Forecasting Code for Mutivariate Time Series Data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Forecasting Code for Mutivariate Time Series Data\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Getting Started\n",
    "\n",
    "In this notebook, we will create an AI and time serie driven forecasting engine based on a set of 5 AI models and 5 time series models and  employ several algorithms to perform feature engennering and selection on a multivaraite time series dataset. We are trying to find the best drivers of the HK flat price using various macro economics data.The dataset we are using is the 'HK macroeconomics data' from Kaggle. It is a daily dataset starting from 2nd January 2003 and goes on till 26th November 2019. The target variable is 'Private Domestic (Price Index)' and there 12 features which can help us forecast HK flat price. The dataset can be found [here](https://www.kaggle.com/stanley11291985/hk-macroeconomics-data?)\n",
    "\n",
    "\n",
    "We will first do exploratory data analysis on the data and then move on to data cleaning where we do the following:\n",
    "\n",
    "1. Check if any dates are missing and add the missing dates\n",
    "2. Use linear interpolation to fill the nulls (we will also provide the code for forward and backward fill\n",
    "3. Perform outlier detection\n",
    "4. Perform outlier treatment\n",
    "5. Check time series components, ACF and PACF plots of target variable\n",
    "\n",
    "Once the data has been cleaned up, we will create the following additional features:\n",
    "\n",
    "1. 3 lags for each feature\n",
    "2. Daily change for each feature\n",
    "3. A flag for holidays\n",
    "5. Flag for S&P 500 index\n",
    "4. Calculate First hand sales price and drop the sales amount variable\n",
    "\n",
    "The next step is to perform featue selection using the following techniques:\n",
    "\n",
    "1. Select the top n features based on feature importance from random forest\n",
    "2. Select the top n features based on absolute correlation with target variable\n",
    "3. Select the features identified by Lasso regression\n",
    "4. Select features by recursively considering smaller and smaller sets of features\n",
    "5. Select the top n features based on absolute value of beta coefficients of features\n",
    "\n",
    "The final features are the ones selected in at least 3 out of 5 models.\n",
    "\n",
    "After feature selection, we train a set of 10 models (and perform hyper parameter tuning wherever required):\n",
    "\n",
    "1. SARIMA (Univariate Time Series)\n",
    "2. Holt Winters (Univariate Time Series)\n",
    "3. SARIMAX(Multi-variate Time Series)\n",
    "4. VAR(Multi-variate Time Series)\n",
    "5. VECM(Multi-variate Time Series)\n",
    "6. LSTM (Univariate Time Series)\n",
    "7. Random Forest (Multi-variate AI)\n",
    "8. XG Boost (Multi-variate AI)\n",
    "9. Linear Regression (Multi-variate AI)\n",
    "10. SVR (Multi-variate AI)\n",
    "\n",
    "We make predictions for the next year using all the models and then take their average as the final predictions."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Exploring the Data\n",
    "In this section, we load necessary Python libraries and load the multivariate time series data. Note that the column 'Private Domestic (Price Index)' from this dataset will be our target label. All other columns are features. We perform EDA like visualising and understaind the mean, min, max etc of the fields."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Necessary Libraries\n",
    "\n",
    "#!pip install pyod\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn import linear_model\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.linear_model import Lasso, LassoCV\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.feature_selection import SelectFromModel\n",
    "from sklearn.feature_selection import RFE\n",
    "from sklearn.feature_selection import RFECV\n",
    "import statsmodels.api as sm\n",
    "from scipy.stats.mstats import zscore\n",
    "from pyod.models.abod import ABOD\n",
    "from pyod.models.knn import KNN\n",
    "import seaborn as sns\n",
    "from pandas.tseries.holiday import USFederalHolidayCalendar as calendar\n",
    "from pandas import Series\n",
    "from matplotlib import pyplot\n",
    "\n",
    "import pmdarima as pm\n",
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "from statsmodels.graphics.tsaplots import plot_acf\n",
    "from statsmodels.graphics.tsaplots import plot_pacf\n",
    "from statsmodels.tsa.vector_ar.var_model import VAR\n",
    "from statsmodels.tsa.statespace.sarimax import SARIMAX\n",
    "from statsmodels.tsa.stattools import grangercausalitytests\n",
    "from statsmodels.tsa.arima_model import ARIMA\n",
    "from statsmodels.tsa.holtwinters import ExponentialSmoothing\n",
    "from statsmodels.tsa.holtwinters import SimpleExpSmoothing\n",
    "from statsmodels.tsa.statespace.sarimax import SARIMAX\n",
    "from statsmodels.tsa.statespace.varmax import VARMAX\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from  statsmodels.tsa.vector_ar.vecm import *\n",
    "\n",
    "\n",
    "import itertools\n",
    "import math\n",
    "import random\n",
    "import xgboost as xgb\n",
    "import tensorflow\n",
    "import keras\n",
    "from keras.layers import Dropout\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense\n",
    "from keras.layers import LSTM\n",
    "from keras.preprocessing.sequence import TimeseriesGenerator\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.svm import SVR\n",
    "\n",
    "#!pip install --trusted-host=pypi.org --trusted-host=files.pythonhosted.org --user pmdarima"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>Date</th>\n",
       "      <th>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2003-01-02</td>\n",
       "      <td>64.559769</td>\n",
       "      <td>1854.689208</td>\n",
       "      <td>5.846897e+09</td>\n",
       "      <td>33983.762413</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.44420</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>126907400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2003-01-03</td>\n",
       "      <td>64.506848</td>\n",
       "      <td>1892.762372</td>\n",
       "      <td>5.927523e+09</td>\n",
       "      <td>33932.818955</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.41741</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>291454400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2003-01-06</td>\n",
       "      <td>64.453848</td>\n",
       "      <td>1931.338866</td>\n",
       "      <td>5.998876e+09</td>\n",
       "      <td>33882.041253</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.40402</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>234912800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2003-01-07</td>\n",
       "      <td>64.400768</td>\n",
       "      <td>1970.418692</td>\n",
       "      <td>6.060958e+09</td>\n",
       "      <td>33831.429307</td>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>19872.438591</td>\n",
       "      <td>74111.243068</td>\n",
       "      <td>7.294081</td>\n",
       "      <td>73.708301</td>\n",
       "      <td>1.39063</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>9652.400391</td>\n",
       "      <td>267021800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2003-01-08</td>\n",
       "      <td>64.347610</td>\n",
       "      <td>2010.001848</td>\n",
       "      <td>6.113768e+09</td>\n",
       "      <td>33780.983117</td>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>19858.946877</td>\n",
       "      <td>74089.062450</td>\n",
       "      <td>7.291601</td>\n",
       "      <td>73.703930</td>\n",
       "      <td>1.39732</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>9688.209961</td>\n",
       "      <td>202439200.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Date  Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0 2003-01-02                       64.559769                1854.689208   \n",
       "1 2003-01-03                       64.506848                1892.762372   \n",
       "2 2003-01-06                       64.453848                1931.338866   \n",
       "3 2003-01-07                       64.400768                1970.418692   \n",
       "4 2003-01-08                       64.347610                2010.001848   \n",
       "\n",
       "   First hand sales amount  Total completions   Total stock  Total take up  \\\n",
       "0             5.846897e+09       33983.762413  1.092844e+06   19914.083323   \n",
       "1             5.927523e+09       33932.818955  1.092244e+06   19900.006814   \n",
       "2             5.998876e+09       33882.041253  1.091646e+06   19886.125236   \n",
       "3             6.060958e+09       33831.429307  1.091050e+06   19872.438591   \n",
       "4             6.113768e+09       33780.983117  1.090457e+06   19858.946877   \n",
       "\n",
       "   Total vacancy  Unemployment rate (seasonally adjusted) (%)        CPI  \\\n",
       "0   74177.818318                                     7.301525  73.721486   \n",
       "1   74155.621001                                     7.299043  73.717079   \n",
       "2   74133.429251                                     7.296561  73.712684   \n",
       "3   74111.243068                                     7.294081  73.708301   \n",
       "4   74089.062450                                     7.291601  73.703930   \n",
       "\n",
       "   HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  \n",
       "0          1.44420      3.562189e+06  9365.519531   126907400.0  \n",
       "1          1.41741      3.562586e+06  9583.849609   291454400.0  \n",
       "2          1.40402      3.563044e+06  9665.959961   234912800.0  \n",
       "3          1.39063      3.563562e+06  9652.400391   267021800.0  \n",
       "4          1.39732      3.564141e+06  9688.209961   202439200.0  "
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Input the dataset\n",
    "df1 = pd.read_excel('Housing market data.xlsx',\n",
    "              sheet_name='Upsampled')\n",
    "\n",
    "df1.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4233 entries, 0 to 4232\n",
      "Data columns (total 14 columns):\n",
      " #   Column                                       Non-Null Count  Dtype         \n",
      "---  ------                                       --------------  -----         \n",
      " 0   Date                                         4233 non-null   datetime64[ns]\n",
      " 1   Private Domestic (Price Index)               4233 non-null   float64       \n",
      " 2   First hand sales quantity                    4233 non-null   float64       \n",
      " 3   First hand sales amount                      4233 non-null   float64       \n",
      " 4   Total completions                            4233 non-null   float64       \n",
      " 5   Total stock                                  4233 non-null   float64       \n",
      " 6   Total take up                                4233 non-null   float64       \n",
      " 7   Total vacancy                                4233 non-null   float64       \n",
      " 8   Unemployment rate (seasonally adjusted) (%)  4233 non-null   float64       \n",
      " 9   CPI                                          4233 non-null   float64       \n",
      " 10  HIBOR (1-month)                              4233 non-null   float64       \n",
      " 11  M3 (HK$ million)                             4233 non-null   float64       \n",
      " 12  HSI - close                                  4233 non-null   float64       \n",
      " 13  HSI - volume                                 4233 non-null   float64       \n",
      "dtypes: datetime64[ns](1), float64(13)\n",
      "memory usage: 463.1 KB\n"
     ]
    }
   ],
   "source": [
    "#Check number of rows and columns, type of each columns\n",
    "df1.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "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>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4.233000e+03</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4.233000e+03</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4.233000e+03</td>\n",
       "      <td>4233.000000</td>\n",
       "      <td>4.233000e+03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>197.064993</td>\n",
       "      <td>1406.732017</td>\n",
       "      <td>1.145308e+10</td>\n",
       "      <td>15387.011646</td>\n",
       "      <td>1.106643e+06</td>\n",
       "      <td>14036.055785</td>\n",
       "      <td>52835.021903</td>\n",
       "      <td>4.202867</td>\n",
       "      <td>88.188533</td>\n",
       "      <td>1.180317</td>\n",
       "      <td>8.375635e+06</td>\n",
       "      <td>20781.490429</td>\n",
       "      <td>1.492191e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>104.268475</td>\n",
       "      <td>758.310792</td>\n",
       "      <td>6.934301e+09</td>\n",
       "      <td>6413.960753</td>\n",
       "      <td>5.073772e+04</td>\n",
       "      <td>6483.743342</td>\n",
       "      <td>9013.878325</td>\n",
       "      <td>1.273481</td>\n",
       "      <td>12.713668</td>\n",
       "      <td>1.415258</td>\n",
       "      <td>3.553012e+06</td>\n",
       "      <td>5303.246267</td>\n",
       "      <td>9.137858e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>58.400000</td>\n",
       "      <td>-2.690601</td>\n",
       "      <td>4.424446e+05</td>\n",
       "      <td>6792.576078</td>\n",
       "      <td>1.006630e+06</td>\n",
       "      <td>-4077.411618</td>\n",
       "      <td>41975.928483</td>\n",
       "      <td>2.800000</td>\n",
       "      <td>71.700000</td>\n",
       "      <td>0.040180</td>\n",
       "      <td>3.548561e+06</td>\n",
       "      <td>8409.009766</td>\n",
       "      <td>0.000000e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>96.834930</td>\n",
       "      <td>863.244685</td>\n",
       "      <td>6.124111e+09</td>\n",
       "      <td>9942.919441</td>\n",
       "      <td>1.073121e+06</td>\n",
       "      <td>9516.701115</td>\n",
       "      <td>45758.833205</td>\n",
       "      <td>3.134502</td>\n",
       "      <td>75.520451</td>\n",
       "      <td>0.212860</td>\n",
       "      <td>5.279677e+06</td>\n",
       "      <td>17048.419922</td>\n",
       "      <td>9.293672e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>181.300974</td>\n",
       "      <td>1292.923682</td>\n",
       "      <td>1.038705e+10</td>\n",
       "      <td>13810.857627</td>\n",
       "      <td>1.106639e+06</td>\n",
       "      <td>12693.239693</td>\n",
       "      <td>49973.213330</td>\n",
       "      <td>3.649037</td>\n",
       "      <td>86.187916</td>\n",
       "      <td>0.345000</td>\n",
       "      <td>7.660015e+06</td>\n",
       "      <td>21654.160156</td>\n",
       "      <td>1.495348e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>284.135754</td>\n",
       "      <td>1845.392958</td>\n",
       "      <td>1.568220e+10</td>\n",
       "      <td>19112.456257</td>\n",
       "      <td>1.142288e+06</td>\n",
       "      <td>16988.464263</td>\n",
       "      <td>62994.641229</td>\n",
       "      <td>5.077025</td>\n",
       "      <td>100.297864</td>\n",
       "      <td>1.747140</td>\n",
       "      <td>1.147862e+07</td>\n",
       "      <td>23919.949219</td>\n",
       "      <td>1.947201e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>397.100022</td>\n",
       "      <td>4151.298770</td>\n",
       "      <td>3.834408e+10</td>\n",
       "      <td>33983.762413</td>\n",
       "      <td>1.214837e+06</td>\n",
       "      <td>31536.196716</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>8.500000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>5.703570</td>\n",
       "      <td>1.470108e+07</td>\n",
       "      <td>33154.121094</td>\n",
       "      <td>9.799120e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "count                     4233.000000                4233.000000   \n",
       "mean                       197.064993                1406.732017   \n",
       "std                        104.268475                 758.310792   \n",
       "min                         58.400000                  -2.690601   \n",
       "25%                         96.834930                 863.244685   \n",
       "50%                        181.300974                1292.923682   \n",
       "75%                        284.135754                1845.392958   \n",
       "max                        397.100022                4151.298770   \n",
       "\n",
       "       First hand sales amount  Total completions   Total stock  \\\n",
       "count             4.233000e+03        4233.000000  4.233000e+03   \n",
       "mean              1.145308e+10       15387.011646  1.106643e+06   \n",
       "std               6.934301e+09        6413.960753  5.073772e+04   \n",
       "min               4.424446e+05        6792.576078  1.006630e+06   \n",
       "25%               6.124111e+09        9942.919441  1.073121e+06   \n",
       "50%               1.038705e+10       13810.857627  1.106639e+06   \n",
       "75%               1.568220e+10       19112.456257  1.142288e+06   \n",
       "max               3.834408e+10       33983.762413  1.214837e+06   \n",
       "\n",
       "       Total take up  Total vacancy  \\\n",
       "count    4233.000000    4233.000000   \n",
       "mean    14036.055785   52835.021903   \n",
       "std      6483.743342    9013.878325   \n",
       "min     -4077.411618   41975.928483   \n",
       "25%      9516.701115   45758.833205   \n",
       "50%     12693.239693   49973.213330   \n",
       "75%     16988.464263   62994.641229   \n",
       "max     31536.196716   74177.818318   \n",
       "\n",
       "       Unemployment rate (seasonally adjusted) (%)          CPI  \\\n",
       "count                                  4233.000000  4233.000000   \n",
       "mean                                      4.202867    88.188533   \n",
       "std                                       1.273481    12.713668   \n",
       "min                                       2.800000    71.700000   \n",
       "25%                                       3.134502    75.520451   \n",
       "50%                                       3.649037    86.187916   \n",
       "75%                                       5.077025   100.297864   \n",
       "max                                       8.500000   111.000000   \n",
       "\n",
       "       HIBOR (1-month)  M3 (HK$ million)   HSI - close  HSI - volume  \n",
       "count      4233.000000      4.233000e+03   4233.000000  4.233000e+03  \n",
       "mean          1.180317      8.375635e+06  20781.490429  1.492191e+09  \n",
       "std           1.415258      3.553012e+06   5303.246267  9.137858e+08  \n",
       "min           0.040180      3.548561e+06   8409.009766  0.000000e+00  \n",
       "25%           0.212860      5.279677e+06  17048.419922  9.293672e+08  \n",
       "50%           0.345000      7.660015e+06  21654.160156  1.495348e+09  \n",
       "75%           1.747140      1.147862e+07  23919.949219  1.947201e+09  \n",
       "max           5.703570      1.470108e+07  33154.121094  9.799120e+09  "
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Understanding the numeric fields\n",
    "df1.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Histogram of all numeric fields\n",
    "df_hist = df1.drop(columns=['Date'],axis=1)\n",
    "df_hist.hist(figsize=(15,15));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\cbook\\__init__.py:1377: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n",
      "  x[:, None]\n",
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axes\\_base.py:237: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n",
      "  x = x[:, np.newaxis]\n",
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axes\\_base.py:239: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n",
      "  y = y[:, np.newaxis]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x24ce5bdbac8>]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error in callback <function install_repl_displayhook.<locals>.post_execute at 0x0000024CD8FD0400> (for post_execute):\n"
     ]
    },
    {
     "ename": "OverflowError",
     "evalue": "Python int too large to convert to C long",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mOverflowError\u001b[0m                             Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\pyplot.py\u001b[0m in \u001b[0;36mpost_execute\u001b[1;34m()\u001b[0m\n\u001b[0;32m    107\u001b[0m             \u001b[1;32mdef\u001b[0m \u001b[0mpost_execute\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    108\u001b[0m                 \u001b[1;32mif\u001b[0m \u001b[0mmatplotlib\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mis_interactive\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 109\u001b[1;33m                     \u001b[0mdraw_all\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    110\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    111\u001b[0m             \u001b[1;31m# IPython >= 2\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\_pylab_helpers.py\u001b[0m in \u001b[0;36mdraw_all\u001b[1;34m(cls, force)\u001b[0m\n\u001b[0;32m    130\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mf_mgr\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mcls\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_all_fig_managers\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    131\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mforce\u001b[0m \u001b[1;32mor\u001b[0m \u001b[0mf_mgr\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcanvas\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstale\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 132\u001b[1;33m                 \u001b[0mf_mgr\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcanvas\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdraw_idle\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    133\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    134\u001b[0m \u001b[0matexit\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mregister\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mGcf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdestroy_all\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\backend_bases.py\u001b[0m in \u001b[0;36mdraw_idle\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1897\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_is_idle_drawing\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1898\u001b[0m             \u001b[1;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_idle_draw_cntx\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1899\u001b[1;33m                 \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1900\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1901\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mdraw_cursor\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mevent\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\backends\\backend_agg.py\u001b[0m in \u001b[0;36mdraw\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    400\u001b[0m         \u001b[0mtoolbar\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtoolbar\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    401\u001b[0m         \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 402\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    403\u001b[0m             \u001b[1;31m# A GUI class may be need to update a window using this draw, so\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    404\u001b[0m             \u001b[1;31m# don't forget to call the superclass.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\artist.py\u001b[0m in \u001b[0;36mdraw_wrapper\u001b[1;34m(artist, renderer, *args, **kwargs)\u001b[0m\n\u001b[0;32m     48\u001b[0m                 \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstart_filter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     49\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 50\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0martist\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     51\u001b[0m         \u001b[1;32mfinally\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     52\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0martist\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_agg_filter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\figure.py\u001b[0m in \u001b[0;36mdraw\u001b[1;34m(self, renderer)\u001b[0m\n\u001b[0;32m   1647\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1648\u001b[0m             mimage._draw_list_compositing_images(\n\u001b[1;32m-> 1649\u001b[1;33m                 renderer, self, artists, self.suppressComposite)\n\u001b[0m\u001b[0;32m   1650\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1651\u001b[0m             \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mclose_group\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'figure'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\image.py\u001b[0m in \u001b[0;36m_draw_list_compositing_images\u001b[1;34m(renderer, parent, artists, suppress_composite)\u001b[0m\n\u001b[0;32m    136\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mnot_composite\u001b[0m \u001b[1;32mor\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mhas_images\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    137\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0ma\u001b[0m \u001b[1;32min\u001b[0m \u001b[0martists\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 138\u001b[1;33m             \u001b[0ma\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    139\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    140\u001b[0m         \u001b[1;31m# Composite any adjacent images together\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\artist.py\u001b[0m in \u001b[0;36mdraw_wrapper\u001b[1;34m(artist, renderer, *args, **kwargs)\u001b[0m\n\u001b[0;32m     48\u001b[0m                 \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstart_filter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     49\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 50\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0martist\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     51\u001b[0m         \u001b[1;32mfinally\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     52\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0martist\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_agg_filter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axes\\_base.py\u001b[0m in \u001b[0;36mdraw\u001b[1;34m(self, renderer, inframe)\u001b[0m\n\u001b[0;32m   2626\u001b[0m             \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstop_rasterizing\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2627\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 2628\u001b[1;33m         \u001b[0mmimage\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_draw_list_compositing_images\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0martists\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   2629\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   2630\u001b[0m         \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mclose_group\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'axes'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\image.py\u001b[0m in \u001b[0;36m_draw_list_compositing_images\u001b[1;34m(renderer, parent, artists, suppress_composite)\u001b[0m\n\u001b[0;32m    136\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[0mnot_composite\u001b[0m \u001b[1;32mor\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mhas_images\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    137\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0ma\u001b[0m \u001b[1;32min\u001b[0m \u001b[0martists\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 138\u001b[1;33m             \u001b[0ma\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    139\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    140\u001b[0m         \u001b[1;31m# Composite any adjacent images together\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\artist.py\u001b[0m in \u001b[0;36mdraw_wrapper\u001b[1;34m(artist, renderer, *args, **kwargs)\u001b[0m\n\u001b[0;32m     48\u001b[0m                 \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstart_filter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     49\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 50\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mdraw\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0martist\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mrenderer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     51\u001b[0m         \u001b[1;32mfinally\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     52\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0martist\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_agg_filter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axis.py\u001b[0m in \u001b[0;36mdraw\u001b[1;34m(self, renderer, *args, **kwargs)\u001b[0m\n\u001b[0;32m   1183\u001b[0m         \u001b[0mrenderer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mopen_group\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m__name__\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1184\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1185\u001b[1;33m         \u001b[0mticks_to_draw\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_update_ticks\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrenderer\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1186\u001b[0m         ticklabelBoxes, ticklabelBoxes2 = self._get_tick_bboxes(ticks_to_draw,\n\u001b[0;32m   1187\u001b[0m                                                                 renderer)\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axis.py\u001b[0m in \u001b[0;36m_update_ticks\u001b[1;34m(self, renderer)\u001b[0m\n\u001b[0;32m   1021\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1022\u001b[0m         \u001b[0minterval\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_view_interval\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1023\u001b[1;33m         \u001b[0mtick_tups\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlist\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0miter_ticks\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# iter_ticks calls the locator\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1024\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_smart_bounds\u001b[0m \u001b[1;32mand\u001b[0m \u001b[0mtick_tups\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1025\u001b[0m             \u001b[1;31m# handle inverted limits\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axis.py\u001b[0m in \u001b[0;36miter_ticks\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    965\u001b[0m         \u001b[0mIterate\u001b[0m \u001b[0mthrough\u001b[0m \u001b[0mall\u001b[0m \u001b[0mof\u001b[0m \u001b[0mthe\u001b[0m \u001b[0mmajor\u001b[0m \u001b[1;32mand\u001b[0m \u001b[0mminor\u001b[0m \u001b[0mticks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    966\u001b[0m         \"\"\"\n\u001b[1;32m--> 967\u001b[1;33m         \u001b[0mmajorLocs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmajor\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlocator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    968\u001b[0m         \u001b[0mmajorTicks\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_major_ticks\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmajorLocs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    969\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmajor\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformatter\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mset_locs\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mmajorLocs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1228\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1229\u001b[0m         \u001b[1;34m'Return the locations of the ticks'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1230\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrefresh\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1231\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_locator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1232\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36mrefresh\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1248\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mrefresh\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1249\u001b[0m         \u001b[1;34m'Refresh internal information based on current limits.'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1250\u001b[1;33m         \u001b[0mdmin\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdmax\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mviewlim_to_dt\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1251\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_locator\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_locator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdmin\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdmax\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1252\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36mviewlim_to_dt\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1000\u001b[0m                              \u001b[1;34m'value to an axis that has datetime units'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1001\u001b[0m                              .format(vmin))\n\u001b[1;32m-> 1002\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mnum2date\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvmin\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtz\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnum2date\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvmax\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtz\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1003\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1004\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_get_unit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36mnum2date\u001b[1;34m(x, tz)\u001b[0m\n\u001b[0;32m    496\u001b[0m         \u001b[0mtz\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_get_rc_timezone\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    497\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mcbook\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0miterable\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 498\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0m_from_ordinalf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtz\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    499\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    500\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36m_from_ordinalf\u001b[1;34m(x, tz)\u001b[0m\n\u001b[0;32m    295\u001b[0m                          \u001b[1;34m'non-datetime values are passed to an axis that '\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    296\u001b[0m                          'expects datetime objects.'.format(ix))\n\u001b[1;32m--> 297\u001b[1;33m     \u001b[0mdt\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdatetime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdatetime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfromordinal\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mix\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreplace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtzinfo\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mUTC\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    298\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    299\u001b[0m     \u001b[1;31m# Since the input date `x` float is unable to preserve microsecond\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mOverflowError\u001b[0m: Python int too large to convert to C long"
     ]
    },
    {
     "ename": "OverflowError",
     "evalue": "Python int too large to convert to C long",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mOverflowError\u001b[0m                             Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel\\pylab\\backend_inline.py\u001b[0m in \u001b[0;36mshow\u001b[1;34m(close, block)\u001b[0m\n\u001b[0;32m     37\u001b[0m             display(\n\u001b[0;32m     38\u001b[0m                 \u001b[0mfigure_manager\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcanvas\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 39\u001b[1;33m                 \u001b[0mmetadata\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0m_fetch_figure_metadata\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfigure_manager\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcanvas\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     40\u001b[0m             )\n\u001b[0;32m     41\u001b[0m     \u001b[1;32mfinally\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel\\pylab\\backend_inline.py\u001b[0m in \u001b[0;36m_fetch_figure_metadata\u001b[1;34m(fig)\u001b[0m\n\u001b[0;32m    175\u001b[0m         \u001b[1;31m# the background is transparent\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    176\u001b[0m         ticksLight = _is_light([label.get_color()\n\u001b[1;32m--> 177\u001b[1;33m                                 \u001b[1;32mfor\u001b[0m \u001b[0maxes\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mfig\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0maxes\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    178\u001b[0m                                 \u001b[1;32mfor\u001b[0m \u001b[0maxis\u001b[0m \u001b[1;32min\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0maxes\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mxaxis\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0maxes\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0myaxis\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    179\u001b[0m                                 for label in axis.get_ticklabels()])\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel\\pylab\\backend_inline.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[1;34m(.0)\u001b[0m\n\u001b[0;32m    177\u001b[0m                                 \u001b[1;32mfor\u001b[0m \u001b[0maxes\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mfig\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0maxes\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    178\u001b[0m                                 \u001b[1;32mfor\u001b[0m \u001b[0maxis\u001b[0m \u001b[1;32min\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0maxes\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mxaxis\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0maxes\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0myaxis\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 179\u001b[1;33m                                 for label in axis.get_ticklabels()])\n\u001b[0m\u001b[0;32m    180\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mticksLight\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msize\u001b[0m \u001b[1;32mand\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0mticksLight\u001b[0m \u001b[1;33m==\u001b[0m \u001b[0mticksLight\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mall\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    181\u001b[0m             \u001b[1;31m# there are one or more tick labels, all with the same lightness\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axis.py\u001b[0m in \u001b[0;36mget_ticklabels\u001b[1;34m(self, minor, which)\u001b[0m\n\u001b[0;32m   1284\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mminor\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1285\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_minorticklabels\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1286\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_majorticklabels\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1287\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1288\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mget_majorticklines\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axis.py\u001b[0m in \u001b[0;36mget_majorticklabels\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1238\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mget_majorticklabels\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1239\u001b[0m         \u001b[1;34m'Return a list of Text instances for the major ticklabels'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1240\u001b[1;33m         \u001b[0mticks\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_major_ticks\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1241\u001b[0m         \u001b[0mlabels1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mtick\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlabel1\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mtick\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mticks\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mtick\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlabel1On\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1242\u001b[0m         \u001b[0mlabels2\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mtick\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlabel2\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mtick\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mticks\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mtick\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlabel2On\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\axis.py\u001b[0m in \u001b[0;36mget_major_ticks\u001b[1;34m(self, numticks)\u001b[0m\n\u001b[0;32m   1389\u001b[0m         \u001b[1;34m'get the tick instances; grow as necessary'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1390\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mnumticks\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1391\u001b[1;33m             \u001b[0mnumticks\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_major_locator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1392\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1393\u001b[0m         \u001b[1;32mwhile\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmajorTicks\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m<\u001b[0m \u001b[0mnumticks\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1228\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m__call__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1229\u001b[0m         \u001b[1;34m'Return the locations of the ticks'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1230\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrefresh\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1231\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_locator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1232\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36mrefresh\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1248\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mrefresh\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1249\u001b[0m         \u001b[1;34m'Refresh internal information based on current limits.'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1250\u001b[1;33m         \u001b[0mdmin\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdmax\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mviewlim_to_dt\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1251\u001b[0m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_locator\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_locator\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdmin\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdmax\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1252\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36mviewlim_to_dt\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m   1000\u001b[0m                              \u001b[1;34m'value to an axis that has datetime units'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1001\u001b[0m                              .format(vmin))\n\u001b[1;32m-> 1002\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mnum2date\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvmin\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtz\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnum2date\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvmax\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtz\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1003\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1004\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_get_unit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36mnum2date\u001b[1;34m(x, tz)\u001b[0m\n\u001b[0;32m    496\u001b[0m         \u001b[0mtz\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_get_rc_timezone\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    497\u001b[0m     \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mcbook\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0miterable\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 498\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0m_from_ordinalf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtz\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    499\u001b[0m     \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    500\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Annaconda\\lib\\site-packages\\matplotlib\\dates.py\u001b[0m in \u001b[0;36m_from_ordinalf\u001b[1;34m(x, tz)\u001b[0m\n\u001b[0;32m    295\u001b[0m                          \u001b[1;34m'non-datetime values are passed to an axis that '\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    296\u001b[0m                          'expects datetime objects.'.format(ix))\n\u001b[1;32m--> 297\u001b[1;33m     \u001b[0mdt\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdatetime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdatetime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfromordinal\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mix\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreplace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtzinfo\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mUTC\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    298\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    299\u001b[0m     \u001b[1;31m# Since the input date `x` float is unable to preserve microsecond\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mOverflowError\u001b[0m: Python int too large to convert to C long"
     ]
    }
   ],
   "source": [
    "# Visaluise the target variable\n",
    "plt.plot(df1['Date'], df1['Private Domestic (Price Index)'])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cleaning the Data\n",
    "\n",
    "In this section we do the following:\n",
    "\n",
    "1. Check if any dates are missing and add the missing dates\n",
    "2. Use linear interpolation to fill the nulls (we will also provide the code for forward and backward fill\n",
    "3. Perform outlier detection\n",
    "4. Perform outlier treatment\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel_launcher.py:4: FutureWarning: Treating datetime data as categorical rather than numeric in `.describe` is deprecated and will be removed in a future version of pandas. Specify `datetime_is_numeric=True` to silence this warning and adopt the future behavior now.\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "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>Date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>6173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>6173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>2006-07-20 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>first</th>\n",
       "      <td>2003-01-02 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last</th>\n",
       "      <td>2019-11-26 00:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Date\n",
       "count                  6173\n",
       "unique                 6173\n",
       "top     2006-07-20 00:00:00\n",
       "freq                      1\n",
       "first   2003-01-02 00:00:00\n",
       "last    2019-11-26 00:00:00"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check if any dates are missing\n",
    "daily_data = pd.DataFrame(pd.date_range(start=df1['Date'].min(),end=df1['Date'].max()))\n",
    "daily_data.rename(columns={ daily_data.columns[0]: \"Date\" }, inplace = True)\n",
    "daily_data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 6173 records between \"2003-01-02\" and \"2019-11-26\" but the original data only had 4233 record. So a few dates are missing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "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>Date</th>\n",
       "      <th>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2003-01-02</td>\n",
       "      <td>64.559769</td>\n",
       "      <td>1854.689208</td>\n",
       "      <td>5.846897e+09</td>\n",
       "      <td>33983.762413</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.44420</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>126907400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2003-01-03</td>\n",
       "      <td>64.506848</td>\n",
       "      <td>1892.762372</td>\n",
       "      <td>5.927523e+09</td>\n",
       "      <td>33932.818955</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.41741</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>291454400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>2003-01-04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>2003-01-05</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2003-01-06</td>\n",
       "      <td>64.453848</td>\n",
       "      <td>1931.338866</td>\n",
       "      <td>5.998876e+09</td>\n",
       "      <td>33882.041253</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.40402</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>234912800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2003-01-07</td>\n",
       "      <td>64.400768</td>\n",
       "      <td>1970.418692</td>\n",
       "      <td>6.060958e+09</td>\n",
       "      <td>33831.429307</td>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>19872.438591</td>\n",
       "      <td>74111.243068</td>\n",
       "      <td>7.294081</td>\n",
       "      <td>73.708301</td>\n",
       "      <td>1.39063</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>9652.400391</td>\n",
       "      <td>267021800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2003-01-08</td>\n",
       "      <td>64.347610</td>\n",
       "      <td>2010.001848</td>\n",
       "      <td>6.113768e+09</td>\n",
       "      <td>33780.983117</td>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>19858.946877</td>\n",
       "      <td>74089.062450</td>\n",
       "      <td>7.291601</td>\n",
       "      <td>73.703930</td>\n",
       "      <td>1.39732</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>9688.209961</td>\n",
       "      <td>202439200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2003-01-09</td>\n",
       "      <td>64.294373</td>\n",
       "      <td>2050.088335</td>\n",
       "      <td>6.157306e+09</td>\n",
       "      <td>33730.702683</td>\n",
       "      <td>1.089865e+06</td>\n",
       "      <td>19845.650096</td>\n",
       "      <td>74066.887400</td>\n",
       "      <td>7.289122</td>\n",
       "      <td>73.699570</td>\n",
       "      <td>1.43080</td>\n",
       "      <td>3.564781e+06</td>\n",
       "      <td>9675.410156</td>\n",
       "      <td>129623200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2003-01-10</td>\n",
       "      <td>64.241057</td>\n",
       "      <td>2090.678153</td>\n",
       "      <td>6.191572e+09</td>\n",
       "      <td>33680.588005</td>\n",
       "      <td>1.089275e+06</td>\n",
       "      <td>19832.548246</td>\n",
       "      <td>74044.717915</td>\n",
       "      <td>7.286643</td>\n",
       "      <td>73.695223</td>\n",
       "      <td>1.43973</td>\n",
       "      <td>3.565481e+06</td>\n",
       "      <td>9721.500000</td>\n",
       "      <td>172862200.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4235</th>\n",
       "      <td>2003-01-11</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Date  Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0    2003-01-02                       64.559769                1854.689208   \n",
       "1    2003-01-03                       64.506848                1892.762372   \n",
       "4233 2003-01-04                             NaN                        NaN   \n",
       "4234 2003-01-05                             NaN                        NaN   \n",
       "2    2003-01-06                       64.453848                1931.338866   \n",
       "3    2003-01-07                       64.400768                1970.418692   \n",
       "4    2003-01-08                       64.347610                2010.001848   \n",
       "5    2003-01-09                       64.294373                2050.088335   \n",
       "6    2003-01-10                       64.241057                2090.678153   \n",
       "4235 2003-01-11                             NaN                        NaN   \n",
       "\n",
       "      First hand sales amount  Total completions   Total stock  Total take up  \\\n",
       "0                5.846897e+09       33983.762413  1.092844e+06   19914.083323   \n",
       "1                5.927523e+09       33932.818955  1.092244e+06   19900.006814   \n",
       "4233                      NaN                NaN           NaN            NaN   \n",
       "4234                      NaN                NaN           NaN            NaN   \n",
       "2                5.998876e+09       33882.041253  1.091646e+06   19886.125236   \n",
       "3                6.060958e+09       33831.429307  1.091050e+06   19872.438591   \n",
       "4                6.113768e+09       33780.983117  1.090457e+06   19858.946877   \n",
       "5                6.157306e+09       33730.702683  1.089865e+06   19845.650096   \n",
       "6                6.191572e+09       33680.588005  1.089275e+06   19832.548246   \n",
       "4235                      NaN                NaN           NaN            NaN   \n",
       "\n",
       "      Total vacancy  Unemployment rate (seasonally adjusted) (%)        CPI  \\\n",
       "0      74177.818318                                     7.301525  73.721486   \n",
       "1      74155.621001                                     7.299043  73.717079   \n",
       "4233            NaN                                          NaN        NaN   \n",
       "4234            NaN                                          NaN        NaN   \n",
       "2      74133.429251                                     7.296561  73.712684   \n",
       "3      74111.243068                                     7.294081  73.708301   \n",
       "4      74089.062450                                     7.291601  73.703930   \n",
       "5      74066.887400                                     7.289122  73.699570   \n",
       "6      74044.717915                                     7.286643  73.695223   \n",
       "4235            NaN                                          NaN        NaN   \n",
       "\n",
       "      HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  \n",
       "0             1.44420      3.562189e+06  9365.519531   126907400.0  \n",
       "1             1.41741      3.562586e+06  9583.849609   291454400.0  \n",
       "4233              NaN               NaN          NaN           NaN  \n",
       "4234              NaN               NaN          NaN           NaN  \n",
       "2             1.40402      3.563044e+06  9665.959961   234912800.0  \n",
       "3             1.39063      3.563562e+06  9652.400391   267021800.0  \n",
       "4             1.39732      3.564141e+06  9688.209961   202439200.0  \n",
       "5             1.43080      3.564781e+06  9675.410156   129623200.0  \n",
       "6             1.43973      3.565481e+06  9721.500000   172862200.0  \n",
       "4235              NaN               NaN          NaN           NaN  "
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add the missing dates\n",
    "df2 = pd.merge(df1,daily_data,on='Date',how='outer')\n",
    "df2 = df2.sort_values(by=['Date'])\n",
    "df2.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "scrolled": false
   },
   "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>Date</th>\n",
       "      <th>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2003-01-02</td>\n",
       "      <td>64.559769</td>\n",
       "      <td>1854.689208</td>\n",
       "      <td>5.846897e+09</td>\n",
       "      <td>33983.762413</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.444200</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>1.269074e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2003-01-03</td>\n",
       "      <td>64.506848</td>\n",
       "      <td>1892.762372</td>\n",
       "      <td>5.927523e+09</td>\n",
       "      <td>33932.818955</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.417410</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>2.914544e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>2003-01-04</td>\n",
       "      <td>64.489181</td>\n",
       "      <td>1905.621203</td>\n",
       "      <td>5.951307e+09</td>\n",
       "      <td>33915.893054</td>\n",
       "      <td>1.092045e+06</td>\n",
       "      <td>19895.379621</td>\n",
       "      <td>74148.223751</td>\n",
       "      <td>7.298216</td>\n",
       "      <td>73.715614</td>\n",
       "      <td>1.412947</td>\n",
       "      <td>3.562739e+06</td>\n",
       "      <td>9611.219726</td>\n",
       "      <td>2.726072e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>2003-01-05</td>\n",
       "      <td>64.471514</td>\n",
       "      <td>1918.480035</td>\n",
       "      <td>5.975092e+09</td>\n",
       "      <td>33898.967153</td>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>19890.752429</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.297389</td>\n",
       "      <td>73.714149</td>\n",
       "      <td>1.408483</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>9638.589844</td>\n",
       "      <td>2.537600e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2003-01-06</td>\n",
       "      <td>64.453848</td>\n",
       "      <td>1931.338866</td>\n",
       "      <td>5.998876e+09</td>\n",
       "      <td>33882.041253</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.404020</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>2.349128e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2003-01-07</td>\n",
       "      <td>64.400768</td>\n",
       "      <td>1970.418692</td>\n",
       "      <td>6.060958e+09</td>\n",
       "      <td>33831.429307</td>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>19872.438591</td>\n",
       "      <td>74111.243068</td>\n",
       "      <td>7.294081</td>\n",
       "      <td>73.708301</td>\n",
       "      <td>1.390630</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>9652.400391</td>\n",
       "      <td>2.670218e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2003-01-08</td>\n",
       "      <td>64.347610</td>\n",
       "      <td>2010.001848</td>\n",
       "      <td>6.113768e+09</td>\n",
       "      <td>33780.983117</td>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>19858.946877</td>\n",
       "      <td>74089.062450</td>\n",
       "      <td>7.291601</td>\n",
       "      <td>73.703930</td>\n",
       "      <td>1.397320</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>9688.209961</td>\n",
       "      <td>2.024392e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2003-01-09</td>\n",
       "      <td>64.294373</td>\n",
       "      <td>2050.088335</td>\n",
       "      <td>6.157306e+09</td>\n",
       "      <td>33730.702683</td>\n",
       "      <td>1.089865e+06</td>\n",
       "      <td>19845.650096</td>\n",
       "      <td>74066.887400</td>\n",
       "      <td>7.289122</td>\n",
       "      <td>73.699570</td>\n",
       "      <td>1.430800</td>\n",
       "      <td>3.564781e+06</td>\n",
       "      <td>9675.410156</td>\n",
       "      <td>1.296232e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2003-01-10</td>\n",
       "      <td>64.241057</td>\n",
       "      <td>2090.678153</td>\n",
       "      <td>6.191572e+09</td>\n",
       "      <td>33680.588005</td>\n",
       "      <td>1.089275e+06</td>\n",
       "      <td>19832.548246</td>\n",
       "      <td>74044.717915</td>\n",
       "      <td>7.286643</td>\n",
       "      <td>73.695223</td>\n",
       "      <td>1.439730</td>\n",
       "      <td>3.565481e+06</td>\n",
       "      <td>9721.500000</td>\n",
       "      <td>1.728622e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4235</th>\n",
       "      <td>2003-01-11</td>\n",
       "      <td>64.223258</td>\n",
       "      <td>2104.375869</td>\n",
       "      <td>6.199903e+09</td>\n",
       "      <td>33663.938365</td>\n",
       "      <td>1.089080e+06</td>\n",
       "      <td>19828.245940</td>\n",
       "      <td>74037.329943</td>\n",
       "      <td>7.285817</td>\n",
       "      <td>73.693777</td>\n",
       "      <td>1.436010</td>\n",
       "      <td>3.565735e+06</td>\n",
       "      <td>9759.026693</td>\n",
       "      <td>1.812307e+08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Date  Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0    2003-01-02                       64.559769                1854.689208   \n",
       "1    2003-01-03                       64.506848                1892.762372   \n",
       "4233 2003-01-04                       64.489181                1905.621203   \n",
       "4234 2003-01-05                       64.471514                1918.480035   \n",
       "2    2003-01-06                       64.453848                1931.338866   \n",
       "3    2003-01-07                       64.400768                1970.418692   \n",
       "4    2003-01-08                       64.347610                2010.001848   \n",
       "5    2003-01-09                       64.294373                2050.088335   \n",
       "6    2003-01-10                       64.241057                2090.678153   \n",
       "4235 2003-01-11                       64.223258                2104.375869   \n",
       "\n",
       "      First hand sales amount  Total completions   Total stock  Total take up  \\\n",
       "0                5.846897e+09       33983.762413  1.092844e+06   19914.083323   \n",
       "1                5.927523e+09       33932.818955  1.092244e+06   19900.006814   \n",
       "4233             5.951307e+09       33915.893054  1.092045e+06   19895.379621   \n",
       "4234             5.975092e+09       33898.967153  1.091846e+06   19890.752429   \n",
       "2                5.998876e+09       33882.041253  1.091646e+06   19886.125236   \n",
       "3                6.060958e+09       33831.429307  1.091050e+06   19872.438591   \n",
       "4                6.113768e+09       33780.983117  1.090457e+06   19858.946877   \n",
       "5                6.157306e+09       33730.702683  1.089865e+06   19845.650096   \n",
       "6                6.191572e+09       33680.588005  1.089275e+06   19832.548246   \n",
       "4235             6.199903e+09       33663.938365  1.089080e+06   19828.245940   \n",
       "\n",
       "      Total vacancy  Unemployment rate (seasonally adjusted) (%)        CPI  \\\n",
       "0      74177.818318                                     7.301525  73.721486   \n",
       "1      74155.621001                                     7.299043  73.717079   \n",
       "4233   74148.223751                                     7.298216  73.715614   \n",
       "4234   74140.826501                                     7.297389  73.714149   \n",
       "2      74133.429251                                     7.296561  73.712684   \n",
       "3      74111.243068                                     7.294081  73.708301   \n",
       "4      74089.062450                                     7.291601  73.703930   \n",
       "5      74066.887400                                     7.289122  73.699570   \n",
       "6      74044.717915                                     7.286643  73.695223   \n",
       "4235   74037.329943                                     7.285817  73.693777   \n",
       "\n",
       "      HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  \n",
       "0            1.444200      3.562189e+06  9365.519531  1.269074e+08  \n",
       "1            1.417410      3.562586e+06  9583.849609  2.914544e+08  \n",
       "4233         1.412947      3.562739e+06  9611.219726  2.726072e+08  \n",
       "4234         1.408483      3.562891e+06  9638.589844  2.537600e+08  \n",
       "2            1.404020      3.563044e+06  9665.959961  2.349128e+08  \n",
       "3            1.390630      3.563562e+06  9652.400391  2.670218e+08  \n",
       "4            1.397320      3.564141e+06  9688.209961  2.024392e+08  \n",
       "5            1.430800      3.564781e+06  9675.410156  1.296232e+08  \n",
       "6            1.439730      3.565481e+06  9721.500000  1.728622e+08  \n",
       "4235         1.436010      3.565735e+06  9759.026693  1.812307e+08  "
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Use linear interpolation to fill up nulls\n",
    "df3 = df2.interpolate(method='linear', axis=0).ffill().bfill()\n",
    "df3.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "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>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4235</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0                              False                      False   \n",
       "1                              False                      False   \n",
       "4233                           False                      False   \n",
       "4234                           False                      False   \n",
       "2                              False                      False   \n",
       "3                              False                      False   \n",
       "4                              False                      False   \n",
       "5                              False                      False   \n",
       "6                              False                      False   \n",
       "4235                           False                      False   \n",
       "\n",
       "      First hand sales amount  Total completions  Total stock  Total take up  \\\n",
       "0                       False               True        False          False   \n",
       "1                       False               True        False          False   \n",
       "4233                    False               True        False          False   \n",
       "4234                    False               True        False          False   \n",
       "2                       False               True        False          False   \n",
       "3                       False               True        False          False   \n",
       "4                       False               True        False          False   \n",
       "5                       False               True        False          False   \n",
       "6                       False               True        False          False   \n",
       "4235                    False               True        False          False   \n",
       "\n",
       "      Total vacancy  Unemployment rate (seasonally adjusted) (%)    CPI  \\\n",
       "0             False                                        False  False   \n",
       "1             False                                        False  False   \n",
       "4233          False                                        False  False   \n",
       "4234          False                                        False  False   \n",
       "2             False                                        False  False   \n",
       "3             False                                        False  False   \n",
       "4             False                                        False  False   \n",
       "5             False                                        False  False   \n",
       "6             False                                        False  False   \n",
       "4235          False                                        False  False   \n",
       "\n",
       "      HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  \n",
       "0               False             False        False         False  \n",
       "1               False             False        False         False  \n",
       "4233            False             False        False         False  \n",
       "4234            False             False        False         False  \n",
       "2               False             False        False         False  \n",
       "3               False             False        False         False  \n",
       "4               False             False        False         False  \n",
       "5               False             False        False         False  \n",
       "6               False             False        False         False  \n",
       "4235            False             False        False         False  "
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Outlier Detection using Inter Quartile Range\n",
    "\n",
    "def out_iqr(s, k=1.5, return_thresholds=False):\n",
    "    \"\"\"\n",
    "    Return a boolean mask of outliers for a series\n",
    "    using interquartile range, works column-wise.\n",
    "    param k:\n",
    "        some cutoff to multiply by the iqr\n",
    "    :type k: ``float``\n",
    "    param return_thresholds:\n",
    "        True returns the lower and upper bounds, good for plotting.\n",
    "        False returns the masked array \n",
    "    :type return_thresholds: ``bool``\n",
    "    \"\"\"\n",
    "    # calculate interquartile range\n",
    "    q25, q75 = np.percentile(s, 25), np.percentile(s, 75)\n",
    "    iqr = q75 - q25\n",
    "    # calculate the outlier cutoff\n",
    "    cut_off = iqr * k\n",
    "    lower, upper = q25 - cut_off, q75 + cut_off\n",
    "    if return_thresholds:\n",
    "        return lower, upper\n",
    "    else: # identify outliers\n",
    "        return [True if x < lower or x > upper else False for x in s]\n",
    "    \n",
    "    \n",
    "# For comparison, make one array each at varying values of k.\n",
    "df4 = df3.drop(columns=['Date'],axis=1)\n",
    "iqr1 = df4.apply(out_iqr, k=1.5)\n",
    "\n",
    "iqr1.head(10)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#### Outlier Detection using Inter Quartile Range\n",
    "\n",
    "def out_std(s, nstd=3.0, return_thresholds=False):\n",
    "    \"\"\"\n",
    "    Return a boolean mask of outliers for a series\n",
    "    using standard deviation, works column-wise.\n",
    "    param nstd:\n",
    "        Set number of standard deviations from the mean\n",
    "        to consider an outlier\n",
    "    :type nstd: ``float``\n",
    "    param return_thresholds:\n",
    "        True returns the lower and upper bounds, good for plotting.\n",
    "        False returns the masked array \n",
    "    :type return_thresholds: ``bool``\n",
    "    \"\"\"\n",
    "    data_mean, data_std = s.mean(), s.std()\n",
    "    cut_off = data_std * nstd\n",
    "    lower, upper = data_mean - cut_off, data_mean + cut_off\n",
    "    if return_thresholds:\n",
    "        return lower, upper\n",
    "    else:\n",
    "        return [True if x < lower or x > upper else False for x in s]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "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>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>64.559769</td>\n",
       "      <td>1854.689208</td>\n",
       "      <td>5.846897e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.444200</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>1.269074e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>64.506848</td>\n",
       "      <td>1892.762372</td>\n",
       "      <td>5.927523e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.417410</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>2.914544e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>64.489181</td>\n",
       "      <td>1905.621203</td>\n",
       "      <td>5.951307e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.092045e+06</td>\n",
       "      <td>19895.379621</td>\n",
       "      <td>74148.223751</td>\n",
       "      <td>7.298216</td>\n",
       "      <td>73.715614</td>\n",
       "      <td>1.412947</td>\n",
       "      <td>3.562739e+06</td>\n",
       "      <td>9611.219726</td>\n",
       "      <td>2.726072e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>64.471514</td>\n",
       "      <td>1918.480035</td>\n",
       "      <td>5.975092e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>19890.752429</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.297389</td>\n",
       "      <td>73.714149</td>\n",
       "      <td>1.408483</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>9638.589844</td>\n",
       "      <td>2.537600e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>64.453848</td>\n",
       "      <td>1931.338866</td>\n",
       "      <td>5.998876e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.404020</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>2.349128e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>64.400768</td>\n",
       "      <td>1970.418692</td>\n",
       "      <td>6.060958e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>19872.438591</td>\n",
       "      <td>74111.243068</td>\n",
       "      <td>7.294081</td>\n",
       "      <td>73.708301</td>\n",
       "      <td>1.390630</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>9652.400391</td>\n",
       "      <td>2.670218e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>64.347610</td>\n",
       "      <td>2010.001848</td>\n",
       "      <td>6.113768e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>19858.946877</td>\n",
       "      <td>74089.062450</td>\n",
       "      <td>7.291601</td>\n",
       "      <td>73.703930</td>\n",
       "      <td>1.397320</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>9688.209961</td>\n",
       "      <td>2.024392e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>64.294373</td>\n",
       "      <td>2050.088335</td>\n",
       "      <td>6.157306e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.089865e+06</td>\n",
       "      <td>19845.650096</td>\n",
       "      <td>74066.887400</td>\n",
       "      <td>7.289122</td>\n",
       "      <td>73.699570</td>\n",
       "      <td>1.430800</td>\n",
       "      <td>3.564781e+06</td>\n",
       "      <td>9675.410156</td>\n",
       "      <td>1.296232e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>64.241057</td>\n",
       "      <td>2090.678153</td>\n",
       "      <td>6.191572e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.089275e+06</td>\n",
       "      <td>19832.548246</td>\n",
       "      <td>74044.717915</td>\n",
       "      <td>7.286643</td>\n",
       "      <td>73.695223</td>\n",
       "      <td>1.439730</td>\n",
       "      <td>3.565481e+06</td>\n",
       "      <td>9721.500000</td>\n",
       "      <td>1.728622e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4235</th>\n",
       "      <td>64.223258</td>\n",
       "      <td>2104.375869</td>\n",
       "      <td>6.199903e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.089080e+06</td>\n",
       "      <td>19828.245940</td>\n",
       "      <td>74037.329943</td>\n",
       "      <td>7.285817</td>\n",
       "      <td>73.693777</td>\n",
       "      <td>1.436010</td>\n",
       "      <td>3.565735e+06</td>\n",
       "      <td>9759.026693</td>\n",
       "      <td>1.812307e+08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0                          64.559769                1854.689208   \n",
       "1                          64.506848                1892.762372   \n",
       "4233                       64.489181                1905.621203   \n",
       "4234                       64.471514                1918.480035   \n",
       "2                          64.453848                1931.338866   \n",
       "3                          64.400768                1970.418692   \n",
       "4                          64.347610                2010.001848   \n",
       "5                          64.294373                2050.088335   \n",
       "6                          64.241057                2090.678153   \n",
       "4235                       64.223258                2104.375869   \n",
       "\n",
       "      First hand sales amount  Total completions   Total stock  Total take up  \\\n",
       "0                5.846897e+09                NaN  1.092844e+06   19914.083323   \n",
       "1                5.927523e+09                NaN  1.092244e+06   19900.006814   \n",
       "4233             5.951307e+09                NaN  1.092045e+06   19895.379621   \n",
       "4234             5.975092e+09                NaN  1.091846e+06   19890.752429   \n",
       "2                5.998876e+09                NaN  1.091646e+06   19886.125236   \n",
       "3                6.060958e+09                NaN  1.091050e+06   19872.438591   \n",
       "4                6.113768e+09                NaN  1.090457e+06   19858.946877   \n",
       "5                6.157306e+09                NaN  1.089865e+06   19845.650096   \n",
       "6                6.191572e+09                NaN  1.089275e+06   19832.548246   \n",
       "4235             6.199903e+09                NaN  1.089080e+06   19828.245940   \n",
       "\n",
       "      Total vacancy  Unemployment rate (seasonally adjusted) (%)        CPI  \\\n",
       "0      74177.818318                                     7.301525  73.721486   \n",
       "1      74155.621001                                     7.299043  73.717079   \n",
       "4233   74148.223751                                     7.298216  73.715614   \n",
       "4234   74140.826501                                     7.297389  73.714149   \n",
       "2      74133.429251                                     7.296561  73.712684   \n",
       "3      74111.243068                                     7.294081  73.708301   \n",
       "4      74089.062450                                     7.291601  73.703930   \n",
       "5      74066.887400                                     7.289122  73.699570   \n",
       "6      74044.717915                                     7.286643  73.695223   \n",
       "4235   74037.329943                                     7.285817  73.693777   \n",
       "\n",
       "      HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  \n",
       "0            1.444200      3.562189e+06  9365.519531  1.269074e+08  \n",
       "1            1.417410      3.562586e+06  9583.849609  2.914544e+08  \n",
       "4233         1.412947      3.562739e+06  9611.219726  2.726072e+08  \n",
       "4234         1.408483      3.562891e+06  9638.589844  2.537600e+08  \n",
       "2            1.404020      3.563044e+06  9665.959961  2.349128e+08  \n",
       "3            1.390630      3.563562e+06  9652.400391  2.670218e+08  \n",
       "4            1.397320      3.564141e+06  9688.209961  2.024392e+08  \n",
       "5            1.430800      3.564781e+06  9675.410156  1.296232e+08  \n",
       "6            1.439730      3.565481e+06  9721.500000  1.728622e+08  \n",
       "4235         1.436010      3.565735e+06  9759.026693  1.812307e+08  "
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for column in df4:\n",
    "    df4[column] = np.where(iqr1[column] == True,'NaN',df4[column])\n",
    "\n",
    "cols = df4.columns\n",
    "df4[cols] = df4[cols].apply(pd.to_numeric, errors='coerce')\n",
    "df4.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "scrolled": true
   },
   "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>Date</th>\n",
       "      <th>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2003-01-02</td>\n",
       "      <td>64.559769</td>\n",
       "      <td>1854.689208</td>\n",
       "      <td>5.846897e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.444200</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>1.269074e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2003-01-03</td>\n",
       "      <td>64.506848</td>\n",
       "      <td>1892.762372</td>\n",
       "      <td>5.927523e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.417410</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>2.914544e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>2003-01-04</td>\n",
       "      <td>64.489181</td>\n",
       "      <td>1905.621203</td>\n",
       "      <td>5.951307e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092045e+06</td>\n",
       "      <td>19895.379621</td>\n",
       "      <td>74148.223751</td>\n",
       "      <td>7.298216</td>\n",
       "      <td>73.715614</td>\n",
       "      <td>1.412947</td>\n",
       "      <td>3.562739e+06</td>\n",
       "      <td>9611.219726</td>\n",
       "      <td>2.726072e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>2003-01-05</td>\n",
       "      <td>64.471514</td>\n",
       "      <td>1918.480035</td>\n",
       "      <td>5.975092e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>19890.752429</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.297389</td>\n",
       "      <td>73.714149</td>\n",
       "      <td>1.408483</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>9638.589844</td>\n",
       "      <td>2.537600e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2003-01-06</td>\n",
       "      <td>64.453848</td>\n",
       "      <td>1931.338866</td>\n",
       "      <td>5.998876e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.404020</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>2.349128e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2003-01-07</td>\n",
       "      <td>64.400768</td>\n",
       "      <td>1970.418692</td>\n",
       "      <td>6.060958e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>19872.438591</td>\n",
       "      <td>74111.243068</td>\n",
       "      <td>7.294081</td>\n",
       "      <td>73.708301</td>\n",
       "      <td>1.390630</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>9652.400391</td>\n",
       "      <td>2.670218e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2003-01-08</td>\n",
       "      <td>64.347610</td>\n",
       "      <td>2010.001848</td>\n",
       "      <td>6.113768e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>19858.946877</td>\n",
       "      <td>74089.062450</td>\n",
       "      <td>7.291601</td>\n",
       "      <td>73.703930</td>\n",
       "      <td>1.397320</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>9688.209961</td>\n",
       "      <td>2.024392e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2003-01-09</td>\n",
       "      <td>64.294373</td>\n",
       "      <td>2050.088335</td>\n",
       "      <td>6.157306e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.089865e+06</td>\n",
       "      <td>19845.650096</td>\n",
       "      <td>74066.887400</td>\n",
       "      <td>7.289122</td>\n",
       "      <td>73.699570</td>\n",
       "      <td>1.430800</td>\n",
       "      <td>3.564781e+06</td>\n",
       "      <td>9675.410156</td>\n",
       "      <td>1.296232e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2003-01-10</td>\n",
       "      <td>64.241057</td>\n",
       "      <td>2090.678153</td>\n",
       "      <td>6.191572e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.089275e+06</td>\n",
       "      <td>19832.548246</td>\n",
       "      <td>74044.717915</td>\n",
       "      <td>7.286643</td>\n",
       "      <td>73.695223</td>\n",
       "      <td>1.439730</td>\n",
       "      <td>3.565481e+06</td>\n",
       "      <td>9721.500000</td>\n",
       "      <td>1.728622e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4235</th>\n",
       "      <td>2003-01-11</td>\n",
       "      <td>64.223258</td>\n",
       "      <td>2104.375869</td>\n",
       "      <td>6.199903e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.089080e+06</td>\n",
       "      <td>19828.245940</td>\n",
       "      <td>74037.329943</td>\n",
       "      <td>7.285817</td>\n",
       "      <td>73.693777</td>\n",
       "      <td>1.436010</td>\n",
       "      <td>3.565735e+06</td>\n",
       "      <td>9759.026693</td>\n",
       "      <td>1.812307e+08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Date  Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0    2003-01-02                       64.559769                1854.689208   \n",
       "1    2003-01-03                       64.506848                1892.762372   \n",
       "4233 2003-01-04                       64.489181                1905.621203   \n",
       "4234 2003-01-05                       64.471514                1918.480035   \n",
       "2    2003-01-06                       64.453848                1931.338866   \n",
       "3    2003-01-07                       64.400768                1970.418692   \n",
       "4    2003-01-08                       64.347610                2010.001848   \n",
       "5    2003-01-09                       64.294373                2050.088335   \n",
       "6    2003-01-10                       64.241057                2090.678153   \n",
       "4235 2003-01-11                       64.223258                2104.375869   \n",
       "\n",
       "      First hand sales amount  Total completions   Total stock  Total take up  \\\n",
       "0                5.846897e+09       32744.302914  1.092844e+06   19914.083323   \n",
       "1                5.927523e+09       32744.302914  1.092244e+06   19900.006814   \n",
       "4233             5.951307e+09       32744.302914  1.092045e+06   19895.379621   \n",
       "4234             5.975092e+09       32744.302914  1.091846e+06   19890.752429   \n",
       "2                5.998876e+09       32744.302914  1.091646e+06   19886.125236   \n",
       "3                6.060958e+09       32744.302914  1.091050e+06   19872.438591   \n",
       "4                6.113768e+09       32744.302914  1.090457e+06   19858.946877   \n",
       "5                6.157306e+09       32744.302914  1.089865e+06   19845.650096   \n",
       "6                6.191572e+09       32744.302914  1.089275e+06   19832.548246   \n",
       "4235             6.199903e+09       32744.302914  1.089080e+06   19828.245940   \n",
       "\n",
       "      Total vacancy  Unemployment rate (seasonally adjusted) (%)        CPI  \\\n",
       "0      74177.818318                                     7.301525  73.721486   \n",
       "1      74155.621001                                     7.299043  73.717079   \n",
       "4233   74148.223751                                     7.298216  73.715614   \n",
       "4234   74140.826501                                     7.297389  73.714149   \n",
       "2      74133.429251                                     7.296561  73.712684   \n",
       "3      74111.243068                                     7.294081  73.708301   \n",
       "4      74089.062450                                     7.291601  73.703930   \n",
       "5      74066.887400                                     7.289122  73.699570   \n",
       "6      74044.717915                                     7.286643  73.695223   \n",
       "4235   74037.329943                                     7.285817  73.693777   \n",
       "\n",
       "      HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  \n",
       "0            1.444200      3.562189e+06  9365.519531  1.269074e+08  \n",
       "1            1.417410      3.562586e+06  9583.849609  2.914544e+08  \n",
       "4233         1.412947      3.562739e+06  9611.219726  2.726072e+08  \n",
       "4234         1.408483      3.562891e+06  9638.589844  2.537600e+08  \n",
       "2            1.404020      3.563044e+06  9665.959961  2.349128e+08  \n",
       "3            1.390630      3.563562e+06  9652.400391  2.670218e+08  \n",
       "4            1.397320      3.564141e+06  9688.209961  2.024392e+08  \n",
       "5            1.430800      3.564781e+06  9675.410156  1.296232e+08  \n",
       "6            1.439730      3.565481e+06  9721.500000  1.728622e+08  \n",
       "4235         1.436010      3.565735e+06  9759.026693  1.812307e+08  "
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Use linear interpolation to fill up nulls\n",
    "df = df4.interpolate(method='linear', axis=0).bfill().ffill()\n",
    "df3['Date'] = pd.to_datetime(df3['Date'])\n",
    "df = pd.concat([df3['Date'],df], axis=1)\n",
    "df.head(10) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel_launcher.py:3: FutureWarning: the 'freq'' keyword is deprecated, use 'period' instead\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_series = df.set_index('Date').asfreq('D')\n",
    "series = pd.Series(df_series['Private Domestic (Price Index)'], index= df_series.index)\n",
    "results = seasonal_decompose(series, model='additive',freq = 365)\n",
    "pyplot.rcParams['figure.figsize'] = (20.0, 10.0)\n",
    "results.plot()\n",
    "pyplot.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ACF Plot of Target Variable\n",
    "plot_acf(series,lags= 100)\n",
    "pyplot.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_pacf(series, lags=50)\n",
    "pyplot.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Feature Engineering\n",
    "Once the data has been cleaned up, we will create the following additional features:\n",
    "\n",
    "1. 3 lags for each feature\n",
    "2. Daily change for each feature\n",
    "3. A flag for holidays\n",
    "4. Calculate First hand sales price and drop the sales amount variable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel_launcher.py:6: FutureWarning: The pandas.np module is deprecated and will be removed from pandas in a future version. Import numpy directly instead\n",
      "  \n"
     ]
    },
    {
     "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>Date</th>\n",
       "      <th>Private Domestic (Price Index)</th>\n",
       "      <th>First hand sales quantity</th>\n",
       "      <th>First hand sales amount</th>\n",
       "      <th>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "      <th>holidays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2003-01-02</td>\n",
       "      <td>64.559769</td>\n",
       "      <td>1854.689208</td>\n",
       "      <td>5.846897e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.444200</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>126907400.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2003-01-03</td>\n",
       "      <td>64.506848</td>\n",
       "      <td>1892.762372</td>\n",
       "      <td>5.927523e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.417410</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>291454400.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>2003-01-04</td>\n",
       "      <td>64.489181</td>\n",
       "      <td>1905.621203</td>\n",
       "      <td>5.951307e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092045e+06</td>\n",
       "      <td>19895.379621</td>\n",
       "      <td>74148.223751</td>\n",
       "      <td>7.298216</td>\n",
       "      <td>73.715614</td>\n",
       "      <td>1.412947</td>\n",
       "      <td>3.562739e+06</td>\n",
       "      <td>9611.219726</td>\n",
       "      <td>272607200.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>2003-01-05</td>\n",
       "      <td>64.471514</td>\n",
       "      <td>1918.480035</td>\n",
       "      <td>5.975092e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>19890.752429</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.297389</td>\n",
       "      <td>73.714149</td>\n",
       "      <td>1.408483</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>9638.589844</td>\n",
       "      <td>253760000.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2003-01-06</td>\n",
       "      <td>64.453848</td>\n",
       "      <td>1931.338866</td>\n",
       "      <td>5.998876e+09</td>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.404020</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>234912800.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Date  Private Domestic (Price Index)  First hand sales quantity  \\\n",
       "0    2003-01-02                       64.559769                1854.689208   \n",
       "1    2003-01-03                       64.506848                1892.762372   \n",
       "4233 2003-01-04                       64.489181                1905.621203   \n",
       "4234 2003-01-05                       64.471514                1918.480035   \n",
       "2    2003-01-06                       64.453848                1931.338866   \n",
       "\n",
       "      First hand sales amount  Total completions   Total stock  Total take up  \\\n",
       "0                5.846897e+09       32744.302914  1.092844e+06   19914.083323   \n",
       "1                5.927523e+09       32744.302914  1.092244e+06   19900.006814   \n",
       "4233             5.951307e+09       32744.302914  1.092045e+06   19895.379621   \n",
       "4234             5.975092e+09       32744.302914  1.091846e+06   19890.752429   \n",
       "2                5.998876e+09       32744.302914  1.091646e+06   19886.125236   \n",
       "\n",
       "      Total vacancy  Unemployment rate (seasonally adjusted) (%)        CPI  \\\n",
       "0      74177.818318                                     7.301525  73.721486   \n",
       "1      74155.621001                                     7.299043  73.717079   \n",
       "4233   74148.223751                                     7.298216  73.715614   \n",
       "4234   74140.826501                                     7.297389  73.714149   \n",
       "2      74133.429251                                     7.296561  73.712684   \n",
       "\n",
       "      HIBOR (1-month)  M3 (HK$ million)  HSI - close  HSI - volume  holidays  \n",
       "0            1.444200      3.562189e+06  9365.519531   126907400.0         0  \n",
       "1            1.417410      3.562586e+06  9583.849609   291454400.0         0  \n",
       "4233         1.412947      3.562739e+06  9611.219726   272607200.0         0  \n",
       "4234         1.408483      3.562891e+06  9638.589844   253760000.0         0  \n",
       "2            1.404020      3.563044e+06  9665.959961   234912800.0         0  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create Holiday Flag\n",
    "\n",
    "df['Date'] = pd.to_datetime(df['Date'])\n",
    "cal = calendar()\n",
    "holidays = cal.holidays(start=df['Date'].min(), end=df['Date'].max())\n",
    "df['holidays'] = pd.np.where(df['Date'].dt.date.isin(holidays), 1, 0)\n",
    "df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "scrolled": true
   },
   "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>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "      <th>First hand sales price</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092844e+06</td>\n",
       "      <td>19914.083323</td>\n",
       "      <td>74177.818318</td>\n",
       "      <td>7.301525</td>\n",
       "      <td>73.721486</td>\n",
       "      <td>1.444200</td>\n",
       "      <td>3.562189e+06</td>\n",
       "      <td>9365.519531</td>\n",
       "      <td>126907400.0</td>\n",
       "      <td>3.152494e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092244e+06</td>\n",
       "      <td>19900.006814</td>\n",
       "      <td>74155.621001</td>\n",
       "      <td>7.299043</td>\n",
       "      <td>73.717079</td>\n",
       "      <td>1.417410</td>\n",
       "      <td>3.562586e+06</td>\n",
       "      <td>9583.849609</td>\n",
       "      <td>291454400.0</td>\n",
       "      <td>3.131678e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4233</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.092045e+06</td>\n",
       "      <td>19895.379621</td>\n",
       "      <td>74148.223751</td>\n",
       "      <td>7.298216</td>\n",
       "      <td>73.715614</td>\n",
       "      <td>1.412947</td>\n",
       "      <td>3.562739e+06</td>\n",
       "      <td>9611.219726</td>\n",
       "      <td>272607200.0</td>\n",
       "      <td>3.123027e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>19890.752429</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.297389</td>\n",
       "      <td>73.714149</td>\n",
       "      <td>1.408483</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>9638.589844</td>\n",
       "      <td>253760000.0</td>\n",
       "      <td>3.114493e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.404020</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>234912800.0</td>\n",
       "      <td>3.106071e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Total completions   Total stock  Total take up  Total vacancy  \\\n",
       "0          32744.302914  1.092844e+06   19914.083323   74177.818318   \n",
       "1          32744.302914  1.092244e+06   19900.006814   74155.621001   \n",
       "4233       32744.302914  1.092045e+06   19895.379621   74148.223751   \n",
       "4234       32744.302914  1.091846e+06   19890.752429   74140.826501   \n",
       "2          32744.302914  1.091646e+06   19886.125236   74133.429251   \n",
       "\n",
       "      Unemployment rate (seasonally adjusted) (%)        CPI  HIBOR (1-month)  \\\n",
       "0                                        7.301525  73.721486         1.444200   \n",
       "1                                        7.299043  73.717079         1.417410   \n",
       "4233                                     7.298216  73.715614         1.412947   \n",
       "4234                                     7.297389  73.714149         1.408483   \n",
       "2                                        7.296561  73.712684         1.404020   \n",
       "\n",
       "      M3 (HK$ million)  HSI - close  HSI - volume  First hand sales price  \n",
       "0         3.562189e+06  9365.519531   126907400.0            3.152494e+06  \n",
       "1         3.562586e+06  9583.849609   291454400.0            3.131678e+06  \n",
       "4233      3.562739e+06  9611.219726   272607200.0            3.123027e+06  \n",
       "4234      3.562891e+06  9638.589844   253760000.0            3.114493e+06  \n",
       "2         3.563044e+06  9665.959961   234912800.0            3.106071e+06  "
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Create a new feature =  First hand price\n",
    "df['First hand sales price'] = df['First hand sales amount']/df['First hand sales quantity']\n",
    "predictors1 = df.drop(columns=['Date','Private Domestic (Price Index)','holidays','First hand sales amount','First hand sales quantity'],axis=1) # Add  if applicable\n",
    "predictors1.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check Optimal Number of Lags\n",
    "for i in range(3,12):\n",
    "    pyplot.figure(i, figsize=(15,2)) # add this statement before your plot\n",
    "    pyplot.xcorr(df.iloc[:,2],df.iloc[:,i],maxlags=45, usevlines=1)\n",
    "    pyplot.title('Cross Correlation Between Private Domestic (Price Index) and '+ df.columns[i])\n",
    "    pyplot.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "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>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "      <th>...</th>\n",
       "      <th>Total take up_Diff</th>\n",
       "      <th>Total vacancy_Diff</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Diff</th>\n",
       "      <th>CPI_Diff</th>\n",
       "      <th>HIBOR (1-month)_Diff</th>\n",
       "      <th>M3 (HK$ million)_Diff</th>\n",
       "      <th>HSI - close_Diff</th>\n",
       "      <th>HSI - volume_Diff</th>\n",
       "      <th>First hand sales price_Diff</th>\n",
       "      <th>holidays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>19890.752429</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.297389</td>\n",
       "      <td>73.714149</td>\n",
       "      <td>1.408483</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>9638.589844</td>\n",
       "      <td>2.537600e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-4.627192</td>\n",
       "      <td>-7.397250</td>\n",
       "      <td>-0.000827</td>\n",
       "      <td>-0.001465</td>\n",
       "      <td>-0.004463</td>\n",
       "      <td>152.611911</td>\n",
       "      <td>27.370117</td>\n",
       "      <td>-1.884720e+07</td>\n",
       "      <td>-8534.865986</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>19886.125236</td>\n",
       "      <td>74133.429251</td>\n",
       "      <td>7.296561</td>\n",
       "      <td>73.712684</td>\n",
       "      <td>1.404020</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>9665.959961</td>\n",
       "      <td>2.349128e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-4.627192</td>\n",
       "      <td>-7.397250</td>\n",
       "      <td>-0.000827</td>\n",
       "      <td>-0.001465</td>\n",
       "      <td>-0.004463</td>\n",
       "      <td>152.611911</td>\n",
       "      <td>27.370117</td>\n",
       "      <td>-1.884720e+07</td>\n",
       "      <td>-8421.215910</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>19872.438591</td>\n",
       "      <td>74111.243068</td>\n",
       "      <td>7.294081</td>\n",
       "      <td>73.708301</td>\n",
       "      <td>1.390630</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>9652.400391</td>\n",
       "      <td>2.670218e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-13.686645</td>\n",
       "      <td>-22.186184</td>\n",
       "      <td>-0.002481</td>\n",
       "      <td>-0.004383</td>\n",
       "      <td>-0.013390</td>\n",
       "      <td>518.438226</td>\n",
       "      <td>-13.559570</td>\n",
       "      <td>3.210900e+07</td>\n",
       "      <td>-30096.688654</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>19858.946877</td>\n",
       "      <td>74089.062450</td>\n",
       "      <td>7.291601</td>\n",
       "      <td>73.703930</td>\n",
       "      <td>1.397320</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>9688.209961</td>\n",
       "      <td>2.024392e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-13.491713</td>\n",
       "      <td>-22.180617</td>\n",
       "      <td>-0.002480</td>\n",
       "      <td>-0.004371</td>\n",
       "      <td>0.006690</td>\n",
       "      <td>579.040720</td>\n",
       "      <td>35.809570</td>\n",
       "      <td>-6.458260e+07</td>\n",
       "      <td>-34301.973468</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.089865e+06</td>\n",
       "      <td>19845.650096</td>\n",
       "      <td>74066.887400</td>\n",
       "      <td>7.289122</td>\n",
       "      <td>73.699570</td>\n",
       "      <td>1.430800</td>\n",
       "      <td>3.564781e+06</td>\n",
       "      <td>9675.410156</td>\n",
       "      <td>1.296232e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-13.296781</td>\n",
       "      <td>-22.175051</td>\n",
       "      <td>-0.002479</td>\n",
       "      <td>-0.004360</td>\n",
       "      <td>0.033480</td>\n",
       "      <td>639.643214</td>\n",
       "      <td>-12.799805</td>\n",
       "      <td>-7.281600e+07</td>\n",
       "      <td>-38238.405936</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.089275e+06</td>\n",
       "      <td>19832.548246</td>\n",
       "      <td>74044.717915</td>\n",
       "      <td>7.286643</td>\n",
       "      <td>73.695223</td>\n",
       "      <td>1.439730</td>\n",
       "      <td>3.565481e+06</td>\n",
       "      <td>9721.500000</td>\n",
       "      <td>1.728622e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-13.101850</td>\n",
       "      <td>-22.169484</td>\n",
       "      <td>-0.002478</td>\n",
       "      <td>-0.004348</td>\n",
       "      <td>0.008930</td>\n",
       "      <td>700.245708</td>\n",
       "      <td>46.089844</td>\n",
       "      <td>4.323900e+07</td>\n",
       "      <td>-41920.790237</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4235</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.089080e+06</td>\n",
       "      <td>19828.245940</td>\n",
       "      <td>74037.329943</td>\n",
       "      <td>7.285817</td>\n",
       "      <td>73.693777</td>\n",
       "      <td>1.436010</td>\n",
       "      <td>3.565735e+06</td>\n",
       "      <td>9759.026693</td>\n",
       "      <td>1.812307e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-4.302306</td>\n",
       "      <td>-7.387973</td>\n",
       "      <td>-0.000826</td>\n",
       "      <td>-0.001445</td>\n",
       "      <td>-0.003720</td>\n",
       "      <td>253.616067</td>\n",
       "      <td>37.526693</td>\n",
       "      <td>8.368467e+06</td>\n",
       "      <td>-15317.896731</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4236</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.088884e+06</td>\n",
       "      <td>19823.943634</td>\n",
       "      <td>74029.941970</td>\n",
       "      <td>7.284991</td>\n",
       "      <td>73.692332</td>\n",
       "      <td>1.432290</td>\n",
       "      <td>3.565988e+06</td>\n",
       "      <td>9796.553385</td>\n",
       "      <td>1.895991e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-4.302306</td>\n",
       "      <td>-7.387973</td>\n",
       "      <td>-0.000826</td>\n",
       "      <td>-0.001445</td>\n",
       "      <td>-0.003720</td>\n",
       "      <td>253.616067</td>\n",
       "      <td>37.526693</td>\n",
       "      <td>8.368467e+06</td>\n",
       "      <td>-15119.773112</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.088688e+06</td>\n",
       "      <td>19819.641329</td>\n",
       "      <td>74022.553998</td>\n",
       "      <td>7.284165</td>\n",
       "      <td>73.690887</td>\n",
       "      <td>1.428570</td>\n",
       "      <td>3.566242e+06</td>\n",
       "      <td>9834.080078</td>\n",
       "      <td>1.979676e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-4.302306</td>\n",
       "      <td>-7.387973</td>\n",
       "      <td>-0.000826</td>\n",
       "      <td>-0.001445</td>\n",
       "      <td>-0.003720</td>\n",
       "      <td>253.616067</td>\n",
       "      <td>37.526693</td>\n",
       "      <td>8.368467e+06</td>\n",
       "      <td>-14925.468628</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.088102e+06</td>\n",
       "      <td>19806.929343</td>\n",
       "      <td>74000.395646</td>\n",
       "      <td>7.281688</td>\n",
       "      <td>73.686562</td>\n",
       "      <td>1.433040</td>\n",
       "      <td>3.567063e+06</td>\n",
       "      <td>9796.309570</td>\n",
       "      <td>1.681520e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-12.711986</td>\n",
       "      <td>-22.158351</td>\n",
       "      <td>-0.002477</td>\n",
       "      <td>-0.004324</td>\n",
       "      <td>0.004470</td>\n",
       "      <td>821.450696</td>\n",
       "      <td>-37.770508</td>\n",
       "      <td>-2.981560e+07</td>\n",
       "      <td>-48578.710184</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 56 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Total completions   Total stock  Total take up  Total vacancy  \\\n",
       "4234       32744.302914  1.091846e+06   19890.752429   74140.826501   \n",
       "2          32744.302914  1.091646e+06   19886.125236   74133.429251   \n",
       "3          32744.302914  1.091050e+06   19872.438591   74111.243068   \n",
       "4          32744.302914  1.090457e+06   19858.946877   74089.062450   \n",
       "5          32744.302914  1.089865e+06   19845.650096   74066.887400   \n",
       "6          32744.302914  1.089275e+06   19832.548246   74044.717915   \n",
       "4235       32744.302914  1.089080e+06   19828.245940   74037.329943   \n",
       "4236       32744.302914  1.088884e+06   19823.943634   74029.941970   \n",
       "7          32744.302914  1.088688e+06   19819.641329   74022.553998   \n",
       "8          32744.302914  1.088102e+06   19806.929343   74000.395646   \n",
       "\n",
       "      Unemployment rate (seasonally adjusted) (%)        CPI  HIBOR (1-month)  \\\n",
       "4234                                     7.297389  73.714149         1.408483   \n",
       "2                                        7.296561  73.712684         1.404020   \n",
       "3                                        7.294081  73.708301         1.390630   \n",
       "4                                        7.291601  73.703930         1.397320   \n",
       "5                                        7.289122  73.699570         1.430800   \n",
       "6                                        7.286643  73.695223         1.439730   \n",
       "4235                                     7.285817  73.693777         1.436010   \n",
       "4236                                     7.284991  73.692332         1.432290   \n",
       "7                                        7.284165  73.690887         1.428570   \n",
       "8                                        7.281688  73.686562         1.433040   \n",
       "\n",
       "      M3 (HK$ million)  HSI - close  HSI - volume  ...  Total take up_Diff  \\\n",
       "4234      3.562891e+06  9638.589844  2.537600e+08  ...           -4.627192   \n",
       "2         3.563044e+06  9665.959961  2.349128e+08  ...           -4.627192   \n",
       "3         3.563562e+06  9652.400391  2.670218e+08  ...          -13.686645   \n",
       "4         3.564141e+06  9688.209961  2.024392e+08  ...          -13.491713   \n",
       "5         3.564781e+06  9675.410156  1.296232e+08  ...          -13.296781   \n",
       "6         3.565481e+06  9721.500000  1.728622e+08  ...          -13.101850   \n",
       "4235      3.565735e+06  9759.026693  1.812307e+08  ...           -4.302306   \n",
       "4236      3.565988e+06  9796.553385  1.895991e+08  ...           -4.302306   \n",
       "7         3.566242e+06  9834.080078  1.979676e+08  ...           -4.302306   \n",
       "8         3.567063e+06  9796.309570  1.681520e+08  ...          -12.711986   \n",
       "\n",
       "      Total vacancy_Diff  Unemployment rate (seasonally adjusted) (%)_Diff  \\\n",
       "4234           -7.397250                                         -0.000827   \n",
       "2              -7.397250                                         -0.000827   \n",
       "3             -22.186184                                         -0.002481   \n",
       "4             -22.180617                                         -0.002480   \n",
       "5             -22.175051                                         -0.002479   \n",
       "6             -22.169484                                         -0.002478   \n",
       "4235           -7.387973                                         -0.000826   \n",
       "4236           -7.387973                                         -0.000826   \n",
       "7              -7.387973                                         -0.000826   \n",
       "8             -22.158351                                         -0.002477   \n",
       "\n",
       "      CPI_Diff  HIBOR (1-month)_Diff  M3 (HK$ million)_Diff  HSI - close_Diff  \\\n",
       "4234 -0.001465             -0.004463             152.611911         27.370117   \n",
       "2    -0.001465             -0.004463             152.611911         27.370117   \n",
       "3    -0.004383             -0.013390             518.438226        -13.559570   \n",
       "4    -0.004371              0.006690             579.040720         35.809570   \n",
       "5    -0.004360              0.033480             639.643214        -12.799805   \n",
       "6    -0.004348              0.008930             700.245708         46.089844   \n",
       "4235 -0.001445             -0.003720             253.616067         37.526693   \n",
       "4236 -0.001445             -0.003720             253.616067         37.526693   \n",
       "7    -0.001445             -0.003720             253.616067         37.526693   \n",
       "8    -0.004324              0.004470             821.450696        -37.770508   \n",
       "\n",
       "      HSI - volume_Diff  First hand sales price_Diff  holidays  \n",
       "4234      -1.884720e+07                 -8534.865986         0  \n",
       "2         -1.884720e+07                 -8421.215910         0  \n",
       "3          3.210900e+07                -30096.688654         0  \n",
       "4         -6.458260e+07                -34301.973468         0  \n",
       "5         -7.281600e+07                -38238.405936         0  \n",
       "6          4.323900e+07                -41920.790237         0  \n",
       "4235       8.368467e+06                -15317.896731         0  \n",
       "4236       8.368467e+06                -15119.773112         0  \n",
       "7          8.368467e+06                -14925.468628         0  \n",
       "8         -2.981560e+07                -48578.710184         0  \n",
       "\n",
       "[10 rows x 56 columns]"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Engineering features based on creating lags and perdiodic changes of features\n",
    "\n",
    "#Add Lags\n",
    "predictors2 = predictors1.shift(1)\n",
    "predictors2 = predictors2.add_suffix('_Lag1')\n",
    "predictors3 = predictors1.shift(2)\n",
    "predictors3 = predictors3.add_suffix('_Lag2')\n",
    "predictors4 = predictors1.shift(3)\n",
    "predictors4 = predictors4.add_suffix('_Lag3')\n",
    "\n",
    "#Add Periodic Difference\n",
    "predictors5 = predictors1.diff()\n",
    "predictors5 = predictors5.add_suffix('_Diff')\n",
    "\n",
    "predictors = pd.concat([predictors1, predictors2, predictors3,predictors4, predictors5,df['holidays']], axis=1) #Add df['holidays'] if applicable\n",
    "predictors=predictors.dropna(axis=0, how='any')\n",
    "\n",
    "predictors.head(10)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#### Additional features: Stock Price\n",
    "#!pip install --trusted-host=pypi.org --trusted-host=files.pythonhosted.org --user yfinance\n",
    "#!pip install --trusted-host=pypi.org --trusted-host=files.pythonhosted.org --user yahoofinancials\n",
    "\n",
    "import yfinance as yf\n",
    "from yahoofinancials import YahooFinancials\n",
    "\n",
    "yahoo_financials = YahooFinancials('CRM')\n",
    "\n",
    "stock_price = yahoo_financials.get_historical_price_data(start_date='2018-08-01', \n",
    "                                                  end_date='2020-08-01', \n",
    "                                                  time_interval='monthly')\n",
    "\n",
    "stock_price_df = pd.DataFrame(stock_price['CRM']['prices'])\n",
    "stock_price_df = stock_price_df.drop('date', axis=1).set_index('formatted_date')\n",
    "stock_price_df = pd.DataFrame(stock_price_df['close'])\n",
    "stock_price_df.columns = ['Stock_Price']\n",
    "\n",
    "stock_price_df "
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "# Additional features:  Stock Price Index from multiple companies               \n",
    "\n",
    "scaler = StandardScaler()\n",
    "# transform x data\n",
    "scaled_stock = scaler.fit_transform(stock_price_df_c)\n",
    "column_name = competitor_stock[i]\n",
    "stock_price_df_comp[column_name] = scaled_stock[:,0]\n",
    "                    \n",
    "col = stock_price_df_comp.loc[: , \"ADBE\":\"ORCL\"]\n",
    "stock_price_df_comp['Competitor_Stock_Index'] = col.mean(axis=1)\n",
    "stock_price_df_comp = pd.DataFrame(stock_price_df_comp['Competitor_Stock_Index'])\n",
    "\n",
    "stock_price_df_comp"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#Additional features:  S&P 500 \n",
    "\n",
    "import datetime\n",
    "\n",
    "pd.core.common.is_list_like = pd.api.types.is_list_like\n",
    "\n",
    "import pandas_datareader.data as web\n",
    "\n",
    "\n",
    "start = datetime.datetime(2018, 8, 1)\n",
    "end = datetime.datetime(2020, 8, 1)\n",
    "\n",
    "SP500 = web.DataReader(['sp500'], 'fred', start, end)\n",
    "\n",
    "\n",
    "SP500_df = SP500.resample('MS').mean()\n",
    "\n",
    "SP500_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "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>Total completions</th>\n",
       "      <th>Total stock</th>\n",
       "      <th>Total take up</th>\n",
       "      <th>Total vacancy</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <th>CPI</th>\n",
       "      <th>HIBOR (1-month)</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>HSI - close</th>\n",
       "      <th>HSI - volume</th>\n",
       "      <th>...</th>\n",
       "      <th>Total take up_Diff</th>\n",
       "      <th>Total vacancy_Diff</th>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Diff</th>\n",
       "      <th>CPI_Diff</th>\n",
       "      <th>HIBOR (1-month)_Diff</th>\n",
       "      <th>M3 (HK$ million)_Diff</th>\n",
       "      <th>HSI - close_Diff</th>\n",
       "      <th>HSI - volume_Diff</th>\n",
       "      <th>First hand sales price_Diff</th>\n",
       "      <th>holidays</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6.170000e+03</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6.170000e+03</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6.170000e+03</td>\n",
       "      <td>...</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6170.000000</td>\n",
       "      <td>6.170000e+03</td>\n",
       "      <td>6.170000e+03</td>\n",
       "      <td>6170.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>15361.946209</td>\n",
       "      <td>1.106765e+06</td>\n",
       "      <td>13925.085988</td>\n",
       "      <td>52796.789304</td>\n",
       "      <td>4.197309</td>\n",
       "      <td>88.223190</td>\n",
       "      <td>1.103977</td>\n",
       "      <td>8.386498e+06</td>\n",
       "      <td>20802.897798</td>\n",
       "      <td>1.453729e+09</td>\n",
       "      <td>...</td>\n",
       "      <td>-3.493241</td>\n",
       "      <td>-0.848153</td>\n",
       "      <td>-0.000655</td>\n",
       "      <td>0.005887</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>1731.692209</td>\n",
       "      <td>2.804327</td>\n",
       "      <td>3.474530e+05</td>\n",
       "      <td>8.058516e+02</td>\n",
       "      <td>0.000162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6386.221684</td>\n",
       "      <td>5.070278e+04</td>\n",
       "      <td>6173.995838</td>\n",
       "      <td>8987.318195</td>\n",
       "      <td>1.267338</td>\n",
       "      <td>12.709861</td>\n",
       "      <td>1.257734</td>\n",
       "      <td>3.551502e+06</td>\n",
       "      <td>5295.961619</td>\n",
       "      <td>8.001701e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>23.476645</td>\n",
       "      <td>19.673321</td>\n",
       "      <td>0.118831</td>\n",
       "      <td>0.226253</td>\n",
       "      <td>0.113443</td>\n",
       "      <td>4928.887779</td>\n",
       "      <td>209.760663</td>\n",
       "      <td>3.312027e+08</td>\n",
       "      <td>4.143880e+05</td>\n",
       "      <td>0.012731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>6792.576078</td>\n",
       "      <td>1.006630e+06</td>\n",
       "      <td>-1657.914651</td>\n",
       "      <td>41975.928483</td>\n",
       "      <td>2.800000</td>\n",
       "      <td>71.700000</td>\n",
       "      <td>0.040180</td>\n",
       "      <td>3.548561e+06</td>\n",
       "      <td>8409.009766</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>...</td>\n",
       "      <td>-86.777479</td>\n",
       "      <td>-58.464668</td>\n",
       "      <td>-1.391505</td>\n",
       "      <td>-3.448806</td>\n",
       "      <td>-2.151193</td>\n",
       "      <td>-27648.009925</td>\n",
       "      <td>-2061.228516</td>\n",
       "      <td>-1.711519e+09</td>\n",
       "      <td>-1.656848e+07</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>9943.585825</td>\n",
       "      <td>1.073241e+06</td>\n",
       "      <td>9518.075367</td>\n",
       "      <td>45722.582067</td>\n",
       "      <td>3.136487</td>\n",
       "      <td>75.570224</td>\n",
       "      <td>0.213570</td>\n",
       "      <td>5.290837e+06</td>\n",
       "      <td>17083.933106</td>\n",
       "      <td>9.291716e+08</td>\n",
       "      <td>...</td>\n",
       "      <td>-12.697966</td>\n",
       "      <td>-9.006723</td>\n",
       "      <td>-0.001276</td>\n",
       "      <td>0.002268</td>\n",
       "      <td>-0.003965</td>\n",
       "      <td>-376.322723</td>\n",
       "      <td>-71.925781</td>\n",
       "      <td>-1.312109e+08</td>\n",
       "      <td>-4.417888e+04</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>13761.512667</td>\n",
       "      <td>1.106819e+06</td>\n",
       "      <td>12662.876161</td>\n",
       "      <td>49958.656160</td>\n",
       "      <td>3.643848</td>\n",
       "      <td>86.249741</td>\n",
       "      <td>0.347573</td>\n",
       "      <td>7.668805e+06</td>\n",
       "      <td>21655.020508</td>\n",
       "      <td>1.497364e+09</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.106193</td>\n",
       "      <td>-2.379272</td>\n",
       "      <td>-0.000483</td>\n",
       "      <td>0.004488</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1122.499073</td>\n",
       "      <td>7.116211</td>\n",
       "      <td>-9.377533e+06</td>\n",
       "      <td>-4.408809e+02</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>19061.267131</td>\n",
       "      <td>1.142407e+06</td>\n",
       "      <td>16988.427021</td>\n",
       "      <td>62792.720675</td>\n",
       "      <td>5.054313</td>\n",
       "      <td>100.307746</td>\n",
       "      <td>1.747771</td>\n",
       "      <td>1.147976e+07</td>\n",
       "      <td>23949.367350</td>\n",
       "      <td>1.940716e+09</td>\n",
       "      <td>...</td>\n",
       "      <td>8.903788</td>\n",
       "      <td>2.886277</td>\n",
       "      <td>-0.000048</td>\n",
       "      <td>0.011017</td>\n",
       "      <td>0.002463</td>\n",
       "      <td>3529.909155</td>\n",
       "      <td>87.710286</td>\n",
       "      <td>1.151925e+08</td>\n",
       "      <td>4.535706e+04</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>32744.302914</td>\n",
       "      <td>1.214837e+06</td>\n",
       "      <td>28187.536217</td>\n",
       "      <td>74140.826501</td>\n",
       "      <td>7.800000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>4.047737</td>\n",
       "      <td>1.470108e+07</td>\n",
       "      <td>33154.121094</td>\n",
       "      <td>3.467110e+09</td>\n",
       "      <td>...</td>\n",
       "      <td>58.243255</td>\n",
       "      <td>97.903823</td>\n",
       "      <td>1.359147</td>\n",
       "      <td>3.400005</td>\n",
       "      <td>1.991575</td>\n",
       "      <td>34181.322349</td>\n",
       "      <td>2332.539063</td>\n",
       "      <td>1.984283e+09</td>\n",
       "      <td>2.214944e+07</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 56 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Total completions   Total stock  Total take up  Total vacancy  \\\n",
       "count        6170.000000  6.170000e+03    6170.000000    6170.000000   \n",
       "mean        15361.946209  1.106765e+06   13925.085988   52796.789304   \n",
       "std          6386.221684  5.070278e+04    6173.995838    8987.318195   \n",
       "min          6792.576078  1.006630e+06   -1657.914651   41975.928483   \n",
       "25%          9943.585825  1.073241e+06    9518.075367   45722.582067   \n",
       "50%         13761.512667  1.106819e+06   12662.876161   49958.656160   \n",
       "75%         19061.267131  1.142407e+06   16988.427021   62792.720675   \n",
       "max         32744.302914  1.214837e+06   28187.536217   74140.826501   \n",
       "\n",
       "       Unemployment rate (seasonally adjusted) (%)          CPI  \\\n",
       "count                                  6170.000000  6170.000000   \n",
       "mean                                      4.197309    88.223190   \n",
       "std                                       1.267338    12.709861   \n",
       "min                                       2.800000    71.700000   \n",
       "25%                                       3.136487    75.570224   \n",
       "50%                                       3.643848    86.249741   \n",
       "75%                                       5.054313   100.307746   \n",
       "max                                       7.800000   111.000000   \n",
       "\n",
       "       HIBOR (1-month)  M3 (HK$ million)   HSI - close  HSI - volume  ...  \\\n",
       "count      6170.000000      6.170000e+03   6170.000000  6.170000e+03  ...   \n",
       "mean          1.103977      8.386498e+06  20802.897798  1.453729e+09  ...   \n",
       "std           1.257734      3.551502e+06   5295.961619  8.001701e+08  ...   \n",
       "min           0.040180      3.548561e+06   8409.009766  0.000000e+00  ...   \n",
       "25%           0.213570      5.290837e+06  17083.933106  9.291716e+08  ...   \n",
       "50%           0.347573      7.668805e+06  21655.020508  1.497364e+09  ...   \n",
       "75%           1.747771      1.147976e+07  23949.367350  1.940716e+09  ...   \n",
       "max           4.047737      1.470108e+07  33154.121094  3.467110e+09  ...   \n",
       "\n",
       "       Total take up_Diff  Total vacancy_Diff  \\\n",
       "count         6170.000000         6170.000000   \n",
       "mean            -3.493241           -0.848153   \n",
       "std             23.476645           19.673321   \n",
       "min            -86.777479          -58.464668   \n",
       "25%            -12.697966           -9.006723   \n",
       "50%             -0.106193           -2.379272   \n",
       "75%              8.903788            2.886277   \n",
       "max             58.243255           97.903823   \n",
       "\n",
       "       Unemployment rate (seasonally adjusted) (%)_Diff     CPI_Diff  \\\n",
       "count                                       6170.000000  6170.000000   \n",
       "mean                                          -0.000655     0.005887   \n",
       "std                                            0.118831     0.226253   \n",
       "min                                           -1.391505    -3.448806   \n",
       "25%                                           -0.001276     0.002268   \n",
       "50%                                           -0.000483     0.004488   \n",
       "75%                                           -0.000048     0.011017   \n",
       "max                                            1.359147     3.400005   \n",
       "\n",
       "       HIBOR (1-month)_Diff  M3 (HK$ million)_Diff  HSI - close_Diff  \\\n",
       "count           6170.000000            6170.000000       6170.000000   \n",
       "mean               0.000106            1731.692209          2.804327   \n",
       "std                0.113443            4928.887779        209.760663   \n",
       "min               -2.151193          -27648.009925      -2061.228516   \n",
       "25%               -0.003965            -376.322723        -71.925781   \n",
       "50%                0.000000            1122.499073          7.116211   \n",
       "75%                0.002463            3529.909155         87.710286   \n",
       "max                1.991575           34181.322349       2332.539063   \n",
       "\n",
       "       HSI - volume_Diff  First hand sales price_Diff     holidays  \n",
       "count       6.170000e+03                 6.170000e+03  6170.000000  \n",
       "mean        3.474530e+05                 8.058516e+02     0.000162  \n",
       "std         3.312027e+08                 4.143880e+05     0.012731  \n",
       "min        -1.711519e+09                -1.656848e+07     0.000000  \n",
       "25%        -1.312109e+08                -4.417888e+04     0.000000  \n",
       "50%        -9.377533e+06                -4.408809e+02     0.000000  \n",
       "75%         1.151925e+08                 4.535706e+04     0.000000  \n",
       "max         1.984283e+09                 2.214944e+07     1.000000  \n",
       "\n",
       "[8 rows x 56 columns]"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking the final predictors data\n",
    "predictors.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "scrolled": true
   },
   "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>Private Domestic (Price Index)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4234</th>\n",
       "      <td>64.471514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>64.453848</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>64.400768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>64.347610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>64.294373</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Private Domestic (Price Index)\n",
       "4234                       64.471514\n",
       "2                          64.453848\n",
       "3                          64.400768\n",
       "4                          64.347610\n",
       "5                          64.294373"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Setting up Target field\n",
    "Target = df.iloc[3: , [1]]\n",
    "Target.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Feature Selection\n",
    "The next step is to perform featue selection using the following techniques:\n",
    "1. Select the top n features based on feature importance from random forest\n",
    "2. Select the top n features based on absolute correlation with target variable\n",
    "3. Select the features identified by Lasso regression\n",
    "4. Perform recursive feature selection and use cross validation to identify the best number of features\n",
    "5. Select the top n features based on absolute value of beta coefficients of features\n",
    "\n",
    "The final features are the ones which were selected in at least 3 out of 5 models.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\ipykernel_launcher.py:8: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24ce5d6f6a0>"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#1.Select the top n features based on feature importance from random forest\n",
    "\n",
    "np.random.seed(10)\n",
    "\n",
    "# define the model\n",
    "model = RandomForestRegressor(random_state = random.seed(10))\n",
    "# fit the model\n",
    "model.fit(predictors, Target)\n",
    "\n",
    "# get importance\n",
    "features = predictors\n",
    "importances = model.feature_importances_\n",
    "indices = np.argsort(importances)\n",
    "\n",
    "feat_importances = pd.Series(model.feature_importances_, index=predictors.columns)\n",
    "feat_importances.nlargest(30).plot(kind='barh')\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Final Features from Random Forest (Select Features with highest feature importance)\n",
    "rf_top_features = pd.DataFrame(feat_importances.nlargest(7)).axes[0].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "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>Correlation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>HSI - close</th>\n",
       "      <td>0.798105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HSI - close_Lag1</th>\n",
       "      <td>0.798300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HSI - close_Lag2</th>\n",
       "      <td>0.798492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HSI - close_Lag3</th>\n",
       "      <td>0.798681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <td>-0.839675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Lag1</th>\n",
       "      <td>-0.839772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Lag2</th>\n",
       "      <td>-0.839865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Lag3</th>\n",
       "      <td>-0.839958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Lag3</th>\n",
       "      <td>0.959957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Lag2</th>\n",
       "      <td>0.960064</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Lag1</th>\n",
       "      <td>0.960169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock</th>\n",
       "      <td>0.960273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI</th>\n",
       "      <td>0.990951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Lag1</th>\n",
       "      <td>0.990956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Lag2</th>\n",
       "      <td>0.990960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Lag3</th>\n",
       "      <td>0.990964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <td>0.992165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Lag1</th>\n",
       "      <td>0.992181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Lag2</th>\n",
       "      <td>0.992196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Lag3</th>\n",
       "      <td>0.992212</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  Correlation\n",
       "HSI - close                                          0.798105\n",
       "HSI - close_Lag1                                     0.798300\n",
       "HSI - close_Lag2                                     0.798492\n",
       "HSI - close_Lag3                                     0.798681\n",
       "Unemployment rate (seasonally adjusted) (%)         -0.839675\n",
       "Unemployment rate (seasonally adjusted) (%)_Lag1    -0.839772\n",
       "Unemployment rate (seasonally adjusted) (%)_Lag2    -0.839865\n",
       "Unemployment rate (seasonally adjusted) (%)_Lag3    -0.839958\n",
       "Total stock_Lag3                                     0.959957\n",
       "Total stock_Lag2                                     0.960064\n",
       "Total stock_Lag1                                     0.960169\n",
       "Total stock                                          0.960273\n",
       "CPI                                                  0.990951\n",
       "CPI_Lag1                                             0.990956\n",
       "CPI_Lag2                                             0.990960\n",
       "CPI_Lag3                                             0.990964\n",
       "M3 (HK$ million)                                     0.992165\n",
       "M3 (HK$ million)_Lag1                                0.992181\n",
       "M3 (HK$ million)_Lag2                                0.992196\n",
       "M3 (HK$ million)_Lag3                                0.992212"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#2.Select the top n features based on absolute correlation with target variable\n",
    "corr_data1 = pd.concat([Target,predictors],axis = 1)\n",
    "corr_data = corr_data1.corr()\n",
    "corr_data = corr_data.iloc[: , [0]]\n",
    "corr_data.columns.values[0] = \"Correlation\"\n",
    "corr_data = corr_data.iloc[corr_data.Correlation.abs().argsort()] \n",
    "corr_data = corr_data[corr_data['Correlation'].notna()]\n",
    "corr_data = corr_data.loc[corr_data['Correlation'] != 1]\n",
    "corr_data.tail(20)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select Features with greater than 90% absolute correlation\n",
    "corr_data2 = corr_data.loc[corr_data['Correlation'].abs() > .9]\n",
    "corr_top_features = corr_data2.axes[0].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Roaming\\Python\\Python37\\site-packages\\sklearn\\utils\\validation.py:72: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  return f(**kwargs)\n",
      "C:\\Users\\ibaruah001\\AppData\\Roaming\\Python\\Python37\\site-packages\\sklearn\\linear_model\\_coordinate_descent.py:531: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 12330.10323860118, tolerance: 6697.469916266773\n",
      "  positive)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['Total completions',\n",
       " 'CPI',\n",
       " 'HSI - close',\n",
       " 'Unemployment rate (seasonally adjusted) (%)_Lag1',\n",
       " 'CPI_Lag1',\n",
       " 'HSI - close_Lag1',\n",
       " 'Total completions_Lag2',\n",
       " 'Unemployment rate (seasonally adjusted) (%)_Lag2',\n",
       " 'CPI_Lag2',\n",
       " 'Total completions_Lag3',\n",
       " 'Total take up_Lag3',\n",
       " 'Unemployment rate (seasonally adjusted) (%)_Lag3',\n",
       " 'CPI_Lag3',\n",
       " 'HSI - close_Lag3',\n",
       " 'Total completions_Diff',\n",
       " 'Total stock_Diff',\n",
       " 'Total take up_Diff',\n",
       " 'Total vacancy_Diff',\n",
       " 'CPI_Diff',\n",
       " 'M3 (HK$ million)_Diff']"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#3.Select the features identified by Lasso regression\n",
    "\n",
    "np.random.seed(10)\n",
    "\n",
    "estimator = LassoCV(cv=5, normalize = True)\n",
    "\n",
    "sfm = SelectFromModel(estimator, prefit=False, norm_order=1, max_features=None)\n",
    "\n",
    "sfm.fit(predictors, Target)\n",
    "\n",
    "feature_idx = sfm.get_support()\n",
    "Lasso_features = predictors.columns[feature_idx].tolist()\n",
    "Lasso_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting estimator with 56 features.\n",
      "Fitting estimator with 46 features.\n",
      "Fitting estimator with 36 features.\n",
      "Fitting estimator with 26 features.\n",
      "Fitting estimator with 16 features.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['CPI',\n",
       " 'CPI_Lag1',\n",
       " 'HIBOR (1-month)_Lag1',\n",
       " 'CPI_Lag2',\n",
       " 'CPI_Lag3',\n",
       " 'HIBOR (1-month)_Diff',\n",
       " 'holidays']"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#4.Perform recursive feature selection and use cross validation to identify the best number of features\n",
    "\n",
    "#Feature ranking with recursive feature elimination and cross-validated selection of the best number of features\n",
    "\n",
    "rfe_selector = RFE(estimator=LinearRegression(), n_features_to_select= 7, step=10, verbose=5)\n",
    "rfe_selector.fit(predictors, Target)\n",
    "rfe_support = rfe_selector.get_support()\n",
    "rfe_feature = predictors.loc[:,rfe_support].columns.tolist()\n",
    "rfe_feature"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#4.Perform recursive feature selection and use cross validation to identify the best number of features\n",
    "\n",
    "#Feature ranking with recursive feature elimination and cross-validated selection of the best number of features\n",
    "\n",
    "\n",
    "\n",
    "names=pd.DataFrame(predictors.columns)\n",
    "\n",
    "#use linear regression as the model\n",
    "lin_reg = LinearRegression()\n",
    "np.random.seed(10)\n",
    "rfe_mod = RFECV(lin_reg, step=1, cv=5)\n",
    "myvalues=rfe_mod.fit(predictors,Target) #to fit\n",
    "myvalues.support_#The mask of selected features.\n",
    "myvalues.ranking_ #The feature ranking, such that ranking_[i] corresponds to the ranking position of the i-th feature. Selected (i.e., estimated best) features are assigned rank 1.\n",
    "\n",
    "rankings=pd.DataFrame(myvalues.ranking_) #Make it into data frame\n",
    "\n",
    "#Concat and name columns\n",
    "ranked=pd.concat([names,rankings], axis=1)\n",
    "ranked.columns = [\"Feature\", \"Rank\"]\n",
    "\n",
    "Top_RFE_d = ranked[ranked['Rank'] < 11] #Get variables with lower rank than 11\n",
    "Top_RFE = Top_RFE_d[\"Feature\"].tolist()\n",
    "Top_RFE\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['M3 (HK$ million)_Lag1',\n",
       " 'M3 (HK$ million)',\n",
       " 'Unemployment rate (seasonally adjusted) (%)',\n",
       " 'Unemployment rate (seasonally adjusted) (%)_Lag1',\n",
       " 'Total completions',\n",
       " 'Total completions_Lag1',\n",
       " 'Total stock_Lag1',\n",
       " 'Total stock',\n",
       " 'Total take up_Lag1',\n",
       " 'Total take up']"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#5.Select the top n features based on absolute value of beta coefficients of features\n",
    "\n",
    "# define standard scaler\n",
    "scaler = StandardScaler()\n",
    "# transform x data\n",
    "scaled_predictors = scaler.fit_transform(predictors)\n",
    "scaled_Target = scaler.fit_transform(Target)\n",
    "\n",
    "sr_reg = LinearRegression(fit_intercept = False).fit(scaled_predictors, scaled_Target)\n",
    "coef_table = pd.DataFrame(list(predictors.columns)).copy()\n",
    "coef_table.insert(len(coef_table.columns),\"Coefs\",sr_reg.coef_.transpose())\n",
    "coef_table = coef_table.iloc[coef_table.Coefs.abs().argsort()] \n",
    "\n",
    "\n",
    "sr_data2 = coef_table.tail(10)\n",
    "sr_top_features = sr_data2.iloc[:,0].tolist()\n",
    "sr_top_features"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#Extra Models for feature selection:\n",
    "\n",
    "\n",
    "#Forward Regression\n",
    "\n",
    "import statsmodels.api as sm\n",
    "\n",
    "def forward_regression(X, y,\n",
    "                       initial_list=[], \n",
    "                       threshold_in=0.01, \n",
    "                       threshold_out = 0.05, \n",
    "                       verbose=True):\n",
    "    initial_list = []\n",
    "    included = list(initial_list)\n",
    "    while True:\n",
    "        changed=False\n",
    "        # forward step\n",
    "        excluded = list(set(X.columns)-set(included))\n",
    "        new_pval = pd.Series(index=excluded)\n",
    "        for new_column in excluded:\n",
    "            model = sm.OLS(y, sm.add_constant(pd.DataFrame(X[included+[new_column]]))).fit()\n",
    "            new_pval[new_column] = model.pvalues[new_column]\n",
    "        best_pval = new_pval.min()\n",
    "        if best_pval < threshold_in:\n",
    "            best_feature = new_pval.argmin()\n",
    "            included.append(best_feature)\n",
    "            changed=True\n",
    "            if verbose:\n",
    "                print('Add   with p-value '.format(best_feature, best_pval))\n",
    "\n",
    "        if not changed:\n",
    "            break\n",
    "\n",
    "    return included\n",
    "\n",
    "forward_regression(predictors, Target)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "#Extra Models for feature selection:\n",
    "\n",
    "\n",
    "#Granger Causality\n",
    "\n",
    "gc_data = pd.concat([Target, predictors], axis=1, sort=False)\n",
    "\n",
    "maxlag=12\n",
    "test = 'ssr_chi2test'\n",
    "def grangers_causation_matrix(data, variables, test='ssr_chi2test', verbose=False):    \n",
    "    \"\"\"Check Granger Causality of all possible combinations of the Time series.\n",
    "    The rows are the response variable, columns are predictors. The values in the table \n",
    "    are the P-Values. P-Values lesser than the significance level (0.05), implies \n",
    "    the Null Hypothesis that the coefficients of the corresponding past values is \n",
    "    zero, that is, the X does not cause Y can be rejected.\n",
    "\n",
    "    data      : pandas dataframe containing the time series variables\n",
    "    variables : list containing names of the time series variables.\n",
    "    \"\"\"\n",
    "    df_gc = pd.DataFrame(np.zeros((len(variables), len(variables))), columns=variables, index=variables)\n",
    "    for c in df_gc.columns:\n",
    "        for r in df_gc.index:\n",
    "            test_result = grangercausalitytests(data[[r, c]], maxlag=maxlag, verbose=False)\n",
    "            p_values = [round(test_result[i+1][0][test][1],4) for i in range(maxlag)]\n",
    "            if verbose: print(f'Y = {r}, X = {c}, P Values = {p_values}')\n",
    "            min_p_value = np.min(p_values)\n",
    "            df_gc.loc[r, c] = min_p_value\n",
    "    df_gc.columns = [var + '_x' for var in variables]\n",
    "    df_gc.index = [var + '_y' for var in variables]\n",
    "    return df_gc\n",
    "\n",
    "grangers_causation_matrix(data = gc_data, variables = gc_data.columns)        "
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "All_drivers = pd.DataFrame(sr_top_features)\n",
    "All_drivers['Model'] = 'SR'\n",
    "RFE_Drivers = pd.DataFrame(Top_RFE)\n",
    "RFE_Drivers['Model'] = 'RFE'\n",
    "Lasso_Drivers = pd.DataFrame(Lasso_features)\n",
    "Lasso_Drivers['Model'] = 'Lasso'\n",
    "Cor_Drivers = pd.DataFrame(corr_top_features)\n",
    "Cor_Drivers['Model'] = 'Cor'\n",
    "RF_Drivers = pd.DataFrame(rf_top_features)\n",
    "RF_Drivers['Model'] = 'RF'\n",
    "All_drivers = All_drivers.append([RFE_Drivers,Lasso_Drivers,Cor_Drivers,RF_Drivers])\n",
    "pd.set_option(\"display.max_rows\", None, \"display.max_columns\", None)\n",
    "All_drivers.head(100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "scrolled": true
   },
   "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>number_of_models</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Lag1</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Lag2</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Lag2</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Lag3</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Lag3</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Lag1</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Lag2</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Lag3</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Lag1</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total completions</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Lag1</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total completions_Diff</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M3 (HK$ million)_Diff</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CPI_Diff</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total vacancy_Diff</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total take up_Diff</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total stock_Diff</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total take up</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HSI - close</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total completions_Lag1</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Lag3</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total take up_Lag3</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total completions_Lag3</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Unemployment rate (seasonally adjusted) (%)_Lag2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total completions_Lag2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HSI - close_Lag1</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Total take up_Lag1</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>HSI - close_Lag3</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  number_of_models\n",
       "M3 (HK$ million)_Lag1                                            3\n",
       "M3 (HK$ million)                                                 3\n",
       "Total stock                                                      3\n",
       "CPI                                                              2\n",
       "M3 (HK$ million)_Lag2                                            2\n",
       "Total stock_Lag2                                                 2\n",
       "Total stock_Lag3                                                 2\n",
       "CPI_Lag3                                                         2\n",
       "CPI_Lag1                                                         2\n",
       "CPI_Lag2                                                         2\n",
       "M3 (HK$ million)_Lag3                                            2\n",
       "Unemployment rate (seasonally adjusted) (%)_Lag1                 2\n",
       "Total completions                                                2\n",
       "Total stock_Lag1                                                 2\n",
       "Total completions_Diff                                           1\n",
       "Unemployment rate (seasonally adjusted) (%)                      1\n",
       "M3 (HK$ million)_Diff                                            1\n",
       "CPI_Diff                                                         1\n",
       "Total vacancy_Diff                                               1\n",
       "Total take up_Diff                                               1\n",
       "Total stock_Diff                                                 1\n",
       "Total take up                                                    1\n",
       "HSI - close                                                      1\n",
       "Total completions_Lag1                                           1\n",
       "Unemployment rate (seasonally adjusted) (%)_Lag3                 1\n",
       "Total take up_Lag3                                               1\n",
       "Total completions_Lag3                                           1\n",
       "Unemployment rate (seasonally adjusted) (%)_Lag2                 1\n",
       "Total completions_Lag2                                           1\n",
       "HSI - close_Lag1                                                 1\n",
       "Total take up_Lag1                                               1\n",
       "HSI - close_Lag3                                                 1"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Combining features from all the models\n",
    "\n",
    "combined_feature_list = sr_top_features + Lasso_features + corr_top_features + rf_top_features\n",
    "\n",
    "combined_feature = {x:combined_feature_list.count(x) for x in combined_feature_list}\n",
    "combined_feature_data = pd.DataFrame.from_dict(combined_feature,orient='index')\n",
    "\n",
    "combined_feature_data.rename(columns={ combined_feature_data.columns[0]: \"number_of_models\" }, inplace = True)\n",
    "\n",
    "\n",
    "combined_feature_data = combined_feature_data.sort_values(['number_of_models'], ascending=[False])\n",
    "\n",
    "combined_feature_data.head(100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['M3 (HK$ million)_Lag1', 'M3 (HK$ million)', 'Total stock']"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Final Features: features which were selected in atleast 3 models\n",
    "\n",
    "combined_feature_data = combined_feature_data.loc[combined_feature_data['number_of_models'] > 2]\n",
    "final_features = combined_feature_data.axes[0].tolist()\n",
    "final_features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Preparation for Forecasting Engine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "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>Total stock</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>M3 (HK$ million)_Lag1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2003-01-05</th>\n",
       "      <td>1.091846e+06</td>\n",
       "      <td>3.562891e+06</td>\n",
       "      <td>3.562739e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-06</th>\n",
       "      <td>1.091646e+06</td>\n",
       "      <td>3.563044e+06</td>\n",
       "      <td>3.562891e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-07</th>\n",
       "      <td>1.091050e+06</td>\n",
       "      <td>3.563562e+06</td>\n",
       "      <td>3.563044e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-08</th>\n",
       "      <td>1.090457e+06</td>\n",
       "      <td>3.564141e+06</td>\n",
       "      <td>3.563562e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-09</th>\n",
       "      <td>1.089865e+06</td>\n",
       "      <td>3.564781e+06</td>\n",
       "      <td>3.564141e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-10</th>\n",
       "      <td>1.089275e+06</td>\n",
       "      <td>3.565481e+06</td>\n",
       "      <td>3.564781e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-11</th>\n",
       "      <td>1.089080e+06</td>\n",
       "      <td>3.565735e+06</td>\n",
       "      <td>3.565481e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-12</th>\n",
       "      <td>1.088884e+06</td>\n",
       "      <td>3.565988e+06</td>\n",
       "      <td>3.565735e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-13</th>\n",
       "      <td>1.088688e+06</td>\n",
       "      <td>3.566242e+06</td>\n",
       "      <td>3.565988e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2003-01-14</th>\n",
       "      <td>1.088102e+06</td>\n",
       "      <td>3.567063e+06</td>\n",
       "      <td>3.566242e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Total stock  M3 (HK$ million)  M3 (HK$ million)_Lag1\n",
       "Date                                                             \n",
       "2003-01-05  1.091846e+06      3.562891e+06           3.562739e+06\n",
       "2003-01-06  1.091646e+06      3.563044e+06           3.562891e+06\n",
       "2003-01-07  1.091050e+06      3.563562e+06           3.563044e+06\n",
       "2003-01-08  1.090457e+06      3.564141e+06           3.563562e+06\n",
       "2003-01-09  1.089865e+06      3.564781e+06           3.564141e+06\n",
       "2003-01-10  1.089275e+06      3.565481e+06           3.564781e+06\n",
       "2003-01-11  1.089080e+06      3.565735e+06           3.565481e+06\n",
       "2003-01-12  1.088884e+06      3.565988e+06           3.565735e+06\n",
       "2003-01-13  1.088688e+06      3.566242e+06           3.565988e+06\n",
       "2003-01-14  1.088102e+06      3.567063e+06           3.566242e+06"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Final Master Data on Selected Features\n",
    "features = predictors[predictors.columns.intersection(final_features)]\n",
    "features_date =  df.iloc[3:,0]\n",
    "features_final = features.set_index(features_date).asfreq('D')\n",
    "features_final.head(10)                                           "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot all Final Features\n",
    "fig, axes = pyplot.subplots(nrows=3, ncols=1, dpi=120, figsize=(10,6))\n",
    "for i, ax in enumerate(axes.flatten()):\n",
    "    plot_data = features_final[features_final.columns[i]]\n",
    "    ax.plot(plot_data, color='blue', linewidth=1)\n",
    "    # Decorations\n",
    "    ax.set_title(features_final.columns[i])\n",
    "    ax.xaxis.set_ticks_position('none')\n",
    "    ax.yaxis.set_ticks_position('none')\n",
    "    ax.spines[\"top\"].set_alpha(0)\n",
    "    ax.tick_params(labelsize=6)\n",
    "\n",
    "pyplot.tight_layout();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date\n",
       "2019-11-22    366.678619\n",
       "2019-11-23    366.534478\n",
       "2019-11-24    366.390336\n",
       "2019-11-25    366.246194\n",
       "2019-11-26    365.810997\n",
       "Freq: D, Name: Private Domestic (Price Index), dtype: float64"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Final Target Variable Dataset\n",
    "target_final = series.iloc[3:,]\n",
    "target_final.tail(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Univariate Time Series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\statsmodels\\tsa\\holtwinters.py:725: RuntimeWarning: invalid value encountered in less_equal\n",
      "  loc = initial_p <= lb\n",
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\statsmodels\\tsa\\holtwinters.py:731: RuntimeWarning: invalid value encountered in greater_equal\n",
      "  loc = initial_p >= ub\n"
     ]
    },
    {
     "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>HWES_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>365.584796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>365.319148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>365.063393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>364.761805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>364.444793</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            HWES_Forecast\n",
       "2019-11-27     365.584796\n",
       "2019-11-28     365.319148\n",
       "2019-11-29     365.063393\n",
       "2019-11-30     364.761805\n",
       "2019-12-01     364.444793"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Holt Winter’s Exponential Smoothing (HWES)or Triple Smoothing\n",
    "# fit model\n",
    "\n",
    "random.seed(10)\n",
    "\n",
    "model = ExponentialSmoothing(series, trend = 'add',seasonal= 'add', seasonal_periods= 7)\n",
    "model_fit = model.fit()\n",
    "# make prediction\n",
    "HWES_Forecast = pd.DataFrame(model_fit.forecast(steps=365))\n",
    "HWES_Forecast.columns = ['HWES_Forecast']\n",
    "HWES_Forecast.head(5)\n",
    "\n",
    "\n",
    "#Plot\n",
    "#HWES_Forecast.plot(marker='o', color='red', legend=True)\n",
    "#model_fit.fittedvalues.plot(marker='o',  color='blue')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing stepwise search to minimize aic\n",
      " ARIMA(1,2,1)(0,0,0)[7]             : AIC=-197.062, Time=1.42 sec\n",
      " ARIMA(0,2,0)(0,0,0)[7]             : AIC=5058.119, Time=0.23 sec\n",
      " ARIMA(1,2,0)(1,0,0)[7]             : AIC=3074.520, Time=1.06 sec\n",
      " ARIMA(0,2,1)(0,0,1)[7]             : AIC=0.776, Time=1.85 sec\n",
      " ARIMA(1,2,1)(1,0,0)[7]             : AIC=-230.999, Time=2.31 sec\n",
      " ARIMA(1,2,1)(2,0,0)[7]             : AIC=-265.196, Time=5.17 sec\n",
      " ARIMA(1,2,1)(3,0,0)[7]             : AIC=-292.867, Time=9.47 sec\n",
      " ARIMA(1,2,1)(3,0,1)[7]             : AIC=-338.219, Time=21.87 sec\n",
      " ARIMA(1,2,1)(2,0,1)[7]             : AIC=-339.959, Time=9.65 sec\n",
      " ARIMA(1,2,1)(1,0,1)[7]             : AIC=-340.231, Time=5.27 sec\n",
      " ARIMA(1,2,1)(0,0,1)[7]             : AIC=-226.245, Time=2.27 sec\n",
      " ARIMA(1,2,1)(1,0,2)[7]             : AIC=-339.988, Time=11.16 sec\n",
      " ARIMA(1,2,1)(0,0,2)[7]             : AIC=-254.390, Time=5.05 sec\n",
      " ARIMA(1,2,1)(2,0,2)[7]             : AIC=-336.264, Time=11.90 sec\n",
      " ARIMA(0,2,1)(1,0,1)[7]             : AIC=-79.167, Time=3.88 sec\n",
      " ARIMA(1,2,0)(1,0,1)[7]             : AIC=3038.408, Time=2.43 sec\n",
      " ARIMA(2,2,1)(1,0,1)[7]             : AIC=-563.822, Time=5.98 sec\n",
      " ARIMA(2,2,1)(0,0,1)[7]             : AIC=-434.301, Time=2.77 sec\n",
      " ARIMA(2,2,1)(1,0,0)[7]             : AIC=-440.657, Time=2.82 sec\n",
      " ARIMA(2,2,1)(2,0,1)[7]             : AIC=-563.715, Time=9.93 sec\n",
      " ARIMA(2,2,1)(1,0,2)[7]             : AIC=-563.741, Time=12.99 sec\n",
      " ARIMA(2,2,1)(0,0,0)[7]             : AIC=-398.592, Time=1.16 sec\n",
      " ARIMA(2,2,1)(0,0,2)[7]             : AIC=-468.531, Time=6.11 sec\n",
      " ARIMA(2,2,1)(2,0,0)[7]             : AIC=-482.807, Time=6.04 sec\n",
      " ARIMA(2,2,1)(2,0,2)[7]             : AIC=-559.938, Time=11.56 sec\n",
      " ARIMA(2,2,0)(1,0,1)[7]             : AIC=2073.323, Time=3.30 sec\n",
      " ARIMA(3,2,1)(1,0,1)[7]             : AIC=-748.973, Time=7.93 sec\n",
      " ARIMA(3,2,1)(0,0,1)[7]             : AIC=-626.111, Time=3.63 sec\n",
      " ARIMA(3,2,1)(1,0,0)[7]             : AIC=-631.622, Time=4.77 sec\n",
      " ARIMA(3,2,1)(2,0,1)[7]             : AIC=-749.429, Time=19.56 sec\n",
      " ARIMA(3,2,1)(2,0,0)[7]             : AIC=-671.923, Time=6.92 sec\n",
      " ARIMA(3,2,1)(3,0,1)[7]             : AIC=-747.660, Time=30.57 sec\n",
      " ARIMA(3,2,1)(2,0,2)[7]             : AIC=-745.395, Time=15.83 sec\n",
      " ARIMA(3,2,1)(1,0,2)[7]             : AIC=-749.468, Time=14.96 sec\n",
      " ARIMA(3,2,1)(0,0,2)[7]             : AIC=-658.553, Time=8.64 sec\n",
      " ARIMA(3,2,1)(1,0,3)[7]             : AIC=-747.667, Time=31.43 sec\n",
      " ARIMA(3,2,1)(0,0,3)[7]             : AIC=-687.256, Time=14.90 sec\n",
      " ARIMA(3,2,1)(2,0,3)[7]             : AIC=-745.972, Time=26.15 sec\n",
      " ARIMA(3,2,0)(1,0,2)[7]             : AIC=1380.467, Time=7.16 sec\n",
      " ARIMA(3,2,2)(1,0,2)[7]             : AIC=-605.815, Time=20.87 sec\n",
      " ARIMA(2,2,0)(1,0,2)[7]             : AIC=2075.098, Time=5.79 sec\n",
      " ARIMA(2,2,2)(1,0,2)[7]             : AIC=-356.332, Time=19.54 sec\n",
      " ARIMA(3,2,1)(1,0,2)[7] intercept   : AIC=-660.000, Time=41.57 sec\n",
      "\n",
      "Best model:  ARIMA(3,2,1)(1,0,2)[7]          \n",
      "Total fit time: 437.941 seconds\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>SARIMAX Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>                    <td>y</td>                 <th>  No. Observations:  </th>   <td>6170</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(3, 2, 1)x(1, 0, [1, 2], 7)</td> <th>  Log Likelihood     </th>  <td>382.734</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>                     <td>Fri, 25 Sep 2020</td>          <th>  AIC                </th> <td>-749.468</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                         <td>10:32:42</td>              <th>  BIC                </th> <td>-695.651</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>                           <td>0</td>                 <th>  HQIC               </th> <td>-730.805</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                               <td> - 6170</td>              <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>                 <td>opg</td>                <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>        <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L1</th>    <td>   -0.3012</td> <td>    0.005</td> <td>  -57.764</td> <td> 0.000</td> <td>   -0.311</td> <td>   -0.291</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L2</th>    <td>   -0.2482</td> <td>    0.007</td> <td>  -35.581</td> <td> 0.000</td> <td>   -0.262</td> <td>   -0.235</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L3</th>    <td>   -0.1778</td> <td>    0.008</td> <td>  -22.832</td> <td> 0.000</td> <td>   -0.193</td> <td>   -0.163</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1</th>    <td>   -0.9392</td> <td>    0.004</td> <td> -226.455</td> <td> 0.000</td> <td>   -0.947</td> <td>   -0.931</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.S.L7</th>  <td>    0.8677</td> <td>    0.025</td> <td>   34.616</td> <td> 0.000</td> <td>    0.819</td> <td>    0.917</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L7</th>  <td>   -0.7997</td> <td>    0.030</td> <td>  -26.803</td> <td> 0.000</td> <td>   -0.858</td> <td>   -0.741</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L14</th> <td>    0.0216</td> <td>    0.020</td> <td>    1.096</td> <td> 0.273</td> <td>   -0.017</td> <td>    0.060</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th>   <td>    0.0517</td> <td>    0.000</td> <td>  243.205</td> <td> 0.000</td> <td>    0.051</td> <td>    0.052</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (Q):</th>          <td>298.30</td> <th>  Jarque-Bera (JB):  </th> <td>480097.01</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                 <td>0.00</td>  <th>  Prob(JB):          </th>   <td>0.00</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th>  <td>1.43</td>  <th>  Skew:              </th>   <td>0.78</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>     <td>0.00</td>  <th>  Kurtosis:          </th>   <td>46.19</td>  \n",
       "</tr>\n",
       "</table><br/><br/>Warnings:<br/>[1] Covariance matrix calculated using the outer product of gradients (complex-step)."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                                       SARIMAX Results                                        \n",
       "==============================================================================================\n",
       "Dep. Variable:                                      y   No. Observations:                 6170\n",
       "Model:             SARIMAX(3, 2, 1)x(1, 0, [1, 2], 7)   Log Likelihood                 382.734\n",
       "Date:                                Fri, 25 Sep 2020   AIC                           -749.468\n",
       "Time:                                        10:32:42   BIC                           -695.651\n",
       "Sample:                                             0   HQIC                          -730.805\n",
       "                                               - 6170                                         \n",
       "Covariance Type:                                  opg                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "ar.L1         -0.3012      0.005    -57.764      0.000      -0.311      -0.291\n",
       "ar.L2         -0.2482      0.007    -35.581      0.000      -0.262      -0.235\n",
       "ar.L3         -0.1778      0.008    -22.832      0.000      -0.193      -0.163\n",
       "ma.L1         -0.9392      0.004   -226.455      0.000      -0.947      -0.931\n",
       "ar.S.L7        0.8677      0.025     34.616      0.000       0.819       0.917\n",
       "ma.S.L7       -0.7997      0.030    -26.803      0.000      -0.858      -0.741\n",
       "ma.S.L14       0.0216      0.020      1.096      0.273      -0.017       0.060\n",
       "sigma2         0.0517      0.000    243.205      0.000       0.051       0.052\n",
       "===================================================================================\n",
       "Ljung-Box (Q):                      298.30   Jarque-Bera (JB):            480097.01\n",
       "Prob(Q):                              0.00   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):               1.43   Skew:                             0.78\n",
       "Prob(H) (two-sided):                  0.00   Kurtosis:                        46.19\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fitting a stepwise model to find the best paramters for SARIMA:\n",
    "stepwise_fit = pm.auto_arima(target_final, start_p=1, start_q=1, max_p=3, max_d=2,max_q=3,m=7,\n",
    "                             start_P=0,start_Q=0,max_P=3, max_D=3,max_Q=3, seasonal=True, trace=True,\n",
    "                             error_action='ignore',  # don't want to know if an order does not work\n",
    "                             suppress_warnings=True,  # don't want convergence warnings\n",
    "                             stepwise=True)  # set to stepwise\n",
    "\n",
    "stepwise_fit.summary()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "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>SARIMA_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>365.470756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>365.143184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>364.820468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>364.575509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>364.329815</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            SARIMA_Forecast\n",
       "2019-11-27       365.470756\n",
       "2019-11-28       365.143184\n",
       "2019-11-29       364.820468\n",
       "2019-11-30       364.575509\n",
       "2019-12-01       364.329815"
      ]
     },
     "execution_count": 136,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SARIMA \n",
    "\n",
    "# Using parameters automatically based on grid serach\n",
    "SARIMA_Forecast = pd.DataFrame(stepwise_fit.predict(n_periods= 365))\n",
    "SARIMA_Forecast.columns = ['SARIMA_Forecast']\n",
    "SARIMA_Forecast.index = HWES_Forecast.index\n",
    "SARIMA_Forecast.head(5)\n",
    "\n",
    "\n",
    "# Manually fit the model\n",
    "#sarima_model = SARIMAX(series, order=(5, 2, 2), seasonal_order=(0, 0, 1, 7))\n",
    "#sarima_model_fit = sarima_model.fit(disp=False)\n",
    "# make prediction\n",
    "#SARIMA_Forecast = model_fit.predict(len(data), len(data))\n",
    "#print(yhat)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Prep for multivariate models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to check stationarity\n",
    "\n",
    "def adfuller_test(series, signif=0.05, name='', verbose=False):\n",
    "    \"\"\"Perform ADFuller to test for Stationarity of given series and print report\"\"\"\n",
    "    r = adfuller(series, autolag='AIC')\n",
    "    output = {'test_statistic':round(r[0], 4), 'pvalue':round(r[1], 4), 'n_lags':round(r[2], 4), 'n_obs':r[3]}\n",
    "    p_value = output['pvalue'] \n",
    "    def adjust(val, length= 6): return str(val).ljust(length)\n",
    "\n",
    "    # Print Summary\n",
    "    print(f'    Augmented Dickey-Fuller Test on \"{name}\"', \"\\n   \", '-'*47)\n",
    "    print(f' Null Hypothesis: Data has unit root. Non-Stationary.')\n",
    "    print(f' Significance Level    = {signif}')\n",
    "    print(f' Test Statistic        = {output[\"test_statistic\"]}')\n",
    "    print(f' No. Lags Chosen       = {output[\"n_lags\"]}')\n",
    "\n",
    "    for key,val in r[4].items():\n",
    "        print(f' Critical value {adjust(key)} = {round(val, 3)}')\n",
    "\n",
    "    if p_value <= signif:\n",
    "        print(f\" => P-Value = {p_value}. Rejecting Null Hypothesis.\")\n",
    "        print(f\" => Series is Stationary.\")\n",
    "    else:\n",
    "        print(f\" => P-Value = {p_value}. Weak evidence to reject the Null Hypothesis.\")\n",
    "        print(f\" => Series is Non-Stationary.\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    Augmented Dickey-Fuller Test on \"Private Domestic (Price Index)\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -1.2294\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.6608. Weak evidence to reject the Null Hypothesis.\n",
      " => Series is Non-Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"Total stock\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -1.982\n",
      " No. Lags Chosen       = 33\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.2945. Weak evidence to reject the Null Hypothesis.\n",
      " => Series is Non-Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"M3 (HK$ million)\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = 0.5809\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.9871. Weak evidence to reject the Null Hypothesis.\n",
      " => Series is Non-Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"M3 (HK$ million)_Lag1\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = 0.6292\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.9883. Weak evidence to reject the Null Hypothesis.\n",
      " => Series is Non-Stationary.\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Data Prep\n",
    "\n",
    "AI_Model_data = pd.concat([target_final, features_final], axis=1, sort=False)\n",
    "\n",
    "# Check stationarity\n",
    "\n",
    "# ADF Test on each column\n",
    "for name, column in AI_Model_data.iteritems():\n",
    "    adfuller_test(column, name=column.name)\n",
    "    print('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    Augmented Dickey-Fuller Test on \"Private Domestic (Price Index)\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -3.6108\n",
      " No. Lags Chosen       = 33\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0056. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"Total stock\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -8.7666\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"M3 (HK$ million)\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -10.1037\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"M3 (HK$ million)_Lag1\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -10.1674\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# 1st difference\n",
    "AI_Model_data_differenced = AI_Model_data.diff().dropna()\n",
    "\n",
    "\n",
    "# Check stationarity\n",
    "\n",
    "# ADF Test on each column\n",
    "#for name, column in AI_Model_data_differenced.iteritems():\n",
    "    #adfuller_test(column, name=column.name)\n",
    "    #print('\\n')\n",
    "\n",
    "x = AI_Model_data_differenced.values #returns a numpy array\n",
    "\n",
    "\n",
    "# Standardise Datasets\n",
    "\n",
    "standard_scaler = StandardScaler()\n",
    "x_scaled = standard_scaler.fit_transform(x)\n",
    "scaled_AI_Model_data = pd.DataFrame(x_scaled)\n",
    "scaled_AI_Model_data.columns = AI_Model_data_differenced.columns\n",
    "\n",
    "\n",
    "\n",
    "# Check stationarity on diffrenced and scaled data\n",
    "\n",
    "# ADF Test on each column\n",
    "for name, column in scaled_AI_Model_data.iteritems():\n",
    "    adfuller_test(column, name=column.name)\n",
    "    print('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "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>Total stock</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>M3 (HK$ million)_Lag1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-2.757881</td>\n",
       "      <td>-0.320427</td>\n",
       "      <td>-0.320983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-7.745738</td>\n",
       "      <td>-0.246206</td>\n",
       "      <td>-0.320983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-7.719380</td>\n",
       "      <td>-0.233910</td>\n",
       "      <td>-0.246720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-7.693022</td>\n",
       "      <td>-0.221615</td>\n",
       "      <td>-0.234418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-7.666664</td>\n",
       "      <td>-0.209319</td>\n",
       "      <td>-0.222116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6164</th>\n",
       "      <td>0.996238</td>\n",
       "      <td>-2.568317</td>\n",
       "      <td>-2.531685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6165</th>\n",
       "      <td>0.165155</td>\n",
       "      <td>-1.103174</td>\n",
       "      <td>-2.570131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6166</th>\n",
       "      <td>0.165155</td>\n",
       "      <td>-1.103174</td>\n",
       "      <td>-1.104168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6167</th>\n",
       "      <td>0.165155</td>\n",
       "      <td>-1.103174</td>\n",
       "      <td>-1.104168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6168</th>\n",
       "      <td>0.997789</td>\n",
       "      <td>-2.645166</td>\n",
       "      <td>-1.104168</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>6169 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Total stock  M3 (HK$ million)  M3 (HK$ million)_Lag1\n",
       "0       -2.757881         -0.320427              -0.320983\n",
       "1       -7.745738         -0.246206              -0.320983\n",
       "2       -7.719380         -0.233910              -0.246720\n",
       "3       -7.693022         -0.221615              -0.234418\n",
       "4       -7.666664         -0.209319              -0.222116\n",
       "...           ...               ...                    ...\n",
       "6164     0.996238         -2.568317              -2.531685\n",
       "6165     0.165155         -1.103174              -2.570131\n",
       "6166     0.165155         -1.103174              -1.104168\n",
       "6167     0.165155         -1.103174              -1.104168\n",
       "6168     0.997789         -2.645166              -1.104168\n",
       "\n",
       "[6169 rows x 3 columns]"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Separate out traget and features\n",
    "\n",
    "scaled_AI_Model_Target = scaled_AI_Model_data.iloc[:,0]\n",
    "scaled_AI_Model_Features = scaled_AI_Model_data.iloc[:,1:]\n",
    "\n",
    "#Mean and Standard Deviation\n",
    "mean_target = AI_Model_data_differenced.iloc[:,0].mean()\n",
    "std_target = AI_Model_data_differenced.iloc[:,0].std()\n",
    "\n",
    "\n",
    "scaled_AI_Model_Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5804, 3)\n",
      "(365, 3)\n",
      "(5804,)\n",
      "(365,)\n"
     ]
    }
   ],
   "source": [
    "# Train Test Split\n",
    "\n",
    "nobs = 365\n",
    "features_train, features_test = scaled_AI_Model_Features[0:-nobs], features_final[-nobs:]\n",
    "target_train, target_test = scaled_AI_Model_Target[0:-nobs], target_final[-nobs:]\n",
    "\n",
    "# Check size\n",
    "print(features_train.shape)  # (5805, 9)\n",
    "print(features_test.shape)  # (365, 9)\n",
    "print(target_train.shape)  # (5805, )\n",
    "print(target_test.shape)  # (365, )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing stepwise search to minimize aic\n",
      " ARIMA(1,1,1)(0,0,0)[7] intercept   : AIC=6763.340, Time=3.74 sec\n",
      " ARIMA(0,1,0)(0,0,0)[7] intercept   : AIC=8526.451, Time=0.51 sec\n",
      " ARIMA(1,1,0)(1,0,0)[7] intercept   : AIC=1680.008, Time=7.21 sec\n",
      " ARIMA(0,1,1)(0,0,1)[7] intercept   : AIC=5447.930, Time=5.03 sec\n",
      " ARIMA(0,1,0)(0,0,0)[7]             : AIC=8524.461, Time=0.17 sec\n",
      " ARIMA(1,1,0)(0,0,0)[7] intercept   : AIC=8526.053, Time=0.57 sec\n",
      " ARIMA(1,1,0)(2,0,0)[7] intercept   : AIC=inf, Time=9.91 sec\n",
      " ARIMA(1,1,0)(1,0,1)[7] intercept   : AIC=-845.626, Time=10.14 sec\n",
      " ARIMA(1,1,0)(0,0,1)[7] intercept   : AIC=5485.680, Time=3.09 sec\n",
      " ARIMA(1,1,0)(2,0,1)[7] intercept   : AIC=-851.121, Time=26.03 sec\n",
      " ARIMA(1,1,0)(2,0,2)[7] intercept   : AIC=inf, Time=28.14 sec\n",
      " ARIMA(1,1,0)(1,0,2)[7] intercept   : AIC=-852.629, Time=21.29 sec\n",
      " ARIMA(1,1,0)(0,0,2)[7] intercept   : AIC=4078.019, Time=6.08 sec\n",
      " ARIMA(0,1,0)(1,0,2)[7] intercept   : AIC=-153.096, Time=18.92 sec\n",
      " ARIMA(2,1,0)(1,0,2)[7] intercept   : AIC=-1189.617, Time=23.33 sec\n",
      " ARIMA(2,1,0)(0,0,2)[7] intercept   : AIC=4034.204, Time=7.02 sec\n",
      " ARIMA(2,1,0)(1,0,1)[7] intercept   : AIC=-1162.585, Time=12.56 sec\n",
      " ARIMA(2,1,0)(2,0,2)[7] intercept   : AIC=inf, Time=32.94 sec\n",
      " ARIMA(2,1,0)(0,0,1)[7] intercept   : AIC=5476.076, Time=2.58 sec\n",
      " ARIMA(2,1,0)(2,0,1)[7] intercept   : AIC=-1184.299, Time=26.86 sec\n",
      " ARIMA(2,1,1)(1,0,2)[7] intercept   : AIC=-2357.944, Time=33.80 sec\n",
      " ARIMA(2,1,1)(0,0,2)[7] intercept   : AIC=2049.488, Time=24.73 sec\n",
      " ARIMA(2,1,1)(1,0,1)[7] intercept   : AIC=-2506.164, Time=19.75 sec\n",
      " ARIMA(2,1,1)(0,0,1)[7] intercept   : AIC=3324.747, Time=11.05 sec\n",
      " ARIMA(2,1,1)(1,0,0)[7] intercept   : AIC=137.931, Time=13.38 sec\n",
      " ARIMA(2,1,1)(2,0,1)[7] intercept   : AIC=-2393.579, Time=30.40 sec\n",
      " ARIMA(2,1,1)(0,0,0)[7] intercept   : AIC=5912.509, Time=6.69 sec\n",
      " ARIMA(2,1,1)(2,0,0)[7] intercept   : AIC=inf, Time=36.63 sec\n",
      " ARIMA(2,1,1)(2,0,2)[7] intercept   : AIC=inf, Time=38.45 sec\n",
      " ARIMA(1,1,1)(1,0,1)[7] intercept   : AIC=-2492.680, Time=15.91 sec\n",
      " ARIMA(2,1,2)(1,0,1)[7] intercept   : AIC=inf, Time=17.60 sec\n",
      " ARIMA(1,1,2)(1,0,1)[7] intercept   : AIC=-2382.083, Time=20.71 sec\n",
      " ARIMA(2,1,1)(1,0,1)[7]             : AIC=-2548.774, Time=5.17 sec\n",
      " ARIMA(2,1,1)(0,0,1)[7]             : AIC=3330.280, Time=2.75 sec\n",
      " ARIMA(2,1,1)(1,0,0)[7]             : AIC=-18.181, Time=5.00 sec\n",
      " ARIMA(2,1,1)(2,0,1)[7]             : AIC=-2554.441, Time=13.33 sec\n",
      " ARIMA(2,1,1)(2,0,0)[7]             : AIC=-1343.462, Time=12.49 sec\n",
      " ARIMA(2,1,1)(2,0,2)[7]             : AIC=inf, Time=18.23 sec\n",
      " ARIMA(2,1,1)(1,0,2)[7]             : AIC=-2555.501, Time=13.32 sec\n",
      " ARIMA(2,1,1)(0,0,2)[7]             : AIC=2054.266, Time=6.24 sec\n",
      " ARIMA(1,1,1)(1,0,2)[7]             : AIC=-2557.006, Time=11.26 sec\n",
      " ARIMA(1,1,1)(0,0,2)[7]             : AIC=2241.459, Time=7.48 sec\n",
      " ARIMA(1,1,1)(1,0,1)[7]             : AIC=-2550.230, Time=4.21 sec\n",
      " ARIMA(1,1,1)(2,0,2)[7]             : AIC=inf, Time=14.28 sec\n",
      " ARIMA(1,1,1)(0,0,1)[7]             : AIC=3663.619, Time=2.07 sec\n",
      " ARIMA(1,1,1)(2,0,1)[7]             : AIC=-2555.930, Time=9.87 sec\n",
      " ARIMA(0,1,1)(1,0,2)[7]             : AIC=-1919.298, Time=9.79 sec\n",
      " ARIMA(1,1,0)(1,0,2)[7]             : AIC=-854.629, Time=7.93 sec\n",
      " ARIMA(1,1,2)(1,0,2)[7]             : AIC=-2555.406, Time=14.08 sec\n",
      " ARIMA(0,1,0)(1,0,2)[7]             : AIC=-155.096, Time=5.23 sec\n",
      " ARIMA(0,1,2)(1,0,2)[7]             : AIC=-2486.538, Time=9.27 sec\n",
      " ARIMA(2,1,0)(1,0,2)[7]             : AIC=-1192.020, Time=8.76 sec\n",
      " ARIMA(2,1,2)(1,0,2)[7]             : AIC=-2582.494, Time=16.99 sec\n",
      " ARIMA(2,1,2)(0,0,2)[7]             : AIC=2228.253, Time=9.33 sec\n",
      " ARIMA(2,1,2)(1,0,1)[7]             : AIC=-2585.526, Time=7.68 sec\n",
      " ARIMA(2,1,2)(0,0,1)[7]             : AIC=3590.703, Time=5.49 sec\n",
      " ARIMA(2,1,2)(1,0,0)[7]             : AIC=-206.739, Time=4.59 sec\n",
      " ARIMA(2,1,2)(2,0,1)[7]             : AIC=-2585.284, Time=15.17 sec\n",
      " ARIMA(2,1,2)(0,0,0)[7]             : AIC=6584.674, Time=1.84 sec\n",
      " ARIMA(2,1,2)(2,0,0)[7]             : AIC=-1385.690, Time=9.36 sec\n",
      " ARIMA(2,1,2)(2,0,2)[7]             : AIC=inf, Time=18.63 sec\n",
      " ARIMA(1,1,2)(1,0,1)[7]             : AIC=-2527.927, Time=6.48 sec\n",
      "\n",
      "Best model:  ARIMA(2,1,2)(1,0,1)[7]          \n",
      "Total fit time: 791.587 seconds\n",
      "Performing stepwise search to minimize aic\n",
      " ARIMA(1,1,1)(0,0,0)[7] intercept   : AIC=8652.591, Time=3.48 sec\n",
      " ARIMA(0,1,0)(0,0,0)[7] intercept   : AIC=9211.875, Time=0.74 sec\n",
      " ARIMA(1,1,0)(1,0,0)[7] intercept   : AIC=5440.622, Time=2.97 sec\n",
      " ARIMA(0,1,1)(0,0,1)[7] intercept   : AIC=6788.853, Time=2.32 sec\n",
      " ARIMA(0,1,0)(0,0,0)[7]             : AIC=9209.879, Time=0.10 sec\n",
      " ARIMA(1,1,0)(0,0,0)[7] intercept   : AIC=9213.413, Time=0.44 sec\n",
      " ARIMA(1,1,0)(2,0,0)[7] intercept   : AIC=5441.940, Time=6.41 sec\n",
      " ARIMA(1,1,0)(1,0,1)[7] intercept   : AIC=5442.179, Time=3.35 sec\n",
      " ARIMA(1,1,0)(0,0,1)[7] intercept   : AIC=6810.885, Time=1.68 sec\n",
      " ARIMA(1,1,0)(2,0,1)[7] intercept   : AIC=5351.482, Time=12.77 sec\n",
      " ARIMA(1,1,0)(2,0,2)[7] intercept   : AIC=5125.780, Time=11.87 sec\n",
      " ARIMA(1,1,0)(1,0,2)[7] intercept   : AIC=5277.043, Time=7.71 sec\n",
      " ARIMA(0,1,0)(2,0,2)[7] intercept   : AIC=5373.764, Time=11.15 sec\n",
      " ARIMA(2,1,0)(2,0,2)[7] intercept   : AIC=5070.470, Time=16.47 sec\n",
      " ARIMA(2,1,0)(1,0,2)[7] intercept   : AIC=5232.289, Time=9.03 sec\n",
      " ARIMA(2,1,0)(2,0,1)[7] intercept   : AIC=5295.699, Time=16.76 sec\n",
      " ARIMA(2,1,0)(1,0,1)[7] intercept   : AIC=5398.041, Time=4.80 sec\n",
      " ARIMA(2,1,1)(2,0,2)[7] intercept   : AIC=4446.700, Time=32.82 sec\n",
      " ARIMA(2,1,1)(1,0,2)[7] intercept   : AIC=4641.930, Time=18.44 sec\n",
      " ARIMA(2,1,1)(2,0,1)[7] intercept   : AIC=4690.755, Time=30.86 sec\n",
      " ARIMA(2,1,1)(1,0,1)[7] intercept   : AIC=4805.982, Time=10.16 sec\n",
      " ARIMA(1,1,1)(2,0,2)[7] intercept   : AIC=4454.996, Time=31.23 sec\n",
      " ARIMA(2,1,2)(2,0,2)[7] intercept   : AIC=4406.695, Time=37.40 sec\n",
      " ARIMA(2,1,2)(1,0,2)[7] intercept   : AIC=4419.357, Time=34.69 sec\n",
      " ARIMA(2,1,2)(2,0,1)[7] intercept   : AIC=4588.460, Time=30.74 sec\n",
      " ARIMA(2,1,2)(1,0,1)[7] intercept   : AIC=4562.193, Time=16.57 sec\n",
      " ARIMA(1,1,2)(2,0,2)[7] intercept   : AIC=4451.831, Time=34.01 sec\n",
      " ARIMA(2,1,2)(2,0,2)[7]             : AIC=4293.892, Time=15.22 sec\n",
      " ARIMA(2,1,2)(1,0,2)[7]             : AIC=4417.220, Time=11.80 sec\n",
      " ARIMA(2,1,2)(2,0,1)[7]             : AIC=4583.401, Time=15.06 sec\n",
      " ARIMA(2,1,2)(1,0,1)[7]             : AIC=4557.597, Time=5.27 sec\n",
      " ARIMA(1,1,2)(2,0,2)[7]             : AIC=4449.718, Time=9.09 sec\n",
      " ARIMA(2,1,1)(2,0,2)[7]             : AIC=4444.695, Time=10.10 sec\n",
      " ARIMA(1,1,1)(2,0,2)[7]             : AIC=4452.995, Time=11.83 sec\n",
      "\n",
      "Best model:  ARIMA(2,1,2)(2,0,2)[7]          \n",
      "Total fit time: 467.382 seconds\n",
      "Performing stepwise search to minimize aic\n",
      " ARIMA(1,1,1)(0,0,0)[7] intercept   : AIC=8651.519, Time=4.67 sec\n",
      " ARIMA(0,1,0)(0,0,0)[7] intercept   : AIC=9209.643, Time=0.91 sec\n",
      " ARIMA(1,1,0)(1,0,0)[7] intercept   : AIC=5445.228, Time=3.12 sec\n",
      " ARIMA(0,1,1)(0,0,1)[7] intercept   : AIC=6789.591, Time=2.11 sec\n",
      " ARIMA(0,1,0)(0,0,0)[7]             : AIC=9207.643, Time=0.10 sec\n",
      " ARIMA(1,1,0)(0,0,0)[7] intercept   : AIC=9211.180, Time=0.93 sec\n",
      " ARIMA(1,1,0)(2,0,0)[7] intercept   : AIC=5446.521, Time=7.18 sec\n",
      " ARIMA(1,1,0)(1,0,1)[7] intercept   : AIC=5446.770, Time=3.99 sec\n",
      " ARIMA(1,1,0)(0,0,1)[7] intercept   : AIC=6811.554, Time=1.63 sec\n",
      " ARIMA(1,1,0)(2,0,1)[7] intercept   : AIC=5355.458, Time=12.38 sec\n",
      " ARIMA(1,1,0)(2,0,2)[7] intercept   : AIC=5129.383, Time=11.79 sec\n",
      " ARIMA(1,1,0)(1,0,2)[7] intercept   : AIC=5281.568, Time=8.44 sec\n",
      " ARIMA(0,1,0)(2,0,2)[7] intercept   : AIC=5377.195, Time=11.82 sec\n",
      " ARIMA(2,1,0)(2,0,2)[7] intercept   : AIC=5074.116, Time=21.17 sec\n",
      " ARIMA(2,1,0)(1,0,2)[7] intercept   : AIC=5236.856, Time=10.45 sec\n",
      " ARIMA(2,1,0)(2,0,1)[7] intercept   : AIC=5299.707, Time=16.40 sec\n",
      " ARIMA(2,1,0)(1,0,1)[7] intercept   : AIC=5402.668, Time=4.84 sec\n",
      " ARIMA(2,1,1)(2,0,2)[7] intercept   : AIC=4450.747, Time=30.00 sec\n",
      " ARIMA(2,1,1)(1,0,2)[7] intercept   : AIC=4646.855, Time=23.94 sec\n",
      " ARIMA(2,1,1)(2,0,1)[7] intercept   : AIC=4680.728, Time=29.91 sec\n",
      " ARIMA(2,1,1)(1,0,1)[7] intercept   : AIC=4810.938, Time=12.56 sec\n",
      " ARIMA(1,1,1)(2,0,2)[7] intercept   : AIC=4458.989, Time=23.08 sec\n",
      " ARIMA(2,1,2)(2,0,2)[7] intercept   : AIC=4409.674, Time=41.85 sec\n",
      " ARIMA(2,1,2)(1,0,2)[7] intercept   : AIC=4424.261, Time=40.68 sec\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " ARIMA(2,1,2)(2,0,1)[7] intercept   : AIC=4590.451, Time=34.35 sec\n",
      " ARIMA(2,1,2)(1,0,1)[7] intercept   : AIC=4565.072, Time=17.30 sec\n",
      " ARIMA(1,1,2)(2,0,2)[7] intercept   : AIC=4456.105, Time=33.29 sec\n",
      " ARIMA(2,1,2)(2,0,2)[7]             : AIC=4298.112, Time=14.19 sec\n",
      " ARIMA(2,1,2)(1,0,2)[7]             : AIC=4422.177, Time=14.42 sec\n",
      " ARIMA(2,1,2)(2,0,1)[7]             : AIC=4588.043, Time=15.24 sec\n",
      " ARIMA(2,1,2)(1,0,1)[7]             : AIC=4562.601, Time=5.29 sec\n",
      " ARIMA(1,1,2)(2,0,2)[7]             : AIC=4453.708, Time=11.14 sec\n",
      " ARIMA(2,1,1)(2,0,2)[7]             : AIC=4448.711, Time=10.04 sec\n",
      " ARIMA(1,1,1)(2,0,2)[7]             : AIC=4456.967, Time=11.54 sec\n",
      "\n",
      "Best model:  ARIMA(2,1,2)(2,0,2)[7]          \n",
      "Total fit time: 490.828 seconds\n"
     ]
    },
    {
     "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>Total stock</th>\n",
       "      <th>M3 (HK$ million)</th>\n",
       "      <th>M3 (HK$ million)_Lag1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>0.997198</td>\n",
       "      <td>-2.682884</td>\n",
       "      <td>-2.086517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>1.008067</td>\n",
       "      <td>-2.544874</td>\n",
       "      <td>-2.269493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>1.000226</td>\n",
       "      <td>-2.385175</td>\n",
       "      <td>-2.293428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>0.185239</td>\n",
       "      <td>-1.503105</td>\n",
       "      <td>-2.221118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>0.176910</td>\n",
       "      <td>-1.511373</td>\n",
       "      <td>-1.332827</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Total stock  M3 (HK$ million)  M3 (HK$ million)_Lag1\n",
       "2019-11-27     0.997198         -2.682884              -2.086517\n",
       "2019-11-28     1.008067         -2.544874              -2.269493\n",
       "2019-11-29     1.000226         -2.385175              -2.293428\n",
       "2019-11-30     0.185239         -1.503105              -2.221118\n",
       "2019-12-01     0.176910         -1.511373              -1.332827"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Driver Forecast using SARIMA\n",
    "\n",
    "forecast_index = pd.date_range('2019/11/27', '2020/11/25')\n",
    "\n",
    "driver_forecasts = pd.DataFrame(index=forecast_index, columns= features_final.columns)\n",
    "\n",
    "scaled_AI_Model_Features_1 = scaled_AI_Model_Features.copy()\n",
    "date_index = pd.date_range('2003-01-06', '2019/11/26')\n",
    "scaled_AI_Model_Features_1.index = date_index\n",
    "\n",
    "\n",
    "for i in range(len(features_final.columns)):\n",
    "\n",
    "\n",
    "                                            df_series = scaled_AI_Model_Features_1.asfreq('D')\n",
    "                                            series_features = pd.Series(df_series.iloc[:,i], index= df_series.index)\n",
    "\n",
    "                                            stepwise_fit_features = pm.auto_arima(series_features, start_p=1, start_q=1, max_p=2, max_d=2,max_q=2,m=7,\n",
    "                                                                         start_P=0,start_Q=0,max_P=2, max_D=2,max_Q=2, seasonal=True, trace=True,\n",
    "                                                                         error_action='ignore',  # don't want to know if an order does not work\n",
    "                                                                         suppress_warnings=True,  # don't want convergence warnings\n",
    "                                                                         stepwise=True)  # set to stepwise\n",
    "\n",
    "                                            SARIMA_Forecast_features = stepwise_fit_features.predict(n_periods= 365)\n",
    "                                            #Add to driver forecast dataset\n",
    "                                            driver_forecasts.iloc[:,i] = SARIMA_Forecast_features\n",
    "\n",
    "driver_forecasts.head(5)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Multivariate Time Series"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing stepwise search to minimize aic\n",
      " ARIMA(1,0,1)(0,0,0)[7] intercept   : AIC=16602.803, Time=3.10 sec\n",
      " ARIMA(0,0,0)(0,0,0)[7] intercept   : AIC=17510.864, Time=0.11 sec\n",
      " ARIMA(1,0,0)(1,0,0)[7] intercept   : AIC=17167.608, Time=1.02 sec\n",
      " ARIMA(0,0,1)(0,0,1)[7] intercept   : AIC=17275.372, Time=1.16 sec\n",
      " ARIMA(0,0,0)(0,0,0)[7]             : AIC=17508.864, Time=0.06 sec\n",
      " ARIMA(1,0,1)(1,0,0)[7] intercept   : AIC=17166.628, Time=2.44 sec\n",
      " ARIMA(1,0,1)(0,0,1)[7] intercept   : AIC=17276.799, Time=2.71 sec\n",
      " ARIMA(1,0,1)(1,0,1)[7] intercept   : AIC=16441.930, Time=5.55 sec\n",
      " ARIMA(1,0,1)(2,0,1)[7] intercept   : AIC=16424.847, Time=13.36 sec\n",
      " ARIMA(1,0,1)(2,0,0)[7] intercept   : AIC=16839.488, Time=7.39 sec\n",
      " ARIMA(1,0,1)(3,0,1)[7] intercept   : AIC=16419.292, Time=23.24 sec\n",
      " ARIMA(1,0,1)(3,0,0)[7] intercept   : AIC=16591.933, Time=14.88 sec\n",
      " ARIMA(1,0,1)(3,0,2)[7] intercept   : AIC=16417.786, Time=61.29 sec\n",
      " ARIMA(1,0,1)(2,0,2)[7] intercept   : AIC=16335.456, Time=28.18 sec\n",
      " ARIMA(1,0,1)(1,0,2)[7] intercept   : AIC=16422.738, Time=17.26 sec\n",
      " ARIMA(1,0,1)(2,0,3)[7] intercept   : AIC=16421.395, Time=40.43 sec\n",
      " ARIMA(1,0,1)(1,0,3)[7] intercept   : AIC=16416.132, Time=42.68 sec\n",
      " ARIMA(1,0,1)(3,0,3)[7] intercept   : AIC=inf, Time=nan sec\n",
      " ARIMA(0,0,1)(2,0,2)[7] intercept   : AIC=16399.441, Time=22.91 sec\n",
      " ARIMA(1,0,0)(2,0,2)[7] intercept   : AIC=16487.217, Time=24.19 sec\n",
      " ARIMA(2,0,1)(2,0,2)[7] intercept   : AIC=16330.912, Time=25.76 sec\n",
      " ARIMA(2,0,1)(1,0,2)[7] intercept   : AIC=16395.010, Time=20.15 sec\n",
      " ARIMA(2,0,1)(2,0,1)[7] intercept   : AIC=16397.154, Time=15.30 sec\n",
      " ARIMA(2,0,1)(3,0,2)[7] intercept   : AIC=16370.288, Time=71.99 sec\n",
      " ARIMA(2,0,1)(2,0,3)[7] intercept   : AIC=16322.364, Time=82.14 sec\n",
      " ARIMA(2,0,1)(1,0,3)[7] intercept   : AIC=16388.333, Time=37.43 sec\n",
      " ARIMA(2,0,1)(3,0,3)[7] intercept   : AIC=inf, Time=71.85 sec\n",
      " ARIMA(2,0,0)(2,0,3)[7] intercept   : AIC=16369.309, Time=56.43 sec\n",
      " ARIMA(3,0,1)(2,0,3)[7] intercept   : AIC=16340.577, Time=150.66 sec\n",
      " ARIMA(2,0,2)(2,0,3)[7] intercept   : AIC=16091.748, Time=179.32 sec\n",
      " ARIMA(2,0,2)(1,0,3)[7] intercept   : AIC=15981.074, Time=157.63 sec\n",
      " ARIMA(2,0,2)(0,0,3)[7] intercept   : AIC=15825.864, Time=165.32 sec\n",
      " ARIMA(2,0,2)(0,0,2)[7] intercept   : AIC=15736.577, Time=72.19 sec\n",
      " ARIMA(2,0,2)(0,0,1)[7] intercept   : AIC=15752.752, Time=40.63 sec\n",
      " ARIMA(2,0,2)(1,0,2)[7] intercept   : AIC=16084.531, Time=80.99 sec\n",
      " ARIMA(2,0,2)(1,0,1)[7] intercept   : AIC=16079.304, Time=45.62 sec\n",
      " ARIMA(1,0,2)(0,0,2)[7] intercept   : AIC=17099.454, Time=24.55 sec\n",
      " ARIMA(2,0,1)(0,0,2)[7] intercept   : AIC=17097.480, Time=20.76 sec\n",
      " ARIMA(3,0,2)(0,0,2)[7] intercept   : AIC=16311.769, Time=75.82 sec\n",
      " ARIMA(2,0,3)(0,0,2)[7] intercept   : AIC=16100.841, Time=81.05 sec\n",
      " ARIMA(1,0,1)(0,0,2)[7] intercept   : AIC=17099.360, Time=13.82 sec\n",
      " ARIMA(1,0,3)(0,0,2)[7] intercept   : AIC=15713.876, Time=76.76 sec\n",
      " ARIMA(1,0,3)(0,0,1)[7] intercept   : AIC=15710.819, Time=36.79 sec\n",
      " ARIMA(1,0,3)(0,0,0)[7] intercept   : AIC=15714.849, Time=20.29 sec\n",
      " ARIMA(1,0,3)(1,0,1)[7] intercept   : AIC=15888.342, Time=44.67 sec\n",
      " ARIMA(1,0,3)(1,0,0)[7] intercept   : AIC=15709.889, Time=27.19 sec\n",
      " ARIMA(1,0,3)(2,0,0)[7] intercept   : AIC=15699.650, Time=64.29 sec\n",
      " ARIMA(1,0,3)(3,0,0)[7] intercept   : AIC=16564.716, Time=77.33 sec\n",
      " ARIMA(1,0,3)(2,0,1)[7] intercept   : AIC=16402.764, Time=69.51 sec\n",
      " ARIMA(1,0,3)(3,0,1)[7] intercept   : AIC=16396.702, Time=124.12 sec\n",
      " ARIMA(0,0,3)(2,0,0)[7] intercept   : AIC=16822.412, Time=14.05 sec\n",
      " ARIMA(1,0,2)(2,0,0)[7] intercept   : AIC=16825.039, Time=21.32 sec\n",
      " ARIMA(2,0,3)(2,0,0)[7] intercept   : AIC=16642.219, Time=76.58 sec\n",
      " ARIMA(0,0,2)(2,0,0)[7] intercept   : AIC=16824.209, Time=12.00 sec\n",
      " ARIMA(2,0,2)(2,0,0)[7] intercept   : AIC=15546.661, Time=62.44 sec\n",
      " ARIMA(2,0,2)(1,0,0)[7] intercept   : AIC=17036.324, Time=20.16 sec\n",
      " ARIMA(2,0,2)(3,0,0)[7] intercept   : AIC=15844.832, Time=170.36 sec\n",
      " ARIMA(2,0,2)(2,0,1)[7] intercept   : AIC=15718.009, Time=70.70 sec\n",
      " ARIMA(2,0,2)(3,0,1)[7] intercept   : AIC=16032.453, Time=158.06 sec\n",
      " ARIMA(2,0,1)(2,0,0)[7] intercept   : AIC=16814.469, Time=19.55 sec\n",
      " ARIMA(3,0,2)(2,0,0)[7] intercept   : AIC=16720.096, Time=59.58 sec\n",
      " ARIMA(3,0,1)(2,0,0)[7] intercept   : AIC=16802.137, Time=23.56 sec\n",
      " ARIMA(3,0,3)(2,0,0)[7] intercept   : AIC=16725.477, Time=50.96 sec\n",
      " ARIMA(2,0,2)(2,0,0)[7]             : AIC=15538.992, Time=28.44 sec\n",
      " ARIMA(2,0,2)(1,0,0)[7]             : AIC=17034.327, Time=7.88 sec\n",
      " ARIMA(2,0,2)(3,0,0)[7]             : AIC=15523.220, Time=59.34 sec\n",
      " ARIMA(2,0,2)(3,0,1)[7]             : AIC=15527.735, Time=65.89 sec\n",
      " ARIMA(2,0,2)(2,0,1)[7]             : AIC=15485.747, Time=32.68 sec\n",
      " ARIMA(2,0,2)(1,0,1)[7]             : AIC=16238.182, Time=10.11 sec\n",
      " ARIMA(2,0,2)(2,0,2)[7]             : AIC=15467.760, Time=41.70 sec\n",
      " ARIMA(2,0,2)(1,0,2)[7]             : AIC=inf, Time=38.77 sec\n",
      " ARIMA(2,0,2)(3,0,2)[7]             : AIC=15494.359, Time=70.73 sec\n",
      " ARIMA(2,0,2)(2,0,3)[7]             : AIC=15491.240, Time=66.33 sec\n",
      " ARIMA(2,0,2)(1,0,3)[7]             : AIC=15480.879, Time=62.43 sec\n",
      " ARIMA(2,0,2)(3,0,3)[7]             : AIC=inf, Time=122.69 sec\n",
      " ARIMA(1,0,2)(2,0,2)[7]             : AIC=16302.578, Time=35.18 sec\n",
      " ARIMA(2,0,1)(2,0,2)[7]             : AIC=16292.568, Time=34.35 sec\n",
      " ARIMA(3,0,2)(2,0,2)[7]             : AIC=16254.276, Time=45.30 sec\n",
      " ARIMA(2,0,3)(2,0,2)[7]             : AIC=16369.957, Time=44.66 sec\n",
      " ARIMA(1,0,1)(2,0,2)[7]             : AIC=16320.474, Time=27.95 sec\n",
      " ARIMA(1,0,3)(2,0,2)[7]             : AIC=16297.668, Time=41.29 sec\n",
      " ARIMA(3,0,1)(2,0,2)[7]             : AIC=16276.088, Time=41.64 sec\n",
      " ARIMA(3,0,3)(2,0,2)[7]             : AIC=15488.731, Time=51.36 sec\n",
      " ARIMA(2,0,2)(2,0,2)[7] intercept   : AIC=15882.547, Time=98.49 sec\n",
      "\n",
      "Best model:  ARIMA(2,0,2)(2,0,2)[7]          \n",
      "Total fit time: 4171.217 seconds\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>SARIMAX Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>                 <td>y</td>               <th>  No. Observations:  </th>   <td>6169</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(2, 0, 2)x(2, 0, 2, 7)</td> <th>  Log Likelihood     </th> <td>-7724.880</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>                  <td>Sun, 27 Sep 2020</td>        <th>  AIC                </th> <td>15467.760</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                      <td>01:01:34</td>            <th>  BIC                </th> <td>15528.306</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>                        <td>0</td>               <th>  HQIC               </th> <td>15488.756</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                            <td> - 6169</td>            <th>                     </th>     <td> </td>    \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>              <td>opg</td>              <th>                     </th>     <td> </td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>        <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L1</th>    <td>    1.5266</td> <td>    0.010</td> <td>  150.004</td> <td> 0.000</td> <td>    1.507</td> <td>    1.547</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L2</th>    <td>   -0.5291</td> <td>    0.010</td> <td>  -52.027</td> <td> 0.000</td> <td>   -0.549</td> <td>   -0.509</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1</th>    <td>   -1.8482</td> <td>    0.008</td> <td> -217.621</td> <td> 0.000</td> <td>   -1.865</td> <td>   -1.832</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L2</th>    <td>    0.8704</td> <td>    0.008</td> <td>  104.301</td> <td> 0.000</td> <td>    0.854</td> <td>    0.887</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.S.L7</th>  <td>    1.7879</td> <td>    0.068</td> <td>   26.442</td> <td> 0.000</td> <td>    1.655</td> <td>    1.920</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.S.L14</th> <td>   -0.8237</td> <td>    0.061</td> <td>  -13.533</td> <td> 0.000</td> <td>   -0.943</td> <td>   -0.704</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L7</th>  <td>   -1.7557</td> <td>    0.064</td> <td>  -27.511</td> <td> 0.000</td> <td>   -1.881</td> <td>   -1.631</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L14</th> <td>    0.8060</td> <td>    0.054</td> <td>   14.988</td> <td> 0.000</td> <td>    0.701</td> <td>    0.911</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th>   <td>    0.7159</td> <td>    0.003</td> <td>  235.346</td> <td> 0.000</td> <td>    0.710</td> <td>    0.722</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (Q):</th>          <td>84.40</td> <th>  Jarque-Bera (JB):  </th> <td>526608.99</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                <td>0.00</td>  <th>  Prob(JB):          </th>   <td>0.00</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th> <td>1.43</td>  <th>  Skew:              </th>   <td>0.83</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>    <td>0.00</td>  <th>  Kurtosis:          </th>   <td>48.23</td>  \n",
       "</tr>\n",
       "</table><br/><br/>Warnings:<br/>[1] Covariance matrix calculated using the outer product of gradients (complex-step)."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                                     SARIMAX Results                                     \n",
       "=========================================================================================\n",
       "Dep. Variable:                                 y   No. Observations:                 6169\n",
       "Model:             SARIMAX(2, 0, 2)x(2, 0, 2, 7)   Log Likelihood               -7724.880\n",
       "Date:                           Sun, 27 Sep 2020   AIC                          15467.760\n",
       "Time:                                   01:01:34   BIC                          15528.306\n",
       "Sample:                                        0   HQIC                         15488.756\n",
       "                                          - 6169                                         \n",
       "Covariance Type:                             opg                                         \n",
       "==============================================================================\n",
       "                 coef    std err          z      P>|z|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "ar.L1          1.5266      0.010    150.004      0.000       1.507       1.547\n",
       "ar.L2         -0.5291      0.010    -52.027      0.000      -0.549      -0.509\n",
       "ma.L1         -1.8482      0.008   -217.621      0.000      -1.865      -1.832\n",
       "ma.L2          0.8704      0.008    104.301      0.000       0.854       0.887\n",
       "ar.S.L7        1.7879      0.068     26.442      0.000       1.655       1.920\n",
       "ar.S.L14      -0.8237      0.061    -13.533      0.000      -0.943      -0.704\n",
       "ma.S.L7       -1.7557      0.064    -27.511      0.000      -1.881      -1.631\n",
       "ma.S.L14       0.8060      0.054     14.988      0.000       0.701       0.911\n",
       "sigma2         0.7159      0.003    235.346      0.000       0.710       0.722\n",
       "===================================================================================\n",
       "Ljung-Box (Q):                       84.40   Jarque-Bera (JB):            526608.99\n",
       "Prob(Q):                              0.00   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):               1.43   Skew:                             0.83\n",
       "Prob(H) (two-sided):                  0.00   Kurtosis:                        48.23\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fitting a stepwise model to find the best paramters for SARIMAX:\n",
    "stepwise_fit_2 = pm.auto_arima(scaled_AI_Model_Target, start_p=1, start_q=1, max_p=3, d=0,max_q=3,m=7,\n",
    "                             start_P=0,start_Q=0,max_P=3, max_D=3,max_Q=3, seasonal=True, trace=True,\n",
    "                             error_action='ignore',  # don't want to know if an order does not work\n",
    "                             suppress_warnings=True,  # don't want convergence warnings\n",
    "                             stepwise=True)  # set to stepwise\n",
    "\n",
    "stepwise_fit_2.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\statsmodels\\tsa\\statespace\\sarimax.py:963: UserWarning: Non-stationary starting autoregressive parameters found. Using zeros as starting parameters.\n",
      "  warn('Non-stationary starting autoregressive parameters'\n",
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\statsmodels\\tsa\\statespace\\sarimax.py:975: UserWarning: Non-invertible starting MA parameters found. Using zeros as starting parameters.\n",
      "  warn('Non-invertible starting MA parameters found.'\n",
      "C:\\Users\\ibaruah001\\AppData\\Local\\Annaconda\\lib\\site-packages\\statsmodels\\base\\model.py:568: ConvergenceWarning: Maximum Likelihood optimization failed to converge. Check mle_retvals\n",
      "  \"Check mle_retvals\", ConvergenceWarning)\n"
     ]
    },
    {
     "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>Raw_Forecast</th>\n",
       "      <th>SARIMAX_Forecast_De_Standardise_R_Sum</th>\n",
       "      <th>SARIMAX_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>-1.336010</td>\n",
       "      <td>-1.336010</td>\n",
       "      <td>364.474987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>-1.395018</td>\n",
       "      <td>-2.731028</td>\n",
       "      <td>363.079969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>-1.404494</td>\n",
       "      <td>-4.135522</td>\n",
       "      <td>361.675475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>-1.083855</td>\n",
       "      <td>-5.219377</td>\n",
       "      <td>360.591620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>-1.106740</td>\n",
       "      <td>-6.326117</td>\n",
       "      <td>359.484880</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Raw_Forecast  SARIMAX_Forecast_De_Standardise_R_Sum  \\\n",
       "2019-11-27     -1.336010                              -1.336010   \n",
       "2019-11-28     -1.395018                              -2.731028   \n",
       "2019-11-29     -1.404494                              -4.135522   \n",
       "2019-11-30     -1.083855                              -5.219377   \n",
       "2019-12-01     -1.106740                              -6.326117   \n",
       "\n",
       "            SARIMAX_Forecast  \n",
       "2019-11-27        364.474987  \n",
       "2019-11-28        363.079969  \n",
       "2019-11-29        361.675475  \n",
       "2019-11-30        360.591620  \n",
       "2019-12-01        359.484880  "
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SARIMAX \n",
    "# contrived dataset\n",
    "data1 = scaled_AI_Model_Target\n",
    "data2 = scaled_AI_Model_Features\n",
    "# fit model\n",
    "model = SARIMAX(data1, exog=data2, order=stepwise_fit_2.order, seasonal_order=stepwise_fit_2.seasonal_order)\n",
    "model_fit = model.fit(disp=False)\n",
    "# make prediction\n",
    "exog2 = driver_forecasts\n",
    "SARIMAX_Forecast = pd.DataFrame(model_fit.predict(len(data1),len(data1)+364, exog=exog2))\n",
    "SARIMAX_Forecast.columns = ['Raw_Forecast']\n",
    "SARIMAX_Forecast['SARIMAX_Forecast_De_Standardise_R_Sum'] = SARIMAX_Forecast['Raw_Forecast'].cumsum()\n",
    "SARIMAX_Forecast['SARIMAX_Forecast'] = SARIMAX_Forecast['SARIMAX_Forecast_De_Standardise_R_Sum'] + Latest_target[0]\n",
    "\n",
    "SARIMAX_Forecast.index = HWES_Forecast.index\n",
    "SARIMAX_Forecast.head(5)\n",
    "#print(SARIMAX_Forecast)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    Augmented Dickey-Fuller Test on \"Private Domestic (Price Index)\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -3.6108\n",
      " No. Lags Chosen       = 33\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0056. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"Total stock\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -8.7666\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n",
      "    Augmented Dickey-Fuller Test on \"M3 (HK$ million)\" \n",
      "    -----------------------------------------------\n",
      " Null Hypothesis: Data has unit root. Non-Stationary.\n",
      " Significance Level    = 0.05\n",
      " Test Statistic        = -10.1037\n",
      " No. Lags Chosen       = 34\n",
      " Critical value 1%     = -3.431\n",
      " Critical value 5%     = -2.862\n",
      " Critical value 10%    = -2.567\n",
      " => P-Value = 0.0. Rejecting Null Hypothesis.\n",
      " => Series is Stationary.\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Data Prep for VAR/VECM - First Difference and check stationarity\n",
    "\n",
    "# Data Prep for VAR - check stationarity\n",
    "\n",
    "varma_vecm_data = pd.concat([target_final, features_final], axis=1, sort=False)\n",
    "varma_vecm_data  = varma_vecm_data.drop(columns=['M3 (HK$ million)_Lag1'])\n",
    "\n",
    "# 1st difference\n",
    "varma_vecm_data_differenced = varma_vecm_data.diff().dropna()\n",
    "\n",
    "# ADF Test on each column\n",
    "for name, column in varma_vecm_data_differenced.iteritems():\n",
    "    adfuller_test(column, name=column.name)\n",
    "    print('\\n')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "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>VAR_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>365.504389</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>365.220096</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>364.928700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>364.694052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>364.437160</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            VAR_Forecast\n",
       "2019-11-27    365.504389\n",
       "2019-11-28    365.220096\n",
       "2019-11-29    364.928700\n",
       "2019-11-30    364.694052\n",
       "2019-12-01    364.437160"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#VAR Model\n",
    "\n",
    "model = VAR(varma_vecm_data_differenced)\n",
    "lag_selection_VAR = model.select_order(maxlags=12)\n",
    "\n",
    "list(lag_selection_VAR.selected_orders.values())[0]\n",
    "\n",
    "\n",
    "# Get the lag order\n",
    "lag_order = list(lag_selection_VAR.selected_orders.values())[0]\n",
    "\n",
    "model_fitted = model.fit(lag_order)\n",
    "\n",
    "\n",
    "# Input data for forecasting\n",
    "forecast_input = varma_vecm_data_differenced.values[-lag_order:]\n",
    "forecast_input\n",
    "\n",
    "# Forecast\n",
    "VAR_FC = model_fitted.forecast(y=forecast_input, steps= 365)\n",
    "VAR_FC_2 = pd.DataFrame(VAR_FC , index= driver_forecasts.index, columns=varma_vecm_data_differenced.columns + '_1d')\n",
    "VAR_FC_2\n",
    "\n",
    "\n",
    "#VAR Forecast\n",
    "\n",
    "def invert_transformation(df_train, df_forecast, second_diff=False):\n",
    "    \"\"\"Revert back the differencing to get the forecast to original scale.\"\"\"\n",
    "    df_fc = df_forecast.copy()\n",
    "    columns = df_train.columns\n",
    "    for col in columns:        \n",
    "        # Roll back 2nd Diff\n",
    "        if second_diff:\n",
    "            df_fc[str(col)+'_1d'] = (df_train[col].iloc[-1]-df_train[col].iloc[-2]) + df_fc[str(col)+'_2d'].cumsum()\n",
    "        # Roll back 1st Diff\n",
    "        df_fc[str(col)+'_forecast'] = df_train[col].iloc[-1] + df_fc[str(col)+'_1d'].cumsum()\n",
    "    return df_fc\n",
    "\n",
    "df_results = invert_transformation(varma_vecm_data, VAR_FC_2, second_diff=False)\n",
    "VAR_Forecast = pd.DataFrame(df_results['Private Domestic (Price Index)_forecast'])\n",
    "VAR_Forecast.columns = ['VAR_Forecast']\n",
    "VAR_Forecast.index = HWES_Forecast.index\n",
    "VAR_Forecast.head(5)\n"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "# VARMAX\n",
    "    \n",
    "for i,j in zip(range(1), range(1)):\n",
    "                                    model = VARMAX(varma_vecm_data_differenced, order=(i+1, j+1))\n",
    "                                    result = model.fit()\n",
    "                                    print('Lag Order =', i+1)\n",
    "                                    print('MA Order =', j+1)\n",
    "                                    print('AIC : ', result.aic)\n",
    "                                    print('BIC : ', result.bic)\n",
    "                                    print('FPE : ', result.fpe)\n",
    "                                    print('HQIC: ', result.hqic, '\\n')\n",
    "                                    \n",
    "#Based on AIC, seelct orders\n",
    "model = VARMAX(varma_vecm_data_differenced, order=(2,2))\n",
    "result = model.fit()\n",
    "\n",
    "# make prediction\n",
    "VARMAX_Forecast= model_fit.forecast(365)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Johansen cointegration test using trace test statistic with 5% significance level\n",
      "=====================================\n",
      "r_0 r_1 test statistic critical value\n",
      "-------------------------------------\n",
      "  0   3          290.5          24.28\n",
      "  1   3          72.75          12.32\n",
      "  2   3          28.74          4.130\n",
      "-------------------------------------\n"
     ]
    }
   ],
   "source": [
    "# VECM - Johansen Cointegration Test\n",
    "\n",
    "rank_test = select_coint_rank(varma_vecm_data_differenced,-1, lag_order, method=\"trace\",\n",
    "                              signif=0.05)\n",
    "print(rank_test)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "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>VECM_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>365.490734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>365.194382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>364.917107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>364.685567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>364.426202</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            VECM_Forecast\n",
       "2019-11-27     365.490734\n",
       "2019-11-28     365.194382\n",
       "2019-11-29     364.917107\n",
       "2019-11-30     364.685567\n",
       "2019-12-01     364.426202"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#VECM\n",
    "\n",
    "model = VECM(varma_vecm_data_differenced, deterministic=\"ci\",\n",
    "             k_ar_diff=lag_order,  \n",
    "             coint_rank=rank_test.rank)  \n",
    "\n",
    "vecm_res = model.fit()\n",
    "\n",
    "VECM_FC = vecm_res.predict(steps=365)\n",
    "\n",
    "VECM_FC_2 = pd.DataFrame(VECM_FC , index= driver_forecasts.index, columns=varma_vecm_data_differenced.columns + '_1d')\n",
    "\n",
    "df_results_2 = invert_transformation(varma_vecm_data, VECM_FC_2, second_diff=False)\n",
    "\n",
    "VECM_Forecast = pd.DataFrame(df_results_2['Private Domestic (Price Index)_forecast'])\n",
    "VECM_Forecast.columns = ['VECM_Forecast']\n",
    "VECM_Forecast.index = HWES_Forecast.index\n",
    "VECM_Forecast.head(5)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### AI Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyper Parameter Tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "5797/5797 [==============================] - 35s 6ms/step - loss: 0.7575\n",
      "Epoch 2/5\n",
      "5797/5797 [==============================] - 39s 7ms/step - loss: 0.6569: 0s - loss: 0.\n",
      "Epoch 3/5\n",
      "5797/5797 [==============================] - 37s 6ms/step - loss: 0.6477\n",
      "Epoch 4/5\n",
      "5797/5797 [==============================] - 41s 7ms/step - loss: 0.6190\n",
      "Epoch 5/5\n",
      "5797/5797 [==============================] - 41s 7ms/step - loss: 0.6019\n",
      "Epoch 1/5\n",
      "5774/5774 [==============================] - 112s 19ms/step - loss: 6.0559\n",
      "Epoch 2/5\n",
      "5774/5774 [==============================] - 118s 20ms/step - loss: 608.7675- ETA: 1s  - ETA: 0s - loss: 609.\n",
      "Epoch 3/5\n",
      "5774/5774 [==============================] - 104s 18ms/step - loss: 0.7135\n",
      "Epoch 4/5\n",
      "5774/5774 [==============================] - 94s 16ms/step - loss: 0.6407 0s - \n",
      "Epoch 5/5\n",
      "5774/5774 [==============================] - 97s 17ms/step - loss: 0.6868\n",
      "Epoch 1/5\n",
      "5684/5684 [==============================] - 407s 72ms/step - loss: nans -\n",
      "Epoch 2/5\n",
      "5684/5684 [==============================] - 444s 78ms/step - loss: nan\n",
      "Epoch 3/5\n",
      "5684/5684 [==============================] - 458s 81ms/step - loss: nan\n",
      "Epoch 4/5\n",
      "5684/5684 [==============================] - 425s 75ms/step - loss: nan\n",
      "Epoch 5/5\n",
      "5684/5684 [==============================] - 431s 76ms/step - loss: nan\n",
      "Epoch 1/7\n",
      "5797/5797 [==============================] - 36s 6ms/step - loss: 0.7708\n",
      "Epoch 2/7\n",
      "5797/5797 [==============================] - 34s 6ms/step - loss: 0.6640\n",
      "Epoch 3/7\n",
      "5797/5797 [==============================] - 33s 6ms/step - loss: 0.6388\n",
      "Epoch 4/7\n",
      "5797/5797 [==============================] - 39s 7ms/step - loss: 0.6183\n",
      "Epoch 5/7\n",
      "5797/5797 [==============================] - 32s 6ms/step - loss: 0.5992\n",
      "Epoch 6/7\n",
      "5797/5797 [==============================] - ETA: 0s - loss: 0.589 - 32s 5ms/step - loss: 0.5885\n",
      "Epoch 7/7\n",
      "5797/5797 [==============================] - 37s 6ms/step - loss: 0.5830\n",
      "Epoch 1/7\n",
      "5774/5774 [==============================] - 102s 18ms/step - loss: 361293.7188\n",
      "Epoch 2/7\n",
      "5774/5774 [==============================] - 113s 20ms/step - loss: 0.7877\n",
      "Epoch 3/7\n",
      "5774/5774 [==============================] - 138s 24ms/step - loss: 0.7223\n",
      "Epoch 4/7\n",
      "5774/5774 [==============================] - 119s 21ms/step - loss: 0.6740\n",
      "Epoch 5/7\n",
      "5774/5774 [==============================] - 110s 19ms/step - loss: 0.6425\n",
      "Epoch 6/7\n",
      "5774/5774 [==============================] - 96s 17ms/step - loss: 0.6327\n",
      "Epoch 7/7\n",
      "5774/5774 [==============================] - 100s 17ms/step - loss: 0.6444\n",
      "Epoch 1/7\n",
      "5684/5684 [==============================] - 407s 72ms/step - loss: nan\n",
      "Epoch 2/7\n",
      "5684/5684 [==============================] - 558s 98ms/step - loss: nan\n",
      "Epoch 3/7\n",
      "5684/5684 [==============================] - 447s 79ms/step - loss: nan\n",
      "Epoch 4/7\n",
      "5684/5684 [==============================] - 368s 65ms/step - loss: nan\n",
      "Epoch 5/7\n",
      "5684/5684 [==============================] - 528s 93ms/step - loss: nan\n",
      "Epoch 6/7\n",
      "5684/5684 [==============================] - 419s 74ms/step - loss: nan\n",
      "Epoch 7/7\n",
      "5684/5684 [==============================] - 389s 68ms/step - loss: nan\n",
      "Epoch 1/10\n",
      "5797/5797 [==============================] - 32s 5ms/step - loss: 1.2709\n",
      "Epoch 2/10\n",
      "5797/5797 [==============================] - 34s 6ms/step - loss: 0.7010\n",
      "Epoch 3/10\n",
      "5797/5797 [==============================] - 30s 5ms/step - loss: 0.6681\n",
      "Epoch 4/10\n",
      "5797/5797 [==============================] - 32s 5ms/step - loss: 0.6288\n",
      "Epoch 5/10\n",
      "5797/5797 [==============================] - 31s 5ms/step - loss: 0.6159\n",
      "Epoch 6/10\n",
      "5797/5797 [==============================] - 31s 5ms/step - loss: 0.6061\n",
      "Epoch 7/10\n",
      "5797/5797 [==============================] - 33s 6ms/step - loss: 0.5741\n",
      "Epoch 8/10\n",
      "5797/5797 [==============================] - 34s 6ms/step - loss: 0.5862\n",
      "Epoch 9/10\n",
      "5797/5797 [==============================] - 31s 5ms/step - loss: 0.5758\n",
      "Epoch 10/10\n",
      "5797/5797 [==============================] - 31s 5ms/step - loss: 0.5621\n",
      "Epoch 1/10\n",
      "5774/5774 [==============================] - 98s 17ms/step - loss: 4.5566\n",
      "Epoch 2/10\n",
      "5774/5774 [==============================] - 93s 16ms/step - loss: 611.4766\n",
      "Epoch 3/10\n",
      "5774/5774 [==============================] - 88s 15ms/step - loss: 1.0208\n",
      "Epoch 4/10\n",
      "5774/5774 [==============================] - 101s 17ms/step - loss: 35499.5352\n",
      "Epoch 5/10\n",
      "5774/5774 [==============================] - 86s 15ms/step - loss: 189111.4688\n",
      "Epoch 6/10\n",
      "5774/5774 [==============================] - 94s 16ms/step - loss: 0.6989\n",
      "Epoch 7/10\n",
      "5774/5774 [==============================] - 88s 15ms/step - loss: 0.6304\n",
      "Epoch 8/10\n",
      "5774/5774 [==============================] - 94s 16ms/step - loss: 0.6129\n",
      "Epoch 9/10\n",
      "5774/5774 [==============================] - 87s 15ms/step - loss: 6852.1753\n",
      "Epoch 10/10\n",
      "5774/5774 [==============================] - 86s 15ms/step - loss: 0.6265\n",
      "Epoch 1/10\n",
      "5684/5684 [==============================] - 352s 62ms/step - loss: nan\n",
      "Epoch 2/10\n",
      "5684/5684 [==============================] - 382s 67ms/step - loss: nan\n",
      "Epoch 3/10\n",
      "5684/5684 [==============================] - 345s 61ms/step - loss: nan\n",
      "Epoch 4/10\n",
      "5684/5684 [==============================] - 352s 62ms/step - loss: nan\n",
      "Epoch 5/10\n",
      "5684/5684 [==============================] - 516s 91ms/step - loss: nan\n",
      "Epoch 6/10\n",
      "5684/5684 [==============================] - 360s 63ms/step - loss: nan\n",
      "Epoch 7/10\n",
      "5684/5684 [==============================] - 333s 59ms/step - loss: nan\n",
      "Epoch 8/10\n",
      "5684/5684 [==============================] - 322s 57ms/step - loss: nan\n",
      "Epoch 9/10\n",
      "5684/5684 [==============================] - 353s 62ms/step - loss: nan\n",
      "Epoch 10/10\n",
      "5684/5684 [==============================] - 357s 63ms/step - loss: nan\n",
      "Epoch 1/5\n",
      "2899/2899 [==============================] - 22s 7ms/step - loss: 0.7849: 0s - l\n",
      "Epoch 2/5\n",
      "2899/2899 [==============================] - 23s 8ms/step - loss: 0.6898\n",
      "Epoch 3/5\n",
      "2899/2899 [==============================] - 23s 8ms/step - loss: 0.6437\n",
      "Epoch 4/5\n",
      "2899/2899 [==============================] - 22s 8ms/step - loss: 0.6277\n",
      "Epoch 5/5\n",
      "2899/2899 [==============================] - 23s 8ms/step - loss: 0.5980\n",
      "Epoch 1/5\n",
      "2887/2887 [==============================] - 119s 41ms/step - loss: 5.2878\n",
      "Epoch 2/5\n",
      "2887/2887 [==============================] - 118s 41ms/step - loss: 0.7385\n",
      "Epoch 3/5\n",
      "2887/2887 [==============================] - 113s 39ms/step - loss: 0.6570\n",
      "Epoch 4/5\n",
      "2887/2887 [==============================] - 112s 39ms/step - loss: 0.6228\n",
      "Epoch 5/5\n",
      "2887/2887 [==============================] - 112s 39ms/step - loss: 0.6312\n",
      "Epoch 1/5\n",
      "2842/2842 [==============================] - 453s 160ms/step - loss: nan\n",
      "Epoch 2/5\n",
      "2842/2842 [==============================] - 471s 166ms/step - loss: nan\n",
      "Epoch 3/5\n",
      "2842/2842 [==============================] - 485s 170ms/step - loss: nan\n",
      "Epoch 4/5\n",
      "2842/2842 [==============================] - 485s 171ms/step - loss: nan\n",
      "Epoch 5/5\n",
      "2842/2842 [==============================] - 467s 164ms/step - loss: nan\n",
      "Epoch 1/7\n",
      "2899/2899 [==============================] - 34s 12ms/step - loss: 0.7854\n",
      "Epoch 2/7\n",
      "2899/2899 [==============================] - 36s 12ms/step - loss: 0.6642\n",
      "Epoch 3/7\n",
      "2899/2899 [==============================] - 35s 12ms/step - loss: 0.6484\n",
      "Epoch 4/7\n",
      "2899/2899 [==============================] - 36s 12ms/step - loss: 0.6123\n",
      "Epoch 5/7\n",
      "2899/2899 [==============================] - 38s 13ms/step - loss: 0.6056\n",
      "Epoch 6/7\n",
      "2899/2899 [==============================] - 34s 12ms/step - loss: 0.5890\n",
      "Epoch 7/7\n",
      "2899/2899 [==============================] - 34s 12ms/step - loss: 0.5753\n",
      "Epoch 1/7\n",
      "2887/2887 [==============================] - 122s 42ms/step - loss: 707.1302\n",
      "Epoch 2/7\n",
      "2887/2887 [==============================] - 141s 49ms/step - loss: 46.5285\n",
      "Epoch 3/7\n",
      "2887/2887 [==============================] - 131s 45ms/step - loss: 13604.7256\n",
      "Epoch 4/7\n",
      "2887/2887 [==============================] - 125s 43ms/step - loss: 172.6297\n",
      "Epoch 5/7\n",
      "2887/2887 [==============================] - 70s 24ms/step - loss: 0.7529\n",
      "Epoch 6/7\n",
      "2887/2887 [==============================] - 84s 29ms/step - loss: 0.7304 1s - lo\n",
      "Epoch 7/7\n",
      "2887/2887 [==============================] - 77s 27ms/step - loss: 0.7004\n",
      "Epoch 1/7\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2842/2842 [==============================] - 283s 100ms/step - loss: nan\n",
      "Epoch 2/7\n",
      "2842/2842 [==============================] - 294s 103ms/step - loss: nan\n",
      "Epoch 3/7\n",
      "2842/2842 [==============================] - 281s 99ms/step - loss: nan\n",
      "Epoch 4/7\n",
      "2842/2842 [==============================] - 263s 93ms/step - loss: nan\n",
      "Epoch 5/7\n",
      "2842/2842 [==============================] - 242s 85ms/step - loss: nan\n",
      "Epoch 6/7\n",
      "2842/2842 [==============================] - 256s 90ms/step - loss: nan\n",
      "Epoch 7/7\n",
      "2842/2842 [==============================] - 244s 86ms/step - loss: nan\n",
      "Epoch 1/10\n",
      "2899/2899 [==============================] - 19s 6ms/step - loss: 0.7854\n",
      "Epoch 2/10\n",
      "2899/2899 [==============================] - 19s 7ms/step - loss: 0.6609\n",
      "Epoch 3/10\n",
      "2899/2899 [==============================] - 19s 7ms/step - loss: 0.6537\n",
      "Epoch 4/10\n",
      "2899/2899 [==============================] - 18s 6ms/step - loss: 0.6230\n",
      "Epoch 5/10\n",
      "2899/2899 [==============================] - 19s 7ms/step - loss: 0.6010\n",
      "Epoch 6/10\n",
      "2899/2899 [==============================] - 20s 7ms/step - loss: 0.5858\n",
      "Epoch 7/10\n",
      "2899/2899 [==============================] - 25s 9ms/step - loss: 0.5769\n",
      "Epoch 8/10\n",
      "2899/2899 [==============================] - 20s 7ms/step - loss: 0.5620\n",
      "Epoch 9/10\n",
      "2899/2899 [==============================] - 24s 8ms/step - loss: 0.5596\n",
      "Epoch 10/10\n",
      "2899/2899 [==============================] - 23s 8ms/step - loss: 0.5431\n",
      "Epoch 1/10\n",
      "2887/2887 [==============================] - 71s 25ms/step - loss: 147.3299\n",
      "Epoch 2/10\n",
      "2887/2887 [==============================] - 65s 23ms/step - loss: 0.8121\n",
      "Epoch 3/10\n",
      "2887/2887 [==============================] - 62s 21ms/step - loss: 0.7054\n",
      "Epoch 4/10\n",
      "2887/2887 [==============================] - 62s 21ms/step - loss: 0.6323\n",
      "Epoch 5/10\n",
      "2887/2887 [==============================] - 61s 21ms/step - loss: 0.6401\n",
      "Epoch 6/10\n",
      "2887/2887 [==============================] - 64s 22ms/step - loss: 0.6195\n",
      "Epoch 7/10\n",
      "2887/2887 [==============================] - 75s 26ms/step - loss: 0.6048\n",
      "Epoch 8/10\n",
      "2887/2887 [==============================] - 72s 25ms/step - loss: 0.8325\n",
      "Epoch 9/10\n",
      "2887/2887 [==============================] - 77s 27ms/step - loss: 0.6102\n",
      "Epoch 10/10\n",
      "2887/2887 [==============================] - 90s 31ms/step - loss: 0.5865 0s - loss: 0.5 - ETA: 0s - loss: 0.\n",
      "Epoch 1/10\n",
      "2842/2842 [==============================] - 324s 114ms/step - loss: nan\n",
      "Epoch 2/10\n",
      "2842/2842 [==============================] - 279s 98ms/step - loss: nan\n",
      "Epoch 3/10\n",
      "2842/2842 [==============================] - 260s 91ms/step - loss: nan\n",
      "Epoch 4/10\n",
      "2842/2842 [==============================] - 270s 95ms/step - loss: nan\n",
      "Epoch 5/10\n",
      "2842/2842 [==============================] - 263s 93ms/step - loss: nan\n",
      "Epoch 6/10\n",
      "2842/2842 [==============================] - 250s 88ms/step - loss: nan\n",
      "Epoch 7/10\n",
      "2842/2842 [==============================] - 253s 89ms/step - loss: nan\n",
      "Epoch 8/10\n",
      "2842/2842 [==============================] - 250s 88ms/step - loss: nan\n",
      "Epoch 9/10\n",
      "2842/2842 [==============================] - 249s 88ms/step - loss: nan\n",
      "Epoch 10/10\n",
      "2842/2842 [==============================] - 248s 87ms/step - loss: nan\n",
      "Epoch 1/5\n",
      "1450/1450 [==============================] - 10s 7ms/step - loss: 0.7975\n",
      "Epoch 2/5\n",
      "1450/1450 [==============================] - 10s 7ms/step - loss: 0.6915\n",
      "Epoch 3/5\n",
      "1450/1450 [==============================] - 11s 7ms/step - loss: 0.6463\n",
      "Epoch 4/5\n",
      "1450/1450 [==============================] - 10s 7ms/step - loss: 0.6241\n",
      "Epoch 5/5\n",
      "1450/1450 [==============================] - 10s 7ms/step - loss: 0.6299\n",
      "Epoch 1/5\n",
      "1444/1444 [==============================] - 33s 23ms/step - loss: 6.5362\n",
      "Epoch 2/5\n",
      "1444/1444 [==============================] - 33s 23ms/step - loss: 2.7125\n",
      "Epoch 3/5\n",
      "1444/1444 [==============================] - 33s 23ms/step - loss: 0.7529\n",
      "Epoch 4/5\n",
      "1444/1444 [==============================] - 33s 23ms/step - loss: 0.8233 0s -\n",
      "Epoch 5/5\n",
      "1444/1444 [==============================] - 35s 24ms/step - loss: 0.7470\n",
      "Epoch 1/5\n",
      "1421/1421 [==============================] - 130s 92ms/step - loss: nan\n",
      "Epoch 2/5\n",
      "1421/1421 [==============================] - 128s 90ms/step - loss: nan\n",
      "Epoch 3/5\n",
      "1421/1421 [==============================] - 129s 91ms/step - loss: nan\n",
      "Epoch 4/5\n",
      "1421/1421 [==============================] - 128s 90ms/step - loss: nan\n",
      "Epoch 5/5\n",
      "1421/1421 [==============================] - 161s 113ms/step - loss: nan\n",
      "Epoch 1/7\n",
      "1450/1450 [==============================] - 12s 8ms/step - loss: 0.7986\n",
      "Epoch 2/7\n",
      "1450/1450 [==============================] - 13s 9ms/step - loss: 0.6728\n",
      "Epoch 3/7\n",
      "1450/1450 [==============================] - 12s 8ms/step - loss: 0.6563\n",
      "Epoch 4/7\n",
      "1450/1450 [==============================] - 11s 8ms/step - loss: 0.6321\n",
      "Epoch 5/7\n",
      "1450/1450 [==============================] - 12s 8ms/step - loss: 0.6154\n",
      "Epoch 6/7\n",
      "1450/1450 [==============================] - 12s 8ms/step - loss: 0.5915\n",
      "Epoch 7/7\n",
      "1450/1450 [==============================] - 11s 8ms/step - loss: 0.5836\n",
      "Epoch 1/7\n",
      "1444/1444 [==============================] - 36s 25ms/step - loss: 0.8038\n",
      "Epoch 2/7\n",
      "1444/1444 [==============================] - 35s 24ms/step - loss: 619.5538\n",
      "Epoch 3/7\n",
      "1444/1444 [==============================] - 38s 27ms/step - loss: 2.2557\n",
      "Epoch 4/7\n",
      "1444/1444 [==============================] - 40s 28ms/step - loss: 0.8178\n",
      "Epoch 5/7\n",
      "1444/1444 [==============================] - 46s 32ms/step - loss: 0.7891\n",
      "Epoch 6/7\n",
      "1444/1444 [==============================] - 49s 34ms/step - loss: 2482.4414\n",
      "Epoch 7/7\n",
      "1444/1444 [==============================] - 48s 33ms/step - loss: 0.7828\n",
      "Epoch 1/7\n",
      "1421/1421 [==============================] - 301s 212ms/step - loss: 169459632.0000\n",
      "Epoch 2/7\n",
      "1421/1421 [==============================] - 289s 203ms/step - loss: 0.8701\n",
      "Epoch 3/7\n",
      "1421/1421 [==============================] - 341s 240ms/step - loss: 0.8541\n",
      "Epoch 4/7\n",
      "1421/1421 [==============================] - 224s 157ms/step - loss: 0.8385\n",
      "Epoch 5/7\n",
      "1421/1421 [==============================] - 195s 137ms/step - loss: nan\n",
      "Epoch 6/7\n",
      "1421/1421 [==============================] - 176s 124ms/step - loss: nan\n",
      "Epoch 7/7\n",
      "1421/1421 [==============================] - 179s 126ms/step - loss: nan\n",
      "Epoch 1/10\n",
      "1450/1450 [==============================] - 16s 11ms/step - loss: 0.7747\n",
      "Epoch 2/10\n",
      "1450/1450 [==============================] - 16s 11ms/step - loss: 0.6676\n",
      "Epoch 3/10\n",
      "1450/1450 [==============================] - 15s 11ms/step - loss: 0.6592\n",
      "Epoch 4/10\n",
      "1450/1450 [==============================] - 15s 10ms/step - loss: 0.6069\n",
      "Epoch 5/10\n",
      "1450/1450 [==============================] - 15s 10ms/step - loss: 0.6142\n",
      "Epoch 6/10\n",
      "1450/1450 [==============================] - 15s 10ms/step - loss: 0.5983\n",
      "Epoch 7/10\n",
      "1450/1450 [==============================] - 15s 11ms/step - loss: 0.5792\n",
      "Epoch 8/10\n",
      "1450/1450 [==============================] - 14s 10ms/step - loss: 0.5669\n",
      "Epoch 9/10\n",
      "1450/1450 [==============================] - 15s 10ms/step - loss: 0.5593\n",
      "Epoch 10/10\n",
      "1450/1450 [==============================] - 15s 10ms/step - loss: 0.5534\n",
      "Epoch 1/10\n",
      "1444/1444 [==============================] - 49s 34ms/step - loss: 0.7599\n",
      "Epoch 2/10\n",
      "1444/1444 [==============================] - 50s 34ms/step - loss: 10.0194\n",
      "Epoch 3/10\n",
      "1444/1444 [==============================] - 48s 33ms/step - loss: 0.6347\n",
      "Epoch 4/10\n",
      "1444/1444 [==============================] - 46s 32ms/step - loss: 0.6095\n",
      "Epoch 5/10\n",
      "1444/1444 [==============================] - 46s 32ms/step - loss: 0.5982\n",
      "Epoch 6/10\n",
      "1444/1444 [==============================] - 48s 33ms/step - loss: 0.5893\n",
      "Epoch 7/10\n",
      "1444/1444 [==============================] - 47s 33ms/step - loss: 0.5842 4s - loss: 0. - ETA: 3s - loss: 0.597 - ETA: 3s - loss\n",
      "Epoch 8/10\n",
      "1444/1444 [==============================] - 46s 32ms/step - loss: 0.5776\n",
      "Epoch 9/10\n",
      "1444/1444 [==============================] - 48s 33ms/step - loss: 0.6312\n",
      "Epoch 10/10\n",
      "1444/1444 [==============================] - 47s 32ms/step - loss: 0.5750\n",
      "Epoch 1/10\n",
      "1421/1421 [==============================] - 185s 130ms/step - loss: 1118401986560.0000\n",
      "Epoch 2/10\n",
      "1421/1421 [==============================] - 191s 134ms/step - loss: 16710153273344.0000\n",
      "Epoch 3/10\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1421/1421 [==============================] - 169s 119ms/step - loss: nan\n",
      "Epoch 4/10\n",
      "1421/1421 [==============================] - 152s 107ms/step - loss: nan\n",
      "Epoch 5/10\n",
      "1421/1421 [==============================] - 143s 101ms/step - loss: nan\n",
      "Epoch 6/10\n",
      "1421/1421 [==============================] - 150s 105ms/step - loss: nan\n",
      "Epoch 7/10\n",
      "1421/1421 [==============================] - 149s 105ms/step - loss: nan\n",
      "Epoch 8/10\n",
      "1421/1421 [==============================] - 138s 97ms/step - loss: nan\n",
      "Epoch 9/10\n",
      "1421/1421 [==============================] - 138s 97ms/step - loss: nan\n",
      "Epoch 10/10\n",
      "1421/1421 [==============================] - 139s 98ms/step - loss: nan\n"
     ]
    },
    {
     "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>batch_size</th>\n",
       "      <th>epoch_size</th>\n",
       "      <th>lenght</th>\n",
       "      <th>Test Accurcay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.383255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>-0.477982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0267117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>30</td>\n",
       "      <td>0.0487681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.0844886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>30</td>\n",
       "      <td>-0.213605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.855603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>-0.165072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.0233251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>30</td>\n",
       "      <td>0.0988616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.195557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>30</td>\n",
       "      <td>-1.66937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.0282652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>30</td>\n",
       "      <td>0.128042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>-0.011418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>30</td>\n",
       "      <td>0.0484387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>-3.20579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>30</td>\n",
       "      <td>-0.705369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>10</td>\n",
       "      <td>120</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  batch_size epoch_size lenght Test Accurcay\n",
       "0          1          5      7     -0.383255\n",
       "0          1          5     30     -0.477982\n",
       "0          1          5    120           NaN\n",
       "0          1          7      7     0.0267117\n",
       "0          1          7     30     0.0487681\n",
       "0          1          7    120           NaN\n",
       "0          1         10      7    -0.0844886\n",
       "0          1         10     30     -0.213605\n",
       "0          1         10    120           NaN\n",
       "0          2          5      7     -0.855603\n",
       "0          2          5     30     -0.165072\n",
       "0          2          5    120           NaN\n",
       "0          2          7      7    -0.0233251\n",
       "0          2          7     30     0.0988616\n",
       "0          2          7    120           NaN\n",
       "0          2         10      7     -0.195557\n",
       "0          2         10     30      -1.66937\n",
       "0          2         10    120           NaN\n",
       "0          4          5      7    -0.0282652\n",
       "0          4          5     30      0.128042\n",
       "0          4          5    120           NaN\n",
       "0          4          7      7     -0.011418\n",
       "0          4          7     30     0.0484387\n",
       "0          4          7    120           NaN\n",
       "0          4         10      7      -3.20579\n",
       "0          4         10     30     -0.705369\n",
       "0          4         10    120           NaN"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#LSTM\n",
    "\n",
    "# Data Prep\n",
    "dataset = pd.DataFrame(target_train).values\n",
    "dataset = dataset.astype('float32')\n",
    "\n",
    "# Batch Size\n",
    "batch_size = [1,2,4]\n",
    "# Epochs\n",
    "epoch_size = [5,7,10]\n",
    "# lenght\n",
    "lenght = [7,30,120]\n",
    "\n",
    "n_features = 1\n",
    "\n",
    "\n",
    "LSTM_Test_Accuracy_Data = pd.DataFrame(columns = ['batch_size','epoch_size','lenght','Test Accurcay'])\n",
    "\n",
    "for x in list(itertools.product(batch_size, epoch_size,lenght)):\n",
    "            generator = TimeseriesGenerator(dataset, dataset,batch_size=x[0], length= x[2])\n",
    "\n",
    "            lstm_model = Sequential()\n",
    "            lstm_model.add(LSTM(200, activation='relu', input_shape=(x[2], n_features)))\n",
    "            lstm_model.add(Dense(1))\n",
    "            lstm_model.compile(optimizer='adam', loss='mse')\n",
    "            lstm_model.fit_generator(generator,epochs=x[1])\n",
    "\n",
    "            lstm_predictions_scaled = list()\n",
    "\n",
    "            batch = dataset[-x[2]:]\n",
    "            current_batch = batch.reshape((1, x[2], n_features))\n",
    "            # Test Data\n",
    "            for i in range(len(target_test)):   \n",
    "                lstm_pred = lstm_model.predict(current_batch)[0]\n",
    "                lstm_predictions_scaled.append(lstm_pred) \n",
    "                current_batch = np.append(current_batch[:,1:,:],[[lstm_pred]],axis=1)    \n",
    "\n",
    "            predictions_test = pd.DataFrame(lstm_predictions_scaled)\n",
    "            predictions_test.index = target_test.index\n",
    "            # Calculate the absolute errors\n",
    "            errors_test = abs(predictions_test.iloc[:,0] - target_test)\n",
    "             # Calculate mean absolute percentage error (MAPE)\n",
    "            mape_test = 100 * (errors_test/ target_test)\n",
    "            # Calculate and display accuracy\n",
    "            accuracy_test = 100 - np.mean(mape_test)\n",
    "\n",
    "            LSTM_Test_Accuracy_Data_One = pd.DataFrame(index = range(1),columns = ['batch_size','epoch_size','lenght','Test Accurcay']) \n",
    "\n",
    "            LSTM_Test_Accuracy_Data_One.loc[:,'batch_size'] = x[0]\n",
    "            LSTM_Test_Accuracy_Data_One.loc[:,'epoch_size'] = x[1]\n",
    "            LSTM_Test_Accuracy_Data_One.loc[:,'lenght'] = x[2]\n",
    "            LSTM_Test_Accuracy_Data_One.loc[:,'Test Accurcay'] = accuracy_test\n",
    "\n",
    "\n",
    "            LSTM_Test_Accuracy_Data = LSTM_Test_Accuracy_Data.append(LSTM_Test_Accuracy_Data_One)\n",
    "            \n",
    "LSTM_Test_Accuracy_Data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "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>n_estimators</th>\n",
       "      <th>max_depth</th>\n",
       "      <th>min_samples_split</th>\n",
       "      <th>min_samples_leaf</th>\n",
       "      <th>Train Accurcay</th>\n",
       "      <th>Test Accurcay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>2</td>\n",
       "      <td>129.586</td>\n",
       "      <td>-0.131677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>4</td>\n",
       "      <td>129.481</td>\n",
       "      <td>-0.119548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>50</td>\n",
       "      <td>10</td>\n",
       "      <td>126.455</td>\n",
       "      <td>-0.0781384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>131.008</td>\n",
       "      <td>-0.08857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>100</td>\n",
       "      <td>4</td>\n",
       "      <td>129.76</td>\n",
       "      <td>-0.0755297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>4</td>\n",
       "      <td>122.364</td>\n",
       "      <td>-0.0701365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>100</td>\n",
       "      <td>10</td>\n",
       "      <td>122.793</td>\n",
       "      <td>-0.0359609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>200</td>\n",
       "      <td>2</td>\n",
       "      <td>130.649</td>\n",
       "      <td>-0.0589393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>200</td>\n",
       "      <td>4</td>\n",
       "      <td>131.866</td>\n",
       "      <td>-0.0500957</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>200</td>\n",
       "      <td>10</td>\n",
       "      <td>131.453</td>\n",
       "      <td>-0.0376685</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>81 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   n_estimators max_depth min_samples_split min_samples_leaf Train Accurcay  \\\n",
       "0           500        10                50                2        129.586   \n",
       "0           500        10                50                4        129.481   \n",
       "0           500        10                50               10        126.455   \n",
       "0           500        10               100                2        131.008   \n",
       "0           500        10               100                4         129.76   \n",
       "..          ...       ...               ...              ...            ...   \n",
       "0          2000       100               100                4        122.364   \n",
       "0          2000       100               100               10        122.793   \n",
       "0          2000       100               200                2        130.649   \n",
       "0          2000       100               200                4        131.866   \n",
       "0          2000       100               200               10        131.453   \n",
       "\n",
       "   Test Accurcay  \n",
       "0      -0.131677  \n",
       "0      -0.119548  \n",
       "0     -0.0781384  \n",
       "0       -0.08857  \n",
       "0     -0.0755297  \n",
       "..           ...  \n",
       "0     -0.0701365  \n",
       "0     -0.0359609  \n",
       "0     -0.0589393  \n",
       "0     -0.0500957  \n",
       "0     -0.0376685  \n",
       "\n",
       "[81 rows x 6 columns]"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Random Forest\n",
    "# Number of trees in random forest\n",
    "n_estimators = [500,1000,2000]\n",
    "# Maximum number of levels in tree\n",
    "max_depth = [10,50,100]\n",
    "# Minimum number of samples required to split a node\n",
    "min_samples_split = [50,100,200]\n",
    "# Minimum number of samples required at each leaf node\n",
    "min_samples_leaf = [2,4,10]\n",
    "\n",
    "RF_Test_Accuracy_Data = pd.DataFrame(columns = ['n_estimators','max_depth','min_samples_split','min_samples_leaf','Train Accurcay','Test Accurcay'])\n",
    "\n",
    "for x in list(itertools.product(n_estimators, max_depth,min_samples_split,min_samples_leaf)):\n",
    "            rf = RandomForestRegressor(n_estimators = x[0],max_depth = x[1], min_samples_split = x[2],min_samples_leaf=x[3], random_state = 10,n_jobs=-1, max_features= \"auto\")\n",
    "            # Train the model on training data\n",
    "            rf.fit(features_train, target_train)\n",
    "\n",
    "            # Train Data\n",
    "            # Use the forest's predict method on the train data\n",
    "            predictions_train = rf.predict(features_train)\n",
    "            # Calculate the absolute errors\n",
    "            errors_train = abs(predictions_train - target_train)\n",
    "             # Calculate mean absolute percentage error (MAPE)\n",
    "            mape_train = 100 * (errors_train/ target_train)\n",
    "            # Calculate and display accuracy\n",
    "            accuracy_train = 100 - np.mean(mape_train)\n",
    "\n",
    "            # Test Data\n",
    "            # Use the forest's predict method on the test data\n",
    "            predictions_test = rf.predict(features_test)\n",
    "            # Calculate the absolute errors\n",
    "            errors_test = abs(predictions_test - target_test)\n",
    "             # Calculate mean absolute percentage error (MAPE)\n",
    "            mape_test = 100 * (errors_test/ target_test)\n",
    "            # Calculate and display accuracy\n",
    "            accuracy_test = 100 - np.mean(mape_test)\n",
    "\n",
    "            RF_Test_Accuracy_Data_One = pd.DataFrame(index = range(1),columns = ['n_estimators','max_depth','min_samples_split','min_samples_leaf','Train Accurcay','Test Accurcay']) \n",
    "\n",
    "            RF_Test_Accuracy_Data_One.loc[:,'n_estimators'] = x[0]\n",
    "            RF_Test_Accuracy_Data_One.loc[:,'max_depth'] = x[1]\n",
    "            RF_Test_Accuracy_Data_One.loc[:,'min_samples_split'] = x[2]\n",
    "            RF_Test_Accuracy_Data_One.loc[:,'min_samples_leaf'] = x[3]\n",
    "            RF_Test_Accuracy_Data_One.loc[:,'Train Accurcay'] = accuracy_train\n",
    "            RF_Test_Accuracy_Data_One.loc[:,'Test Accurcay'] = accuracy_test\n",
    "\n",
    "\n",
    "            RF_Test_Accuracy_Data = RF_Test_Accuracy_Data.append(RF_Test_Accuracy_Data_One)\n",
    "            \n",
    "RF_Test_Accuracy_Data\n",
    "\n",
    "\n",
    "                                   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "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>n_estimators</th>\n",
       "      <th>max_depth</th>\n",
       "      <th>min_child_weight</th>\n",
       "      <th>gamma</th>\n",
       "      <th>learning_rate</th>\n",
       "      <th>Train Accurcay</th>\n",
       "      <th>Test Accurcay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.1</td>\n",
       "      <td>115.653</td>\n",
       "      <td>-0.459981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.05</td>\n",
       "      <td>115.504</td>\n",
       "      <td>-0.341839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.01</td>\n",
       "      <td>117.937</td>\n",
       "      <td>-0.375321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>0.1</td>\n",
       "      <td>129.018</td>\n",
       "      <td>-0.296427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>500</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>0.05</td>\n",
       "      <td>127.328</td>\n",
       "      <td>-0.25741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0.05</td>\n",
       "      <td>103.926</td>\n",
       "      <td>-0.290538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0.01</td>\n",
       "      <td>107.914</td>\n",
       "      <td>-0.339946</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.1</td>\n",
       "      <td>120.039</td>\n",
       "      <td>-0.107394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.05</td>\n",
       "      <td>123.743</td>\n",
       "      <td>-0.0449505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2000</td>\n",
       "      <td>100</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.01</td>\n",
       "      <td>129.729</td>\n",
       "      <td>-0.0509627</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>108 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   n_estimators max_depth min_child_weight gamma learning_rate Train Accurcay  \\\n",
       "0           500        10                1     1           0.1        115.653   \n",
       "0           500        10                1     1          0.05        115.504   \n",
       "0           500        10                1     1          0.01        117.937   \n",
       "0           500        10                1     5           0.1        129.018   \n",
       "0           500        10                1     5          0.05        127.328   \n",
       "..          ...       ...              ...   ...           ...            ...   \n",
       "0          2000       100                2     1          0.05        103.926   \n",
       "0          2000       100                2     1          0.01        107.914   \n",
       "0          2000       100                2     5           0.1        120.039   \n",
       "0          2000       100                2     5          0.05        123.743   \n",
       "0          2000       100                2     5          0.01        129.729   \n",
       "\n",
       "   Test Accurcay  \n",
       "0      -0.459981  \n",
       "0      -0.341839  \n",
       "0      -0.375321  \n",
       "0      -0.296427  \n",
       "0       -0.25741  \n",
       "..           ...  \n",
       "0      -0.290538  \n",
       "0      -0.339946  \n",
       "0      -0.107394  \n",
       "0     -0.0449505  \n",
       "0     -0.0509627  \n",
       "\n",
       "[108 rows x 7 columns]"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#XG Boost\n",
    "# Number of trees\n",
    "n_estimators = [500,1000,2000]\n",
    "# Maximum number of levels in tree\n",
    "max_depth = [10,50,100]\n",
    "#minimum sum of weights of all observations required in a child\n",
    "min_child_weight = [1,2]\n",
    "#Gamma specifies the minimum loss reduction required to make a split\n",
    "gamma = [1,5]\n",
    "# boosting learning rate\n",
    "learning_rate = [.1,.05,.01]\n",
    "\n",
    "\n",
    "xgb_Test_Accuracy_Data = pd.DataFrame(columns = ['n_estimators','max_depth','min_child_weight','gamma','learning_rate','Train Accurcay','Test Accurcay'])\n",
    "\n",
    "for x in list(itertools.product(n_estimators, max_depth,min_child_weight,gamma,learning_rate)):\n",
    "            xgb_reg = xgb.XGBRegressor(n_estimators=x[0],max_depth =x[1],min_child_weight = x[2],gamma = x[3],learning_rate = x[4])\n",
    "        \n",
    "            # Train the model on training data\n",
    "            xgb_reg.fit(features_train, target_train)\n",
    "\n",
    "            # Train Data\n",
    "            # Use the forest's predict method on the train data\n",
    "            predictions_train = xgb_reg.predict(features_train)\n",
    "            # Calculate the absolute errors\n",
    "            errors_train = abs(predictions_train - target_train)\n",
    "             # Calculate mean absolute percentage error (MAPE)\n",
    "            mape_train = 100 * (errors_train/ target_train)\n",
    "            # Calculate and display accuracy\n",
    "            accuracy_train = 100 - np.mean(mape_train)\n",
    "\n",
    "            # Test Data\n",
    "            # Use the forest's predict method on the test data\n",
    "            predictions_test = xgb_reg.predict(features_test)\n",
    "            # Calculate the absolute errors\n",
    "            errors_test = abs(predictions_test - target_test)\n",
    "             # Calculate mean absolute percentage error (MAPE)\n",
    "            mape_test = 100 * (errors_test/ target_test)\n",
    "            # Calculate and display accuracy\n",
    "            accuracy_test = 100 - np.mean(mape_test)\n",
    "\n",
    "            xgb_Test_Accuracy_Data_One = pd.DataFrame(index = range(1),columns = ['n_estimators','max_depth','min_child_weight','gamma','learning_rate','Train Accurcay','Test Accurcay']) \n",
    "\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'n_estimators'] = x[0]\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'max_depth'] = x[1]\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'min_child_weight'] = x[2]\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'gamma'] = x[3]\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'learning_rate'] = x[4]\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'Train Accurcay'] = accuracy_train\n",
    "            xgb_Test_Accuracy_Data_One.loc[:,'Test Accurcay'] = accuracy_test\n",
    "\n",
    "\n",
    "            xgb_Test_Accuracy_Data = xgb_Test_Accuracy_Data.append(xgb_Test_Accuracy_Data_One)\n",
    "            \n",
    "xgb_Test_Accuracy_Data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "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>C</th>\n",
       "      <th>gamma</th>\n",
       "      <th>Train Accurcay</th>\n",
       "      <th>Test Accurcay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.1</td>\n",
       "      <td>1</td>\n",
       "      <td>128.078</td>\n",
       "      <td>0.0439536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.1</td>\n",
       "      <td>5</td>\n",
       "      <td>130.116</td>\n",
       "      <td>0.0377233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>133.55</td>\n",
       "      <td>0.0876367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "      <td>131.436</td>\n",
       "      <td>0.0694628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>132.013</td>\n",
       "      <td>0.143509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10.0</td>\n",
       "      <td>5</td>\n",
       "      <td>134.279</td>\n",
       "      <td>0.0800868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100.0</td>\n",
       "      <td>1</td>\n",
       "      <td>136.386</td>\n",
       "      <td>0.237454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100.0</td>\n",
       "      <td>5</td>\n",
       "      <td>136.41</td>\n",
       "      <td>0.0811142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>137.445</td>\n",
       "      <td>0.362094</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1000.0</td>\n",
       "      <td>5</td>\n",
       "      <td>127.833</td>\n",
       "      <td>0.0708028</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        C gamma Train Accurcay Test Accurcay\n",
       "0     0.1     1        128.078     0.0439536\n",
       "0     0.1     5        130.116     0.0377233\n",
       "0     1.0     1         133.55     0.0876367\n",
       "0     1.0     5        131.436     0.0694628\n",
       "0    10.0     1        132.013      0.143509\n",
       "0    10.0     5        134.279     0.0800868\n",
       "0   100.0     1        136.386      0.237454\n",
       "0   100.0     5         136.41     0.0811142\n",
       "0  1000.0     1        137.445      0.362094\n",
       "0  1000.0     5        127.833     0.0708028"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#SVR\n",
    "# Regularization parameter \n",
    "C = [0.1, 1, 10, 100, 1000]\n",
    "#Kernel coefficient \n",
    "gamma = [1,5]\n",
    "\n",
    "\n",
    "\n",
    "svr_Test_Accuracy_Data = pd.DataFrame(columns = ['C','gamma','Train Accurcay','Test Accurcay'])\n",
    "\n",
    "for x in list(itertools.product(C,gamma)):\n",
    "                svr_reg = SVR(kernel= 'rbf', C = x[0], gamma= x[1])\n",
    "                # Train the model on training data\n",
    "                svr_reg.fit(features_train, target_train)\n",
    "\n",
    "                # Train Data\n",
    "                # Use the forest's predict method on the train data\n",
    "                predictions_train = svr_reg.predict(features_train)\n",
    "                # Calculate the absolute errors\n",
    "                errors_train = abs(predictions_train - target_train)\n",
    "                 # Calculate mean absolute percentage error (MAPE)\n",
    "                mape_train = 100 * (errors_train/ target_train)\n",
    "                # Calculate and display accuracy\n",
    "                accuracy_train = 100 - np.mean(mape_train)\n",
    "\n",
    "                # Test Data\n",
    "                # Use the forest's predict method on the test data\n",
    "                predictions_test = svr_reg.predict(features_test)\n",
    "                # Calculate the absolute errors\n",
    "                errors_test = abs(predictions_test - target_test)\n",
    "                 # Calculate mean absolute percentage error (MAPE)\n",
    "                mape_test = 100 * (errors_test/ target_test)\n",
    "                # Calculate and display accuracy\n",
    "                accuracy_test = 100 - np.mean(mape_test)\n",
    "\n",
    "                svr_Test_Accuracy_Data_One = pd.DataFrame(index = range(1),columns = ['C','gamma','Train Accurcay','Test Accurcay']) \n",
    "\n",
    "                svr_Test_Accuracy_Data_One.loc[:,'C'] = x[0]\n",
    "                svr_Test_Accuracy_Data_One.loc[:,'gamma'] = x[1]\n",
    "                svr_Test_Accuracy_Data_One.loc[:,'Train Accurcay'] = accuracy_train\n",
    "                svr_Test_Accuracy_Data_One.loc[:,'Test Accurcay'] = accuracy_test\n",
    "\n",
    "\n",
    "                svr_Test_Accuracy_Data = svr_Test_Accuracy_Data.append(svr_Test_Accuracy_Data_One)\n",
    "\n",
    "svr_Test_Accuracy_Data\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train and Predict using best fit model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1535/1535 [==============================] - 73s 47ms/step - loss: 184057.3281\n",
      "Epoch 2/5\n",
      "1535/1535 [==============================] - 73s 48ms/step - loss: 0.9244\n",
      "Epoch 3/5\n",
      "1535/1535 [==============================] - 73s 48ms/step - loss: 0.9119\n",
      "Epoch 4/5\n",
      "1535/1535 [==============================] - 76s 49ms/step - loss: 1.3430\n",
      "Epoch 5/5\n",
      "1535/1535 [==============================] - 73s 47ms/step - loss: 0.9205\n"
     ]
    },
    {
     "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>LSTM_Forecast_De_Standardise</th>\n",
       "      <th>LSTM_Forecast_De_Standardise_R_Sum</th>\n",
       "      <th>LSTM_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>0.019059</td>\n",
       "      <td>0.019059</td>\n",
       "      <td>365.830055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>0.022284</td>\n",
       "      <td>0.041343</td>\n",
       "      <td>365.852339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>0.023896</td>\n",
       "      <td>0.065239</td>\n",
       "      <td>365.876235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>0.025874</td>\n",
       "      <td>0.091112</td>\n",
       "      <td>365.902109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>0.028028</td>\n",
       "      <td>0.119141</td>\n",
       "      <td>365.930138</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            LSTM_Forecast_De_Standardise  LSTM_Forecast_De_Standardise_R_Sum  \\\n",
       "2019-11-27                      0.019059                            0.019059   \n",
       "2019-11-28                      0.022284                            0.041343   \n",
       "2019-11-29                      0.023896                            0.065239   \n",
       "2019-11-30                      0.025874                            0.091112   \n",
       "2019-12-01                      0.028028                            0.119141   \n",
       "\n",
       "            LSTM_Forecast  \n",
       "2019-11-27     365.830055  \n",
       "2019-11-28     365.852339  \n",
       "2019-11-29     365.876235  \n",
       "2019-11-30     365.902109  \n",
       "2019-12-01     365.930138  "
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#LSTM\n",
    "\n",
    "\n",
    "# Data Prep\n",
    "dataset = pd.DataFrame(scaled_AI_Model_Target).values\n",
    "dataset = dataset.astype('float32')\n",
    "Latest_target = target_final.tail(1).values.tolist()\n",
    "\n",
    "# Best Fit Model\n",
    "\n",
    "Best_Fit_LSTM = LSTM_Test_Accuracy_Data.loc[LSTM_Test_Accuracy_Data['Test Accurcay'] == max(LSTM_Test_Accuracy_Data['Test Accurcay'])]\n",
    "\n",
    "Best_Fit_LSTM = Best_Fit_LSTM.values.flatten().tolist()\n",
    "\n",
    "# Fit  Model\n",
    "generator = TimeseriesGenerator(dataset, dataset,batch_size=Best_Fit_LSTM[0], length= Best_Fit_LSTM[2])\n",
    "\n",
    "lstm_model = Sequential()\n",
    "lstm_model.add(LSTM(200, activation='relu', input_shape=(Best_Fit_LSTM[2], n_features)))\n",
    "lstm_model.add(Dense(1))\n",
    "lstm_model.compile(optimizer='adam', loss='mse')\n",
    "lstm_model.fit_generator(generator,epochs=Best_Fit_LSTM[1])\n",
    "\n",
    "lstm_predictions_scaled = list()\n",
    "\n",
    "batch = dataset[-Best_Fit_LSTM[2]:]\n",
    "current_batch = batch.reshape((1, Best_Fit_LSTM[2], n_features))\n",
    "# Test Data\n",
    "for i in range(365):   \n",
    "    lstm_pred = lstm_model.predict(current_batch)[0]\n",
    "    lstm_predictions_scaled.append(lstm_pred) \n",
    "    current_batch = np.append(current_batch[:,1:,:],[[lstm_pred]],axis=1)    \n",
    "\n",
    "LSTM_Forecast = pd.DataFrame(lstm_predictions_scaled)\n",
    "\n",
    "            \n",
    "\n",
    "# De-Standardise \n",
    "LSTM_Forecast_Adjusted_1 = (LSTM_Forecast*std_target)+ mean_target\n",
    "LSTM_Forecast_Adjusted_1 \n",
    "\n",
    "LSTM_Forecast_Adjusted_2 = pd.DataFrame({'LSTM_Forecast_De_Standardise': LSTM_Forecast_Adjusted_1.iloc[:,0]})\n",
    "\n",
    "\n",
    "#Roll back first difference\n",
    "\n",
    "#LSTM_Forecast_Adjusted_3 = AI_Model_data\n",
    "\n",
    "\n",
    "\n",
    "LSTM_Forecast_Adjusted_2['LSTM_Forecast_De_Standardise_R_Sum'] = LSTM_Forecast_Adjusted_2['LSTM_Forecast_De_Standardise'].cumsum()\n",
    "LSTM_Forecast_Adjusted_2['LSTM_Forecast'] = LSTM_Forecast_Adjusted_2['LSTM_Forecast_De_Standardise_R_Sum'] + Latest_target[0]\n",
    "\n",
    "LSTM_Forecast_Adjusted_2.index = HWES_Forecast.index\n",
    "LSTM_Forecast_Adjusted_2.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "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>Random_Forest_Forecast_De_Standardise</th>\n",
       "      <th>Random_Forest_Forecast_De_Standardise_R_Sum</th>\n",
       "      <th>Random_Forest_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>0.255585</td>\n",
       "      <td>0.255585</td>\n",
       "      <td>366.066582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>0.251891</td>\n",
       "      <td>0.507476</td>\n",
       "      <td>366.318473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>0.252236</td>\n",
       "      <td>0.759712</td>\n",
       "      <td>366.570709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>0.094024</td>\n",
       "      <td>0.853736</td>\n",
       "      <td>366.664733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>0.111262</td>\n",
       "      <td>0.964998</td>\n",
       "      <td>366.775994</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Random_Forest_Forecast_De_Standardise  \\\n",
       "2019-11-27                               0.255585   \n",
       "2019-11-28                               0.251891   \n",
       "2019-11-29                               0.252236   \n",
       "2019-11-30                               0.094024   \n",
       "2019-12-01                               0.111262   \n",
       "\n",
       "            Random_Forest_Forecast_De_Standardise_R_Sum  \\\n",
       "2019-11-27                                     0.255585   \n",
       "2019-11-28                                     0.507476   \n",
       "2019-11-29                                     0.759712   \n",
       "2019-11-30                                     0.853736   \n",
       "2019-12-01                                     0.964998   \n",
       "\n",
       "            Random_Forest_Forecast  \n",
       "2019-11-27              366.066582  \n",
       "2019-11-28              366.318473  \n",
       "2019-11-29              366.570709  \n",
       "2019-11-30              366.664733  \n",
       "2019-12-01              366.775994  "
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Random Forest\n",
    "\n",
    "# Best Fit Model\n",
    "\n",
    "Best_Fit_Random_Forest = RF_Test_Accuracy_Data.loc[RF_Test_Accuracy_Data['Test Accurcay'] == max(RF_Test_Accuracy_Data['Test Accurcay'])]\n",
    "\n",
    "Best_Fit_Random_Forest = Best_Fit_Random_Forest.values.flatten().tolist()\n",
    "\n",
    "# Fit  Model\n",
    "rf = RandomForestRegressor(n_estimators = Best_Fit_Random_Forest[0],max_depth = Best_Fit_Random_Forest[1], min_samples_split = Best_Fit_Random_Forest[2],min_samples_leaf=Best_Fit_Random_Forest[3], random_state = 10,n_jobs=-1, max_features= \"auto\")# Train the model on training data\n",
    "rf.fit(features_train, target_train)\n",
    "\n",
    "# Use the forest's predict method on the test data\n",
    "Random_Forest_Forecast = rf.predict(driver_forecasts)\n",
    "Random_Forest_Forecast\n",
    "\n",
    "# De-Standardise \n",
    "Random_Forest_Forecast_Adjusted_1 = (Random_Forest_Forecast*std_target)+ mean_target\n",
    "Random_Forest_Forecast_Adjusted_2 = pd.DataFrame({'Random_Forest_Forecast_De_Standardise': Random_Forest_Forecast_Adjusted_1[:]})\n",
    "\n",
    "\n",
    "#Roll back first difference\n",
    "\n",
    "#Random_Forest_Forecast_Adjusted_3 = AI_Model_data\n",
    "\n",
    "\n",
    "\n",
    "Random_Forest_Forecast_Adjusted_2['Random_Forest_Forecast_De_Standardise_R_Sum'] = Random_Forest_Forecast_Adjusted_2['Random_Forest_Forecast_De_Standardise'].cumsum()\n",
    "Random_Forest_Forecast_Adjusted_2['Random_Forest_Forecast'] = Random_Forest_Forecast_Adjusted_2['Random_Forest_Forecast_De_Standardise_R_Sum'] + Latest_target[0]\n",
    "\n",
    "Random_Forest_Forecast_Adjusted_2.index = HWES_Forecast.index\n",
    "Random_Forest_Forecast_Adjusted_2.head(5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "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>XGB_Forecast_De_Standardise</th>\n",
       "      <th>XGB_Forecast_De_Standardise_R_Sum</th>\n",
       "      <th>XGB_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>0.262247</td>\n",
       "      <td>0.262247</td>\n",
       "      <td>366.073242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>0.208593</td>\n",
       "      <td>0.470840</td>\n",
       "      <td>366.281830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>0.171175</td>\n",
       "      <td>0.642016</td>\n",
       "      <td>366.453033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>0.090658</td>\n",
       "      <td>0.732673</td>\n",
       "      <td>366.543671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>0.086032</td>\n",
       "      <td>0.818705</td>\n",
       "      <td>366.629700</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            XGB_Forecast_De_Standardise  XGB_Forecast_De_Standardise_R_Sum  \\\n",
       "2019-11-27                     0.262247                           0.262247   \n",
       "2019-11-28                     0.208593                           0.470840   \n",
       "2019-11-29                     0.171175                           0.642016   \n",
       "2019-11-30                     0.090658                           0.732673   \n",
       "2019-12-01                     0.086032                           0.818705   \n",
       "\n",
       "            XGB_Forecast  \n",
       "2019-11-27    366.073242  \n",
       "2019-11-28    366.281830  \n",
       "2019-11-29    366.453033  \n",
       "2019-11-30    366.543671  \n",
       "2019-12-01    366.629700  "
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#XG Boost\n",
    "\n",
    "# Best Fit Model\n",
    "\n",
    "Best_Fit_XG_Boost = xgb_Test_Accuracy_Data.loc[xgb_Test_Accuracy_Data['Test Accurcay'] == max(xgb_Test_Accuracy_Data['Test Accurcay'])]\n",
    "\n",
    "Best_Fit_XG_Boost = Best_Fit_XG_Boost.values.flatten().tolist()\n",
    "\n",
    "# Fit  Model\n",
    "xgb_reg = xgb.XGBRegressor(n_estimators=Best_Fit_XG_Boost[0],max_depth =Best_Fit_XG_Boost[1],min_child_weight = Best_Fit_XG_Boost[2],gamma = Best_Fit_XG_Boost[3],learning_rate = Best_Fit_XG_Boost[4])\n",
    "xgb_reg.fit(features_train, target_train)\n",
    "\n",
    "# Use the forest's predict method on the test data\n",
    "XGB_Forecast = xgb_reg.predict(driver_forecasts)\n",
    "\n",
    "# De-Standardise \n",
    "XGB_Forecast_Adjusted_1 = (XGB_Forecast*std_target)+ mean_target\n",
    "XGB_Forecast_Adjusted_2 = pd.DataFrame({'XGB_Forecast_De_Standardise': XGB_Forecast_Adjusted_1[:]})\n",
    "\n",
    "\n",
    "#Roll back first difference\n",
    "\n",
    "XGB_Forecast_Adjusted_2['XGB_Forecast_De_Standardise_R_Sum'] = XGB_Forecast_Adjusted_2['XGB_Forecast_De_Standardise'].cumsum()\n",
    "XGB_Forecast_Adjusted_2['XGB_Forecast'] = XGB_Forecast_Adjusted_2['XGB_Forecast_De_Standardise_R_Sum'] + Latest_target[0]\n",
    "\n",
    "XGB_Forecast_Adjusted_2.index = HWES_Forecast.index\n",
    "XGB_Forecast_Adjusted_2.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "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>svr_Forecast_De_Standardise</th>\n",
       "      <th>svr_Forecast_De_Standardise_R_Sum</th>\n",
       "      <th>SVM_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>1.872393</td>\n",
       "      <td>1.872393</td>\n",
       "      <td>367.683390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>1.826521</td>\n",
       "      <td>3.698914</td>\n",
       "      <td>369.509911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>1.883420</td>\n",
       "      <td>5.582334</td>\n",
       "      <td>371.393331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>-0.009321</td>\n",
       "      <td>5.573013</td>\n",
       "      <td>371.384010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>0.109823</td>\n",
       "      <td>5.682836</td>\n",
       "      <td>371.493833</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            svr_Forecast_De_Standardise  svr_Forecast_De_Standardise_R_Sum  \\\n",
       "2019-11-27                     1.872393                           1.872393   \n",
       "2019-11-28                     1.826521                           3.698914   \n",
       "2019-11-29                     1.883420                           5.582334   \n",
       "2019-11-30                    -0.009321                           5.573013   \n",
       "2019-12-01                     0.109823                           5.682836   \n",
       "\n",
       "            SVM_Forecast  \n",
       "2019-11-27    367.683390  \n",
       "2019-11-28    369.509911  \n",
       "2019-11-29    371.393331  \n",
       "2019-11-30    371.384010  \n",
       "2019-12-01    371.493833  "
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# SVR\n",
    "\n",
    "\n",
    "# Best Fit Model\n",
    "\n",
    "Best_Fit_SVR = svr_Test_Accuracy_Data.loc[svr_Test_Accuracy_Data['Test Accurcay'] == max(svr_Test_Accuracy_Data['Test Accurcay'])]\n",
    "\n",
    "Best_Fit_SVR = Best_Fit_SVR.values.flatten().tolist()\n",
    "\n",
    "# Fit  Model\n",
    "svr_reg = SVR(kernel= 'rbf', C = Best_Fit_SVR[0], gamma= Best_Fit_SVR[1])\n",
    "# Train the model on training data\n",
    "svr_reg.fit(features_train, target_train)\n",
    "\n",
    "# Use the forest's predict method on the test data\n",
    "svr_Forecast = svr_reg.predict(driver_forecasts)\n",
    "\n",
    "# De-Standardise \n",
    "svr_Forecast_Adjusted_1 = (svr_Forecast*std_target)+ mean_target\n",
    "svr_Forecast_Adjusted_2 = pd.DataFrame({'svr_Forecast_De_Standardise': svr_Forecast_Adjusted_1[:]})\n",
    "\n",
    "\n",
    "#Roll back first difference\n",
    "\n",
    "svr_Forecast_Adjusted_2['svr_Forecast_De_Standardise_R_Sum'] = svr_Forecast_Adjusted_2['svr_Forecast_De_Standardise'].cumsum()\n",
    "svr_Forecast_Adjusted_2['SVM_Forecast'] = svr_Forecast_Adjusted_2['svr_Forecast_De_Standardise_R_Sum'] + Latest_target[0]\n",
    "\n",
    "svr_Forecast_Adjusted_2.index = HWES_Forecast.index\n",
    "svr_Forecast_Adjusted_2.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "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>Regression_Forecast_De_Standardise</th>\n",
       "      <th>Regression_Forecast_De_Standardise_R_Sum</th>\n",
       "      <th>Regression_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-11-27</th>\n",
       "      <td>0.041990</td>\n",
       "      <td>0.041990</td>\n",
       "      <td>365.852986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-28</th>\n",
       "      <td>0.044190</td>\n",
       "      <td>0.086180</td>\n",
       "      <td>365.897177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-29</th>\n",
       "      <td>0.045961</td>\n",
       "      <td>0.132141</td>\n",
       "      <td>365.943138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-11-30</th>\n",
       "      <td>0.043784</td>\n",
       "      <td>0.175925</td>\n",
       "      <td>365.986921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-12-01</th>\n",
       "      <td>0.041332</td>\n",
       "      <td>0.217257</td>\n",
       "      <td>366.028254</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Regression_Forecast_De_Standardise  \\\n",
       "2019-11-27                            0.041990   \n",
       "2019-11-28                            0.044190   \n",
       "2019-11-29                            0.045961   \n",
       "2019-11-30                            0.043784   \n",
       "2019-12-01                            0.041332   \n",
       "\n",
       "            Regression_Forecast_De_Standardise_R_Sum  Regression_Forecast  \n",
       "2019-11-27                                  0.041990           365.852986  \n",
       "2019-11-28                                  0.086180           365.897177  \n",
       "2019-11-29                                  0.132141           365.943138  \n",
       "2019-11-30                                  0.175925           365.986921  \n",
       "2019-12-01                                  0.217257           366.028254  "
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Linear Regression\n",
    "\n",
    "reg = LinearRegression().fit(features_final, target_final)\n",
    "\n",
    "Regression_Forecast = reg.predict(driver_forecasts)\n",
    "\n",
    "Regression_Forecast\n",
    "\n",
    "\n",
    "#XG Boost\n",
    "\n",
    "reg = LinearRegression().fit(features_train, target_train)\n",
    "\n",
    "# Use the forest's predict method on the test data\n",
    "Regression_Forecast = reg.predict(driver_forecasts)\n",
    "\n",
    "# De-Standardise \n",
    "Regression_Forecast_Adjusted_1 = (Regression_Forecast*std_target)+ mean_target\n",
    "Regression_Forecast_Adjusted_2 = pd.DataFrame({'Regression_Forecast_De_Standardise': Regression_Forecast_Adjusted_1[:]})\n",
    "\n",
    "\n",
    "#Roll back first difference\n",
    "\n",
    "Regression_Forecast_Adjusted_2['Regression_Forecast_De_Standardise_R_Sum'] = Regression_Forecast_Adjusted_2['Regression_Forecast_De_Standardise'].cumsum()\n",
    "Regression_Forecast_Adjusted_2['Regression_Forecast'] = Regression_Forecast_Adjusted_2['Regression_Forecast_De_Standardise_R_Sum'] + Latest_target[0]\n",
    "\n",
    "Regression_Forecast_Adjusted_2.index = HWES_Forecast.index\n",
    "Regression_Forecast_Adjusted_2.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Combine all forecasts by averaging them "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "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>HWES_Forecast</th>\n",
       "      <th>SARIMA_Forecast</th>\n",
       "      <th>SARIMAX_Forecast</th>\n",
       "      <th>VAR_Forecast</th>\n",
       "      <th>VECM_Forecast</th>\n",
       "      <th>LSTM_Forecast</th>\n",
       "      <th>Random_Forest_Forecast</th>\n",
       "      <th>XGB_Forecast</th>\n",
       "      <th>SVM_Forecast</th>\n",
       "      <th>Regression_Forecast</th>\n",
       "      <th>Final_Forecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-11-21</th>\n",
       "      <td>264.700460</td>\n",
       "      <td>253.431638</td>\n",
       "      <td>221.074665</td>\n",
       "      <td>360.940682</td>\n",
       "      <td>355.030531</td>\n",
       "      <td>395.384563</td>\n",
       "      <td>430.469606</td>\n",
       "      <td>431.461151</td>\n",
       "      <td>818.746712</td>\n",
       "      <td>381.244890</td>\n",
       "      <td>391.248490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-11-22</th>\n",
       "      <td>264.383448</td>\n",
       "      <td>253.117618</td>\n",
       "      <td>221.035725</td>\n",
       "      <td>360.985673</td>\n",
       "      <td>355.071635</td>\n",
       "      <td>395.469353</td>\n",
       "      <td>430.573268</td>\n",
       "      <td>431.551880</td>\n",
       "      <td>818.853271</td>\n",
       "      <td>381.282561</td>\n",
       "      <td>391.232443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-11-23</th>\n",
       "      <td>264.077485</td>\n",
       "      <td>252.803600</td>\n",
       "      <td>220.997326</td>\n",
       "      <td>361.030803</td>\n",
       "      <td>355.113054</td>\n",
       "      <td>395.554142</td>\n",
       "      <td>430.676861</td>\n",
       "      <td>431.642578</td>\n",
       "      <td>818.959627</td>\n",
       "      <td>381.320238</td>\n",
       "      <td>391.217571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-11-24</th>\n",
       "      <td>263.820963</td>\n",
       "      <td>252.489509</td>\n",
       "      <td>221.041282</td>\n",
       "      <td>361.076043</td>\n",
       "      <td>355.154732</td>\n",
       "      <td>395.638932</td>\n",
       "      <td>430.912862</td>\n",
       "      <td>431.883698</td>\n",
       "      <td>821.325309</td>\n",
       "      <td>381.368820</td>\n",
       "      <td>391.471215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-11-25</th>\n",
       "      <td>263.561464</td>\n",
       "      <td>252.175422</td>\n",
       "      <td>221.089239</td>\n",
       "      <td>361.121295</td>\n",
       "      <td>355.196461</td>\n",
       "      <td>395.723722</td>\n",
       "      <td>431.113957</td>\n",
       "      <td>432.071747</td>\n",
       "      <td>823.855724</td>\n",
       "      <td>381.417867</td>\n",
       "      <td>391.732690</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            HWES_Forecast  SARIMA_Forecast  SARIMAX_Forecast  VAR_Forecast  \\\n",
       "2020-11-21     264.700460       253.431638        221.074665    360.940682   \n",
       "2020-11-22     264.383448       253.117618        221.035725    360.985673   \n",
       "2020-11-23     264.077485       252.803600        220.997326    361.030803   \n",
       "2020-11-24     263.820963       252.489509        221.041282    361.076043   \n",
       "2020-11-25     263.561464       252.175422        221.089239    361.121295   \n",
       "\n",
       "            VECM_Forecast  LSTM_Forecast  Random_Forest_Forecast  \\\n",
       "2020-11-21     355.030531     395.384563              430.469606   \n",
       "2020-11-22     355.071635     395.469353              430.573268   \n",
       "2020-11-23     355.113054     395.554142              430.676861   \n",
       "2020-11-24     355.154732     395.638932              430.912862   \n",
       "2020-11-25     355.196461     395.723722              431.113957   \n",
       "\n",
       "            XGB_Forecast  SVM_Forecast  Regression_Forecast  Final_Forecast  \n",
       "2020-11-21    431.461151    818.746712           381.244890      391.248490  \n",
       "2020-11-22    431.551880    818.853271           381.282561      391.232443  \n",
       "2020-11-23    431.642578    818.959627           381.320238      391.217571  \n",
       "2020-11-24    431.883698    821.325309           381.368820      391.471215  \n",
       "2020-11-25    432.071747    823.855724           381.417867      391.732690  "
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Final_Forecast = pd.concat([HWES_Forecast,SARIMA_Forecast,SARIMAX_Forecast['SARIMAX_Forecast'],VAR_Forecast,VECM_Forecast,LSTM_Forecast_Adjusted_2['LSTM_Forecast'],Random_Forest_Forecast_Adjusted_2['Random_Forest_Forecast'],XGB_Forecast_Adjusted_2['XGB_Forecast'],svr_Forecast_Adjusted_2['SVM_Forecast'],Regression_Forecast_Adjusted_2['Regression_Forecast']],1)\n",
    "Final_Forecast['Final_Forecast'] = Final_Forecast.mean(axis=1)\n",
    "Final_Forecast.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x24cfa1046a0>"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "Final_Forecast.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}