17 VARMA model.ipynb

{
 "cells": [
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "Good thing with the VARMA model is that we can use auto_arima for optimum AR and MA term of the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### VARMA(p,q) : Vector Auto Regression Moving Average models!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "\n",
    "from statsmodels.tsa.statespace.varmax import VARMAX, VARMAXResults\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from pmdarima import auto_arima\n",
    "from statsmodels.tools.eval_measures import rmse\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "df = pd.read_csv(\"C:\\\\Users\\\\shaleens\\\\Desktop\\\\Projects\\\\TimeSeriesUdacity\\\\TS Udemy Vids\\\\Time Series 2\\\\VAR model dataset\\\\M2.csv\", index_col=0, parse_dates=True)\n",
    "df.index.freq = 'MS'\n",
    "\n",
    "sp = pd.read_csv(\"C:\\\\Users\\\\shaleens\\\\Desktop\\\\Projects\\\\TimeSeriesUdacity\\\\TS Udemy Vids\\\\Time Series 2\\\\VAR model dataset\\\\PCE.csv\", index_col=0, parse_dates=True)\n",
    "sp.index.freq = 'MS'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.rename(columns={'M2NS':'Money'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "sp.rename(columns={'PCE':'Spend'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lets go ahead and combines the 2 dataframes:\n",
    "\n",
    "df = pd.merge(df, sp, how='inner', on='DATE')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Money</th>\n",
       "      <th>Spend</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>1995-01-01</td>\n",
       "      <td>3495.3</td>\n",
       "      <td>4851.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-02-01</td>\n",
       "      <td>3474.9</td>\n",
       "      <td>4850.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-03-01</td>\n",
       "      <td>3491.7</td>\n",
       "      <td>4885.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-04-01</td>\n",
       "      <td>3522.4</td>\n",
       "      <td>4890.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-05-01</td>\n",
       "      <td>3506.7</td>\n",
       "      <td>4933.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Money   Spend\n",
       "DATE                      \n",
       "1995-01-01  3495.3  4851.2\n",
       "1995-02-01  3474.9  4850.8\n",
       "1995-03-01  3491.7  4885.4\n",
       "1995-04-01  3522.4  4890.2\n",
       "1995-05-01  3506.7  4933.1"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(300, 2)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(300, 2)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dropna().shape \n",
    "# This confirms that there is no null value to be treated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.iloc[:252]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>Money</th>\n",
       "      <th>Spend</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>12058.7</td>\n",
       "      <td>12387.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>12126.3</td>\n",
       "      <td>12384.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>12159.8</td>\n",
       "      <td>12387.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>12286.2</td>\n",
       "      <td>12422.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>12416.1</td>\n",
       "      <td>12453.8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Money    Spend\n",
       "DATE                        \n",
       "2015-08-01  12058.7  12387.0\n",
       "2015-09-01  12126.3  12384.5\n",
       "2015-10-01  12159.8  12387.3\n",
       "2015-11-01  12286.2  12422.8\n",
       "2015-12-01  12416.1  12453.8"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b89e5b7688>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot(figsize=(12,8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Now, we will perform auto_arima prediction to identify optimal p and q orders:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ARIMA(maxiter=1000, method='lbfgs', order=(2, 2, 3), out_of_sample_size=0,\n",
       "      scoring='mse', scoring_args=None, seasonal_order=(0, 0, 0, 0),\n",
       "      start_params=None, suppress_warnings=False, trend=None,\n",
       "      with_intercept=True)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auto_arima(df['Money'], maxiter=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ARIMA(maxiter=1000, method='lbfgs', order=(1, 1, 2), out_of_sample_size=0,\n",
       "      scoring='mse', scoring_args=None, seasonal_order=(0, 0, 0, 0),\n",
       "      start_params=None, suppress_warnings=False, trend=None,\n",
       "      with_intercept=True)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auto_arima(df['Spend'], maxiter=1000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Since, auto_arima already informed us abt stationarity of series, we dont need to run Dickey-fuller!\n",
    "##### Rather directly diff the series to majke it Stationary!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Money</th>\n",
       "      <th>Spend</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>1995-03-01</td>\n",
       "      <td>37.2</td>\n",
       "      <td>35.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-04-01</td>\n",
       "      <td>13.9</td>\n",
       "      <td>-29.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-05-01</td>\n",
       "      <td>-46.4</td>\n",
       "      <td>38.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-06-01</td>\n",
       "      <td>54.5</td>\n",
       "      <td>1.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1995-07-01</td>\n",
       "      <td>-14.6</td>\n",
       "      <td>-51.7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Money  Spend\n",
       "DATE                    \n",
       "1995-03-01   37.2   35.0\n",
       "1995-04-01   13.9  -29.8\n",
       "1995-05-01  -46.4   38.1\n",
       "1995-06-01   54.5    1.5\n",
       "1995-07-01  -14.6  -51.7"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Using inbuilt function of pandas dataframe for differencing the series:\n",
    "\n",
    "df_transformed = df.diff().diff()\n",
    "df_transformed = df_transformed.dropna()\n",
    "df_transformed.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Notice that we have performed doubke differencing as suggested by auto_arima and we can see we have lost 2 rows of data i.e for Jan month and Feb month!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Now the columns in dataframe are completely stationary!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(250, 2)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_transformed.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Train-test split:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_obs = 12\n",
    "\n",
    "train = df_transformed[:-n_obs]\n",
    "test = df_transformed[-n_obs:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((238, 2), (12, 2))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.shape, test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Fit the VARMA model:"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "auto_arima for 2 series were suggesting (1,2) and (2,3) for val of (p,q). So, i tried both and used one which is giving leser AIC!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = VARMAX(train, order=(1,2), trend='c')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = model.fit(maxiter=1000, disp=False) #disp is display"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Statespace Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>   <td>['Money', 'Spend']</td> <th>  No. Observations:  </th>    <td>238</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>               <td>VARMA(1,2)</td>     <th>  Log Likelihood     </th> <td>-2389.242</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                     <td>+ intercept</td>    <th>  AIC                </th> <td>4812.484</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>             <td>Mon, 02 Mar 2020</td>  <th>  BIC                </th> <td>4871.512</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                 <td>23:09:55</td>      <th>  HQIC               </th> <td>4836.273</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>              <td>03-01-1995</td>     <th>                     </th>     <td> </td>    \n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                    <td>- 12-01-2014</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",
       "  <th>Ljung-Box (Q):</th>          <td>320.40, 32.04</td> <th>  Jarque-Bera (JB):  </th> <td>4.15, 72.57</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                 <td>0.00, 0.81</td>   <th>  Prob(JB):          </th> <td>0.13, 0.00</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th>  <td>3.76, 2.46</td>   <th>  Skew:              </th> <td>0.10, -0.15</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>     <td>0.00, 0.00</td>   <th>  Kurtosis:          </th> <td>3.62, 5.69</td> \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<caption>Results for equation Money</caption>\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>intercept</th>   <td>    0.3147</td> <td>    0.150</td> <td>    2.092</td> <td> 0.036</td> <td>    0.020</td> <td>    0.609</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.Money</th>    <td>   -0.4227</td> <td>    0.124</td> <td>   -3.419</td> <td> 0.001</td> <td>   -0.665</td> <td>   -0.180</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.Spend</th>    <td>   -0.7996</td> <td>    0.376</td> <td>   -2.128</td> <td> 0.033</td> <td>   -1.536</td> <td>   -0.063</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.e(Money)</th> <td>   -0.1771</td> <td>    5.063</td> <td>   -0.035</td> <td> 0.972</td> <td>  -10.100</td> <td>    9.746</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.e(Spend)</th> <td>    0.8396</td> <td>    2.836</td> <td>    0.296</td> <td> 0.767</td> <td>   -4.719</td> <td>    6.398</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L2.e(Money)</th> <td>   -0.8314</td> <td>    4.987</td> <td>   -0.167</td> <td> 0.868</td> <td>  -10.606</td> <td>    8.943</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L2.e(Spend)</th> <td>   -0.7936</td> <td>    2.698</td> <td>   -0.294</td> <td> 0.769</td> <td>   -6.081</td> <td>    4.494</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<caption>Results for equation Spend</caption>\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>intercept</th>   <td>    0.1467</td> <td>    0.499</td> <td>    0.294</td> <td> 0.769</td> <td>   -0.831</td> <td>    1.124</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.Money</th>    <td>    0.0635</td> <td>    0.081</td> <td>    0.786</td> <td> 0.432</td> <td>   -0.095</td> <td>    0.222</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.Spend</th>    <td>   -1.0875</td> <td>    0.121</td> <td>   -8.967</td> <td> 0.000</td> <td>   -1.325</td> <td>   -0.850</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.e(Money)</th> <td>   -0.0406</td> <td>    0.938</td> <td>   -0.043</td> <td> 0.965</td> <td>   -1.879</td> <td>    1.798</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L1.e(Spend)</th> <td>    0.1779</td> <td>    0.547</td> <td>    0.325</td> <td> 0.745</td> <td>   -0.895</td> <td>    1.251</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L2.e(Money)</th> <td>    0.0082</td> <td>    0.909</td> <td>    0.009</td> <td> 0.993</td> <td>   -1.773</td> <td>    1.790</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>L2.e(Spend)</th> <td>   -1.0045</td> <td>    0.497</td> <td>   -2.022</td> <td> 0.043</td> <td>   -1.978</td> <td>   -0.031</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<caption>Error covariance matrix</caption>\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>sqrt.var.Money</th>       <td>   39.6062</td> <td>   87.619</td> <td>    0.452</td> <td> 0.651</td> <td> -132.124</td> <td>  211.336</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sqrt.cov.Money.Spend</th> <td>   -9.5400</td> <td>   53.948</td> <td>   -0.177</td> <td> 0.860</td> <td> -115.277</td> <td>   96.197</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sqrt.var.Spend</th>       <td>   32.6334</td> <td>   19.575</td> <td>    1.667</td> <td> 0.095</td> <td>   -5.733</td> <td>   71.000</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",
       "                           Statespace Model Results                           \n",
       "==============================================================================\n",
       "Dep. Variable:     ['Money', 'Spend']   No. Observations:                  238\n",
       "Model:                     VARMA(1,2)   Log Likelihood               -2389.242\n",
       "                          + intercept   AIC                           4812.484\n",
       "Date:                Mon, 02 Mar 2020   BIC                           4871.512\n",
       "Time:                        23:09:55   HQIC                          4836.273\n",
       "Sample:                    03-01-1995                                         \n",
       "                         - 12-01-2014                                         \n",
       "Covariance Type:                  opg                                         \n",
       "===================================================================================\n",
       "Ljung-Box (Q):               320.40, 32.04   Jarque-Bera (JB):          4.15, 72.57\n",
       "Prob(Q):                        0.00, 0.81   Prob(JB):                   0.13, 0.00\n",
       "Heteroskedasticity (H):         3.76, 2.46   Skew:                      0.10, -0.15\n",
       "Prob(H) (two-sided):            0.00, 0.00   Kurtosis:                   3.62, 5.69\n",
       "                           Results for equation Money                          \n",
       "===============================================================================\n",
       "                  coef    std err          z      P>|z|      [0.025      0.975]\n",
       "-------------------------------------------------------------------------------\n",
       "intercept       0.3147      0.150      2.092      0.036       0.020       0.609\n",
       "L1.Money       -0.4227      0.124     -3.419      0.001      -0.665      -0.180\n",
       "L1.Spend       -0.7996      0.376     -2.128      0.033      -1.536      -0.063\n",
       "L1.e(Money)    -0.1771      5.063     -0.035      0.972     -10.100       9.746\n",
       "L1.e(Spend)     0.8396      2.836      0.296      0.767      -4.719       6.398\n",
       "L2.e(Money)    -0.8314      4.987     -0.167      0.868     -10.606       8.943\n",
       "L2.e(Spend)    -0.7936      2.698     -0.294      0.769      -6.081       4.494\n",
       "                           Results for equation Spend                          \n",
       "===============================================================================\n",
       "                  coef    std err          z      P>|z|      [0.025      0.975]\n",
       "-------------------------------------------------------------------------------\n",
       "intercept       0.1467      0.499      0.294      0.769      -0.831       1.124\n",
       "L1.Money        0.0635      0.081      0.786      0.432      -0.095       0.222\n",
       "L1.Spend       -1.0875      0.121     -8.967      0.000      -1.325      -0.850\n",
       "L1.e(Money)    -0.0406      0.938     -0.043      0.965      -1.879       1.798\n",
       "L1.e(Spend)     0.1779      0.547      0.325      0.745      -0.895       1.251\n",
       "L2.e(Money)     0.0082      0.909      0.009      0.993      -1.773       1.790\n",
       "L2.e(Spend)    -1.0045      0.497     -2.022      0.043      -1.978      -0.031\n",
       "                                Error covariance matrix                                 \n",
       "========================================================================================\n",
       "                           coef    std err          z      P>|z|      [0.025      0.975]\n",
       "----------------------------------------------------------------------------------------\n",
       "sqrt.var.Money          39.6062     87.619      0.452      0.651    -132.124     211.336\n",
       "sqrt.cov.Money.Spend    -9.5400     53.948     -0.177      0.860    -115.277      96.197\n",
       "sqrt.var.Spend          32.6334     19.575      1.667      0.095      -5.733      71.000\n",
       "========================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.summary()"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "VARMAX forecast() function wont require to pass in a number of previous obs in form of ndarray or something. It directly forecasts output!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_forecast = results.forecast(12) # steps is number of periods to predict."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>Money</th>\n",
       "      <th>Spend</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>-80.624140</td>\n",
       "      <td>16.801971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>-5.900099</td>\n",
       "      <td>16.418502</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>-10.320018</td>\n",
       "      <td>-18.082846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>19.136004</td>\n",
       "      <td>19.155494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>-23.090683</td>\n",
       "      <td>-19.468230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-06-01</td>\n",
       "      <td>25.641614</td>\n",
       "      <td>19.850498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-07-01</td>\n",
       "      <td>-26.396176</td>\n",
       "      <td>-19.810603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>27.312529</td>\n",
       "      <td>20.012759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>-27.232175</td>\n",
       "      <td>-19.880872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>27.722073</td>\n",
       "      <td>20.036047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>-27.423901</td>\n",
       "      <td>-19.880172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>27.802550</td>\n",
       "      <td>20.023102</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Money      Spend\n",
       "2015-01-01 -80.624140  16.801971\n",
       "2015-02-01  -5.900099  16.418502\n",
       "2015-03-01 -10.320018 -18.082846\n",
       "2015-04-01  19.136004  19.155494\n",
       "2015-05-01 -23.090683 -19.468230\n",
       "2015-06-01  25.641614  19.850498\n",
       "2015-07-01 -26.396176 -19.810603\n",
       "2015-08-01  27.312529  20.012759\n",
       "2015-09-01 -27.232175 -19.880872\n",
       "2015-10-01  27.722073  20.036047\n",
       "2015-11-01 -27.423901 -19.880172\n",
       "2015-12-01  27.802550  20.023102"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_forecast"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "But, remember the pred value received is in the 2nd difference!\n",
    "We need to get the original data and then plot out how well we are performing!\n",
    "For this we need to essentially reverse the second difference and then reverse the 1st difference!\n",
    "i.e. Inverting the transformation!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Inverting the Transformation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Money', 'Spend'], dtype='object')"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_forecast.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_forecast['Money_1d'] = (df['Money'].iloc[-n_obs-1] - df['Money'].iloc[-n_obs-2]) + df_forecast['Money'].cumsum()\n",
    "\n",
    "df_forecast['MoneyForecast'] = df['Money'].iloc[-n_obs-1] + df_forecast['Money_1d'].cumsum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b8a1fdb388>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_forecast['MoneyForecast'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_forecast['Spend_1d'] = (df['Spend'].iloc[-n_obs-1] - df['Spend'].iloc[-n_obs-2]) + df_forecast['Spend'].cumsum()\n",
    "\n",
    "df_forecast['SpendForecast'] = df['Spend'].iloc[-n_obs-1] + df_forecast['Spend_1d'].cumsum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b8a359cfc8>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_forecast['SpendForecast'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>Money</th>\n",
       "      <th>Spend</th>\n",
       "      <th>Money_1d</th>\n",
       "      <th>MoneyForecast</th>\n",
       "      <th>Spend_1d</th>\n",
       "      <th>SpendForecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>-80.624140</td>\n",
       "      <td>16.801971</td>\n",
       "      <td>64.475860</td>\n",
       "      <td>11812.375860</td>\n",
       "      <td>31.601971</td>\n",
       "      <td>12096.601971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>-5.900099</td>\n",
       "      <td>16.418502</td>\n",
       "      <td>58.575761</td>\n",
       "      <td>11870.951621</td>\n",
       "      <td>48.020473</td>\n",
       "      <td>12144.622444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>-10.320018</td>\n",
       "      <td>-18.082846</td>\n",
       "      <td>48.255743</td>\n",
       "      <td>11919.207363</td>\n",
       "      <td>29.937628</td>\n",
       "      <td>12174.560072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>19.136004</td>\n",
       "      <td>19.155494</td>\n",
       "      <td>67.391747</td>\n",
       "      <td>11986.599110</td>\n",
       "      <td>49.093121</td>\n",
       "      <td>12223.653193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>-23.090683</td>\n",
       "      <td>-19.468230</td>\n",
       "      <td>44.301063</td>\n",
       "      <td>12030.900174</td>\n",
       "      <td>29.624892</td>\n",
       "      <td>12253.278085</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Money      Spend   Money_1d  MoneyForecast   Spend_1d  \\\n",
       "2015-01-01 -80.624140  16.801971  64.475860   11812.375860  31.601971   \n",
       "2015-02-01  -5.900099  16.418502  58.575761   11870.951621  48.020473   \n",
       "2015-03-01 -10.320018 -18.082846  48.255743   11919.207363  29.937628   \n",
       "2015-04-01  19.136004  19.155494  67.391747   11986.599110  49.093121   \n",
       "2015-05-01 -23.090683 -19.468230  44.301063   12030.900174  29.624892   \n",
       "\n",
       "            SpendForecast  \n",
       "2015-01-01   12096.601971  \n",
       "2015-02-01   12144.622444  \n",
       "2015-03-01   12174.560072  \n",
       "2015-04-01   12223.653193  \n",
       "2015-05-01   12253.278085  "
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_forecast.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>Money</th>\n",
       "      <th>Spend</th>\n",
       "      <th>MoneyForecast</th>\n",
       "      <th>SpendForecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>11746.9</td>\n",
       "      <td>12039.9</td>\n",
       "      <td>11812.375860</td>\n",
       "      <td>12096.601971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>11835.1</td>\n",
       "      <td>12089.1</td>\n",
       "      <td>11870.951621</td>\n",
       "      <td>12144.622444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>11944.9</td>\n",
       "      <td>12145.7</td>\n",
       "      <td>11919.207363</td>\n",
       "      <td>12174.560072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>11992.7</td>\n",
       "      <td>12187.4</td>\n",
       "      <td>11986.599110</td>\n",
       "      <td>12223.653193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>11896.1</td>\n",
       "      <td>12261.1</td>\n",
       "      <td>12030.900174</td>\n",
       "      <td>12253.278085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-06-01</td>\n",
       "      <td>11950.7</td>\n",
       "      <td>12295.6</td>\n",
       "      <td>12100.842851</td>\n",
       "      <td>12302.753475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-07-01</td>\n",
       "      <td>12001.3</td>\n",
       "      <td>12357.1</td>\n",
       "      <td>12144.389353</td>\n",
       "      <td>12332.418262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>12058.7</td>\n",
       "      <td>12387.0</td>\n",
       "      <td>12215.248384</td>\n",
       "      <td>12382.095807</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>12126.3</td>\n",
       "      <td>12384.5</td>\n",
       "      <td>12258.875239</td>\n",
       "      <td>12411.892481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>12159.8</td>\n",
       "      <td>12387.3</td>\n",
       "      <td>12330.224168</td>\n",
       "      <td>12461.725202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>12286.2</td>\n",
       "      <td>12422.8</td>\n",
       "      <td>12374.149195</td>\n",
       "      <td>12491.677752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>12416.1</td>\n",
       "      <td>12453.8</td>\n",
       "      <td>12445.876773</td>\n",
       "      <td>12541.653403</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Money    Spend  MoneyForecast  SpendForecast\n",
       "2015-01-01  11746.9  12039.9   11812.375860   12096.601971\n",
       "2015-02-01  11835.1  12089.1   11870.951621   12144.622444\n",
       "2015-03-01  11944.9  12145.7   11919.207363   12174.560072\n",
       "2015-04-01  11992.7  12187.4   11986.599110   12223.653193\n",
       "2015-05-01  11896.1  12261.1   12030.900174   12253.278085\n",
       "2015-06-01  11950.7  12295.6   12100.842851   12302.753475\n",
       "2015-07-01  12001.3  12357.1   12144.389353   12332.418262\n",
       "2015-08-01  12058.7  12387.0   12215.248384   12382.095807\n",
       "2015-09-01  12126.3  12384.5   12258.875239   12411.892481\n",
       "2015-10-01  12159.8  12387.3   12330.224168   12461.725202\n",
       "2015-11-01  12286.2  12422.8   12374.149195   12491.677752\n",
       "2015-12-01  12416.1  12453.8   12445.876773   12541.653403"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([df.iloc[-12:], df_forecast[['MoneyForecast','SpendForecast']]], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b8a8628508>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['Money'][-n_obs:].plot(figsize=(12,5), legend=True)\n",
    "df_forecast['MoneyForecast'].plot(legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b8a866e188>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['Spend'][-n_obs:].plot(figsize=(12,5),legend=True)\n",
    "df_forecast['SpendForecast'].plot(legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1b8a77ee0c8>"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_forecast['MoneyForecast'].rename('forecast money').plot(legend=True, figsize=(16,12))\n",
    "df_forecast['SpendForecast'].rename('forecast spend').plot(legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "110.75799042535651"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(df['Money'][-n_obs:], df_forecast['MoneyForecast'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12034.566666666668"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Money'][-n_obs:].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "48.254940064250626"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(df['Spend'][-n_obs:], df_forecast['SpendForecast'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### We can see we managed to reduce the error in the Money column upto great extent! However, there is a very very slight increase in error of Spend as compared to VAR model!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Still, we can say VARMA is performing pretty decent on our dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Lets try and compare these results to individual ARMA(1,2) model\n",
    "##### i.e. we are assuming no relationship between the two series!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.tsa.arima_model import ARMA, ARMAResults"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Money:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>ARMA Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>       <td>Money</td>      <th>  No. Observations:  </th>    <td>238</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>            <td>ARMA(1, 2)</td>    <th>  Log Likelihood     </th> <td>-1234.869</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>css-mle</td>     <th>  S.D. of innovations</th>  <td>42.880</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>          <td>Mon, 02 Mar 2020</td> <th>  AIC                </th> <td>2479.738</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>              <td>23:36:50</td>     <th>  BIC                </th> <td>2497.099</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>           <td>03-01-1995</td>    <th>  HQIC               </th> <td>2486.735</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                 <td>- 12-01-2014</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>const</th>       <td>    0.1902</td> <td>    0.045</td> <td>    4.266</td> <td> 0.000</td> <td>    0.103</td> <td>    0.278</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L1.Money</th> <td>   -0.4459</td> <td>    0.129</td> <td>   -3.465</td> <td> 0.001</td> <td>   -0.698</td> <td>   -0.194</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1.Money</th> <td>   -0.3950</td> <td>    0.105</td> <td>   -3.754</td> <td> 0.000</td> <td>   -0.601</td> <td>   -0.189</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L2.Money</th> <td>   -0.6050</td> <td>    0.105</td> <td>   -5.775</td> <td> 0.000</td> <td>   -0.810</td> <td>   -0.400</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<caption>Roots</caption>\n",
       "<tr>\n",
       "    <td></td>   <th>            Real</th>  <th>         Imaginary</th> <th>         Modulus</th>  <th>        Frequency</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>AR.1</th> <td>          -2.2426</td> <td>          +0.0000j</td> <td>           2.2426</td> <td>           0.5000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>MA.1</th> <td>           1.0000</td> <td>          +0.0000j</td> <td>           1.0000</td> <td>           0.0000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>MA.2</th> <td>          -1.6528</td> <td>          +0.0000j</td> <td>           1.6528</td> <td>           0.5000</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                              ARMA Model Results                              \n",
       "==============================================================================\n",
       "Dep. Variable:                  Money   No. Observations:                  238\n",
       "Model:                     ARMA(1, 2)   Log Likelihood               -1234.869\n",
       "Method:                       css-mle   S.D. of innovations             42.880\n",
       "Date:                Mon, 02 Mar 2020   AIC                           2479.738\n",
       "Time:                        23:36:50   BIC                           2497.099\n",
       "Sample:                    03-01-1995   HQIC                          2486.735\n",
       "                         - 12-01-2014                                         \n",
       "===============================================================================\n",
       "                  coef    std err          z      P>|z|      [0.025      0.975]\n",
       "-------------------------------------------------------------------------------\n",
       "const           0.1902      0.045      4.266      0.000       0.103       0.278\n",
       "ar.L1.Money    -0.4459      0.129     -3.465      0.001      -0.698      -0.194\n",
       "ma.L1.Money    -0.3950      0.105     -3.754      0.000      -0.601      -0.189\n",
       "ma.L2.Money    -0.6050      0.105     -5.775      0.000      -0.810      -0.400\n",
       "                                    Roots                                    \n",
       "=============================================================================\n",
       "                  Real          Imaginary           Modulus         Frequency\n",
       "-----------------------------------------------------------------------------\n",
       "AR.1           -2.2426           +0.0000j            2.2426            0.5000\n",
       "MA.1            1.0000           +0.0000j            1.0000            0.0000\n",
       "MA.2           -1.6528           +0.0000j            1.6528            0.5000\n",
       "-----------------------------------------------------------------------------\n",
       "\"\"\""
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = ARMA(train['Money'], order=(1,2))\n",
    "results = model.fit()\n",
    "results.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "start = len(train)\n",
    "end = len(train) + len(test) - 1\n",
    "\n",
    "z1 = results.predict(start=start, end=end).rename('Money')\n",
    "z1 = pd.DataFrame(z1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>Money</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>-75.021344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>-18.790995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>8.653972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>-3.583805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>1.873049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-06-01</td>\n",
       "      <td>-0.560175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-07-01</td>\n",
       "      <td>0.524805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>0.041010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>0.256736</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>0.160543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>0.203436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>0.184310</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Money\n",
       "2015-01-01 -75.021344\n",
       "2015-02-01 -18.790995\n",
       "2015-03-01   8.653972\n",
       "2015-04-01  -3.583805\n",
       "2015-05-01   1.873049\n",
       "2015-06-01  -0.560175\n",
       "2015-07-01   0.524805\n",
       "2015-08-01   0.041010\n",
       "2015-09-01   0.256736\n",
       "2015-10-01   0.160543\n",
       "2015-11-01   0.203436\n",
       "2015-12-01   0.184310"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "z1['Money1d'] = (df['Money'].iloc[-n_obs-1] - df['Money'].iloc[-n_obs-2]) + z1['Money'].cumsum()\n",
    "\n",
    "z1['MoneyForecast'] = df['Money'].iloc[-n_obs-1] + z1['Money1d'].cumsum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>Money</th>\n",
       "      <th>Money1d</th>\n",
       "      <th>MoneyForecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>-75.021344</td>\n",
       "      <td>70.078656</td>\n",
       "      <td>11817.978656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>-18.790995</td>\n",
       "      <td>51.287662</td>\n",
       "      <td>11869.266318</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>8.653972</td>\n",
       "      <td>59.941633</td>\n",
       "      <td>11929.207951</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>-3.583805</td>\n",
       "      <td>56.357829</td>\n",
       "      <td>11985.565780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>1.873049</td>\n",
       "      <td>58.230878</td>\n",
       "      <td>12043.796658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-06-01</td>\n",
       "      <td>-0.560175</td>\n",
       "      <td>57.670703</td>\n",
       "      <td>12101.467361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-07-01</td>\n",
       "      <td>0.524805</td>\n",
       "      <td>58.195508</td>\n",
       "      <td>12159.662869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>0.041010</td>\n",
       "      <td>58.236518</td>\n",
       "      <td>12217.899387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>0.256736</td>\n",
       "      <td>58.493254</td>\n",
       "      <td>12276.392641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>0.160543</td>\n",
       "      <td>58.653797</td>\n",
       "      <td>12335.046438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>0.203436</td>\n",
       "      <td>58.857232</td>\n",
       "      <td>12393.903670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>0.184310</td>\n",
       "      <td>59.041542</td>\n",
       "      <td>12452.945213</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Money    Money1d  MoneyForecast\n",
       "2015-01-01 -75.021344  70.078656   11817.978656\n",
       "2015-02-01 -18.790995  51.287662   11869.266318\n",
       "2015-03-01   8.653972  59.941633   11929.207951\n",
       "2015-04-01  -3.583805  56.357829   11985.565780\n",
       "2015-05-01   1.873049  58.230878   12043.796658\n",
       "2015-06-01  -0.560175  57.670703   12101.467361\n",
       "2015-07-01   0.524805  58.195508   12159.662869\n",
       "2015-08-01   0.041010  58.236518   12217.899387\n",
       "2015-09-01   0.256736  58.493254   12276.392641\n",
       "2015-10-01   0.160543  58.653797   12335.046438\n",
       "2015-11-01   0.203436  58.857232   12393.903670\n",
       "2015-12-01   0.184310  59.041542   12452.945213"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "118.20947449892496"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(df['Money'][-n_obs:], z1['MoneyForecast'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Now, for Spend column:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>ARMA Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>       <td>Spend</td>      <th>  No. Observations:  </th>    <td>238</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>            <td>ARMA(1, 2)</td>    <th>  Log Likelihood     </th> <td>-1182.737</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>css-mle</td>     <th>  S.D. of innovations</th>  <td>34.709</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>          <td>Mon, 02 Mar 2020</td> <th>  AIC                </th> <td>2375.474</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>              <td>23:43:28</td>     <th>  BIC                </th> <td>2392.835</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>           <td>03-01-1995</td>    <th>  HQIC               </th> <td>2382.471</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                 <td>- 12-01-2014</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>const</th>       <td>    0.0881</td> <td>    0.245</td> <td>    0.359</td> <td> 0.720</td> <td>   -0.393</td> <td>    0.569</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.L1.Spend</th> <td>   -0.3246</td> <td>    0.501</td> <td>   -0.648</td> <td> 0.517</td> <td>   -1.307</td> <td>    0.657</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L1.Spend</th> <td>   -0.6638</td> <td>    0.511</td> <td>   -1.300</td> <td> 0.194</td> <td>   -1.665</td> <td>    0.337</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.L2.Spend</th> <td>   -0.1966</td> <td>    0.475</td> <td>   -0.414</td> <td> 0.679</td> <td>   -1.128</td> <td>    0.735</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<caption>Roots</caption>\n",
       "<tr>\n",
       "    <td></td>   <th>            Real</th>  <th>         Imaginary</th> <th>         Modulus</th>  <th>        Frequency</th>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>AR.1</th> <td>          -3.0809</td> <td>          +0.0000j</td> <td>           3.0809</td> <td>           0.5000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>MA.1</th> <td>           1.1290</td> <td>          +0.0000j</td> <td>           1.1290</td> <td>           0.0000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>MA.2</th> <td>          -4.5046</td> <td>          +0.0000j</td> <td>           4.5046</td> <td>           0.5000</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                              ARMA Model Results                              \n",
       "==============================================================================\n",
       "Dep. Variable:                  Spend   No. Observations:                  238\n",
       "Model:                     ARMA(1, 2)   Log Likelihood               -1182.737\n",
       "Method:                       css-mle   S.D. of innovations             34.709\n",
       "Date:                Mon, 02 Mar 2020   AIC                           2375.474\n",
       "Time:                        23:43:28   BIC                           2392.835\n",
       "Sample:                    03-01-1995   HQIC                          2382.471\n",
       "                         - 12-01-2014                                         \n",
       "===============================================================================\n",
       "                  coef    std err          z      P>|z|      [0.025      0.975]\n",
       "-------------------------------------------------------------------------------\n",
       "const           0.0881      0.245      0.359      0.720      -0.393       0.569\n",
       "ar.L1.Spend    -0.3246      0.501     -0.648      0.517      -1.307       0.657\n",
       "ma.L1.Spend    -0.6638      0.511     -1.300      0.194      -1.665       0.337\n",
       "ma.L2.Spend    -0.1966      0.475     -0.414      0.679      -1.128       0.735\n",
       "                                    Roots                                    \n",
       "=============================================================================\n",
       "                  Real          Imaginary           Modulus         Frequency\n",
       "-----------------------------------------------------------------------------\n",
       "AR.1           -3.0809           +0.0000j            3.0809            0.5000\n",
       "MA.1            1.1290           +0.0000j            1.1290            0.0000\n",
       "MA.2           -4.5046           +0.0000j            4.5046            0.5000\n",
       "-----------------------------------------------------------------------------\n",
       "\"\"\""
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = ARMA(train['Spend'], order=(1,2))\n",
    "results = model.fit()\n",
    "results.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>Spend</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>29.743269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>-3.064423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>1.111292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>-0.244044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>0.195865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-06-01</td>\n",
       "      <td>0.053081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-07-01</td>\n",
       "      <td>0.099425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>0.084383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>0.089265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>0.087681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>0.088195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>0.088028</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Spend\n",
       "2015-01-01  29.743269\n",
       "2015-02-01  -3.064423\n",
       "2015-03-01   1.111292\n",
       "2015-04-01  -0.244044\n",
       "2015-05-01   0.195865\n",
       "2015-06-01   0.053081\n",
       "2015-07-01   0.099425\n",
       "2015-08-01   0.084383\n",
       "2015-09-01   0.089265\n",
       "2015-10-01   0.087681\n",
       "2015-11-01   0.088195\n",
       "2015-12-01   0.088028"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "start = len(train)\n",
    "end = len(train) + len(test) - 1\n",
    "\n",
    "z2 = results.predict(start=start, end=end).rename('Spend')\n",
    "z2 = pd.DataFrame(z2)\n",
    "z2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Money', 'Spend'], dtype='object')"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "z2['Spend1d'] = (df['Spend'].iloc[-n_obs-1] - df['Spend'].iloc[-n_obs-2]) + z2['Spend'].cumsum()\n",
    "z2['SpendForecast'] = df['Spend'].iloc[-n_obs-1] + z2['Spend1d'].cumsum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "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>Spend</th>\n",
       "      <th>Spend1d</th>\n",
       "      <th>SpendForecast</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>2015-01-01</td>\n",
       "      <td>29.743269</td>\n",
       "      <td>44.543269</td>\n",
       "      <td>12109.543269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-02-01</td>\n",
       "      <td>-3.064423</td>\n",
       "      <td>41.478846</td>\n",
       "      <td>12151.022115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-03-01</td>\n",
       "      <td>1.111292</td>\n",
       "      <td>42.590138</td>\n",
       "      <td>12193.612253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-04-01</td>\n",
       "      <td>-0.244044</td>\n",
       "      <td>42.346093</td>\n",
       "      <td>12235.958346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-05-01</td>\n",
       "      <td>0.195865</td>\n",
       "      <td>42.541959</td>\n",
       "      <td>12278.500305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-06-01</td>\n",
       "      <td>0.053081</td>\n",
       "      <td>42.595040</td>\n",
       "      <td>12321.095345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-07-01</td>\n",
       "      <td>0.099425</td>\n",
       "      <td>42.694465</td>\n",
       "      <td>12363.789810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-08-01</td>\n",
       "      <td>0.084383</td>\n",
       "      <td>42.778848</td>\n",
       "      <td>12406.568658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-09-01</td>\n",
       "      <td>0.089265</td>\n",
       "      <td>42.868114</td>\n",
       "      <td>12449.436772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-10-01</td>\n",
       "      <td>0.087681</td>\n",
       "      <td>42.955795</td>\n",
       "      <td>12492.392566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-11-01</td>\n",
       "      <td>0.088195</td>\n",
       "      <td>43.043990</td>\n",
       "      <td>12535.436556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2015-12-01</td>\n",
       "      <td>0.088028</td>\n",
       "      <td>43.132018</td>\n",
       "      <td>12578.568574</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Spend    Spend1d  SpendForecast\n",
       "2015-01-01  29.743269  44.543269   12109.543269\n",
       "2015-02-01  -3.064423  41.478846   12151.022115\n",
       "2015-03-01   1.111292  42.590138   12193.612253\n",
       "2015-04-01  -0.244044  42.346093   12235.958346\n",
       "2015-05-01   0.195865  42.541959   12278.500305\n",
       "2015-06-01   0.053081  42.595040   12321.095345\n",
       "2015-07-01   0.099425  42.694465   12363.789810\n",
       "2015-08-01   0.084383  42.778848   12406.568658\n",
       "2015-09-01   0.089265  42.868114   12449.436772\n",
       "2015-10-01   0.087681  42.955795   12492.392566\n",
       "2015-11-01   0.088195  43.043990   12535.436556\n",
       "2015-12-01   0.088028  43.132018   12578.568574"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "69.66223388702355"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rmse(df['Spend'][-n_obs:], z2['SpendForecast'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### We can see that our VARMA is performing better than ARMA! which also indicates some relation between variables!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### But one thing to observe is:\n",
    "\n",
    "There is not a very heavy dependency between money and consumption because the results obtained from ARMA and VARMA are similar (with VARMA performing slightly better). SO, there is a slight dependency between them atleast for the dataset we investigated!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Advice for future:\n",
    "\n",
    "Its always good to run simpler models like ARMA before moving to complex models like VARMA."
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}