Iteration: 86200. Train_MSE: 0.0035488170105963945. Test_MSE: 0.00011281288633904795\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 909 Iteration: 86400. Train_MSE: 0.0001749241928337142. Test_MSE: 0.0001126429900032378\n",
      "Epoch: 911 Iteration: 86600. Train_MSE: 9.194156882585958e-05. Test_MSE: 0.0001124755594015487\n",
      "Epoch: 913 Iteration: 86800. Train_MSE: 0.0005542532890103757. Test_MSE: 0.00011249296350164502\n",
      "Epoch: 915 Iteration: 87000. Train_MSE: 0.0029788773972541094. Test_MSE: 0.00011237214048376817\n",
      "Epoch: 917 Iteration: 87200. Train_MSE: 0.0002685431099962443. Test_MSE: 0.00011247326830122926\n",
      "Epoch: 920 Iteration: 87400. Train_MSE: 0.00021243715309537947. Test_MSE: 0.00011215107578342248\n",
      "Epoch: 922 Iteration: 87600. Train_MSE: 0.00012721138773486018. Test_MSE: 0.00011209595446968472\n",
      "Epoch: 924 Iteration: 87800. Train_MSE: 0.0006184277008287609. Test_MSE: 0.0001119539926171645\n",
      "Epoch: 926 Iteration: 88000. Train_MSE: 0.0002354812022531405. Test_MSE: 0.0001118121182718872\n",
      "Epoch: 928 Iteration: 88200. Train_MSE: 8.767643157625571e-05. Test_MSE: 0.00011174935259330426\n",
      "Epoch: 930 Iteration: 88400. Train_MSE: 0.00017778387700673193. Test_MSE: 0.00011165662108240941\n",
      "Epoch: 932 Iteration: 88600. Train_MSE: 0.00015718245413154364. Test_MSE: 0.00011136894309328986\n",
      "Epoch: 934 Iteration: 88800. Train_MSE: 0.0002683374041225761. Test_MSE: 0.00011137762306617049\n",
      "Epoch: 936 Iteration: 89000. Train_MSE: 0.0003988263779319823. Test_MSE: 0.00011141044841119911\n",
      "Epoch: 938 Iteration: 89200. Train_MSE: 0.00014127398026175797. Test_MSE: 0.00011135322827354002\n",
      "Epoch: 941 Iteration: 89400. Train_MSE: 0.00020203806343488395. Test_MSE: 0.00011110170715558803\n",
      "Epoch: 943 Iteration: 89600. Train_MSE: 0.00019838594016619027. Test_MSE: 0.00011097604698216026\n",
      "Epoch: 945 Iteration: 89800. Train_MSE: 0.00037767470348626375. Test_MSE: 0.00011075343430565447\n",
      "Epoch: 947 Iteration: 90000. Train_MSE: 0.003549800021573901. Test_MSE: 0.0001107704110811581\n",
      "Epoch: 949 Iteration: 90200. Train_MSE: 0.00016907161625567824. Test_MSE: 0.00011061113695922878\n",
      "Epoch: 951 Iteration: 90400. Train_MSE: 8.995604002848268e-05. Test_MSE: 0.00011044514500723986\n",
      "Epoch: 953 Iteration: 90600. Train_MSE: 0.0005472936318255961. Test_MSE: 0.00011047166134109646\n",
      "Epoch: 955 Iteration: 90800. Train_MSE: 0.0029652530793100595. Test_MSE: 0.00011035128319628114\n",
      "Epoch: 957 Iteration: 91000. Train_MSE: 0.0002664166095200926. Test_MSE: 0.00011045295407527406\n",
      "Epoch: 960 Iteration: 91200. Train_MSE: 0.00020348478574305773. Test_MSE: 0.0001101419692688685\n",
      "Epoch: 962 Iteration: 91400. Train_MSE: 0.00012344602146185935. Test_MSE: 0.00011008731529998658\n",
      "Epoch: 964 Iteration: 91600. Train_MSE: 0.0006142263300716877. Test_MSE: 0.00010995160066746731\n",
      "Epoch: 966 Iteration: 91800. Train_MSE: 0.00023395154858008027. Test_MSE: 0.00010980645587685074\n",
      "Epoch: 968 Iteration: 92000. Train_MSE: 8.486653678119183e-05. Test_MSE: 0.00010975273283096428\n",
      "Epoch: 970 Iteration: 92200. Train_MSE: 0.0001721159933367744. Test_MSE: 0.00010966133668528385\n",
      "Epoch: 972 Iteration: 92400. Train_MSE: 0.00015324207197409123. Test_MSE: 0.00010938278564614381\n",
      "Epoch: 974 Iteration: 92600. Train_MSE: 0.0002655656717251986. Test_MSE: 0.00010939728706140261\n",
      "Epoch: 976 Iteration: 92800. Train_MSE: 0.00039718818152323365. Test_MSE: 0.00010942457679618559\n",
      "Epoch: 978 Iteration: 93000. Train_MSE: 0.00013453868450596929. Test_MSE: 0.00010937797682122847\n",
      "Epoch: 981 Iteration: 93200. Train_MSE: 0.00019704998703673482. Test_MSE: 0.00010912695223392693\n",
      "Epoch: 983 Iteration: 93400. Train_MSE: 0.00019529661221895367. Test_MSE: 0.00010900536720937103\n",
      "Epoch: 985 Iteration: 93600. Train_MSE: 0.0003750771575141698. Test_MSE: 0.00010878542067951795\n",
      "Epoch: 987 Iteration: 93800. Train_MSE: 0.0035507730208337307. Test_MSE: 0.00010880365885836941\n",
      "Epoch: 989 Iteration: 94000. Train_MSE: 0.0001635314110899344. Test_MSE: 0.00010865406145796168\n",
      "Epoch: 991 Iteration: 94200. Train_MSE: 8.816764602670446e-05. Test_MSE: 0.00010848957763978823\n",
      "Epoch: 993 Iteration: 94400. Train_MSE: 0.0005407168064266443. Test_MSE: 0.00010852497361035552\n",
      "Epoch: 995 Iteration: 94600. Train_MSE: 0.00295230932533741. Test_MSE: 0.0001084045591817115\n",
      "Epoch: 997 Iteration: 94800. Train_MSE: 0.0002644174383021891. Test_MSE: 0.00010850638823648126\n"
     ]
    }
   ],
   "source": [
    "# Training the model\n",
    "seq_dim = 1\n",
    "\n",
    "n_iter =0\n",
    "num_samples = len(X_train)\n",
    "test_samples = len(X_test)\n",
    "batch_size = 100\n",
    "num_epochs = 1000\n",
    "feat_dim = X_train.shape[1]\n",
    "\n",
    "X_train = X_train.type(torch.FloatTensor)\n",
    "y_train = y_train.type(torch.FloatTensor)\n",
    "X_test = X_test.type(torch.FloatTensor)\n",
    "y_test = y_test.type(torch.FloatTensor)\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    for i in range(0, int(num_samples/batch_size -1)):\n",
    "        \n",
    "        \n",
    "        features = Variable(X_train[i*batch_size:(i+1)*batch_size, :]).view(-1, seq_dim, feat_dim)\n",
    "        Kt_value = Variable(y_train[i*batch_size:(i+1)*batch_size])\n",
    "        \n",
    "        #print(\"Kt_value={}\".format(Kt_value))\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        outputs = model(features)\n",
    "        #print(\"outputs ={}\".format(outputs))\n",
    "        \n",
    "        loss = criterion(outputs, Kt_value)\n",
    "        \n",
    "        train_loss.append(loss.data[0])\n",
    "        train_iter.append(n_iter)\n",
    "\n",
    "        #print(\"loss = {}\".format(loss))\n",
    "        loss.backward()\n",
    "        \n",
    "        optimizer.step()\n",
    "        \n",
    "         \n",
    "            \n",
    "        if n_iter%200 == 0:\n",
    "            for i in range(0,int(test_samples/batch_size -1)):\n",
    "                features = Variable(X_test[i*batch_size:(i+1)*batch_size, :]).view(-1, seq_dim, feat_dim)\n",
    "                Kt_test = Variable(y_test[i*batch_size:(i+1)*batch_size])\n",
    "                \n",
    "                outputs = model(features)\n",
    "                \n",
    "                mse = np.sqrt(np.mean((Kt_test.data.numpy() - outputs.data.numpy().squeeze())**2)/num_samples)\n",
    "                \n",
    "                test_iter.append(n_iter)\n",
    "                test_loss.append(mse)\n",
    "                \n",
    "            print('Epoch: {} Iteration: {}. Train_MSE: {}. Test_MSE: {}'.format(epoch, n_iter, loss.data[0], mse))\n",
    "        \n",
    "        n_iter += 1 \n",
    "               "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19000\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x169cb3c8>]"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD3CAYAAADxJYRbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ni4gERKEvIhKQ/w+Ujt/whOGorAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(len(test_loss))\n",
    "#plt.plot(test_loss)\n",
    "plt.plot(train_loss,'-')\n",
    "#plt.ylim([0.000,0.99])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0xcf19940>]"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAD3CAYAAAAQYlNPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(test_loss,'r')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "#### Demornamization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "rmse = np.sqrt(mse)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "rmse_denorm = (rmse * (df_new_test['Kt'].max() - df_new_test['Kt'].min()))+ df_new_test['Kt'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17.810984861107954"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse_denorm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    4001.000000\n",
       "mean        5.870793\n",
       "std        30.554212\n",
       "min        -0.030773\n",
       "25%        -0.007258\n",
       "50%         0.166769\n",
       "75%         1.241848\n",
       "max      1146.230573\n",
       "Name: Kt, dtype: float64"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_new_test['Kt'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Saving train and test losses to a csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_trainLoss = pd.DataFrame(data={'Train Loss':train_loss, 'iteration':train_iter}, columns=['Train Loss','iteration'])\n",
    "df_trainLoss.to_csv('RNN Paper Results/Exp1_Bondville_TrainLoss.csv')\n",
    "df_testLoss = pd.DataFrame(data={'Test Loss':test_loss,'iteration':test_iter}, columns=['Test Loss','iteration'])\n",
    "df_testLoss.to_csv('RNN Paper Results/Exp1_Bondville_TestLoss.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}