6. LassoModel.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Lasso Regression Model\n",
    "\n",
    "This code tests lags for sentiment to test which ones are most correlated with "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from datetime import datetime\n",
    "from sklearn.linear_model import Lasso\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "df = pd.read_pickle('./IntermediatePickles/model_df.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Make Data Stationary\n",
    "by taking the difference (we are interested in rate-of-change anyway)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['GDP_interp'] = df.GDP_interp.diff()\n",
    "df['sentiment'] = df.sentiment.diff()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create Lags\n",
    "\n",
    "Shift GDP by one and two quarters (data is quarterly)\n",
    "\n",
    "Shift sentiment by 36 months"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['GDP_interp_L4'] = df.GDP_interp.shift(4)\n",
    "#df['GDP_interp_L8'] = df.GDP_interp.shift(8)\n",
    "for i in range(36):\n",
    "    df['sentiment_L' + str(i)] = df.sentiment.shift(i)\n",
    "df.dropna(inplace=True)\n",
    "df.drop(columns='sentiment_L0',inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define alpha\n",
    "\n",
    "This will get increased until we only have one coefficient"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha = 3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define pipeline\n",
    "Lasso regression requires standard scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "pipe = Pipeline([\n",
    "    ('scaler',StandardScaler()),\n",
    "    ('lasso', Lasso(alpha = alpha,max_iter=100000))\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define Cutoff\n",
    "This is the point we will make predictions to test the accuracy of the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "cutoff = datetime.strptime('2014-03-01','%Y-%m-%d')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df = df.loc[df.chart_date > cutoff]\n",
    "train_df = df.loc[df.chart_date <= cutoff]\n",
    "X_train = train_df.drop(columns=['chart_date','GDP_interp'])\n",
    "X_test = test_df.drop(columns=['chart_date','GDP_interp'])\n",
    "y_train = pd.DataFrame(train_df.GDP_interp)\n",
    "y_test = pd.DataFrame(test_df.GDP_interp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Fit Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "lasso_fit = pipe.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Test Model"
   ]
  },
  {
   "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>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>sentiment</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>GDP_interp_L4</td>\n",
       "      <td>25.118756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>sentiment_L1</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>sentiment_L2</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>sentiment_L3</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>sentiment_L4</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>sentiment_L5</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sentiment_L6</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>sentiment_L7</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>sentiment_L8</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>sentiment_L9</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>sentiment_L10</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>sentiment_L11</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>sentiment_L12</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>sentiment_L13</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>sentiment_L14</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>sentiment_L15</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>sentiment_L16</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>sentiment_L17</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>sentiment_L18</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>sentiment_L19</td>\n",
       "      <td>0.094669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>sentiment_L20</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>sentiment_L21</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>sentiment_L22</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>sentiment_L23</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>sentiment_L24</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>sentiment_L25</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>sentiment_L26</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>sentiment_L27</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>sentiment_L28</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>sentiment_L29</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>sentiment_L30</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>sentiment_L31</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>sentiment_L32</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>sentiment_L33</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>sentiment_L34</td>\n",
       "      <td>-0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>sentiment_L35</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                0          1\n",
       "0       sentiment   0.000000\n",
       "1   GDP_interp_L4  25.118756\n",
       "2    sentiment_L1  -0.000000\n",
       "3    sentiment_L2   0.000000\n",
       "4    sentiment_L3   0.000000\n",
       "5    sentiment_L4   0.000000\n",
       "6    sentiment_L5  -0.000000\n",
       "7    sentiment_L6  -0.000000\n",
       "8    sentiment_L7   0.000000\n",
       "9    sentiment_L8  -0.000000\n",
       "10   sentiment_L9   0.000000\n",
       "11  sentiment_L10   0.000000\n",
       "12  sentiment_L11  -0.000000\n",
       "13  sentiment_L12  -0.000000\n",
       "14  sentiment_L13  -0.000000\n",
       "15  sentiment_L14   0.000000\n",
       "16  sentiment_L15   0.000000\n",
       "17  sentiment_L16  -0.000000\n",
       "18  sentiment_L17   0.000000\n",
       "19  sentiment_L18  -0.000000\n",
       "20  sentiment_L19   0.094669\n",
       "21  sentiment_L20  -0.000000\n",
       "22  sentiment_L21  -0.000000\n",
       "23  sentiment_L22   0.000000\n",
       "24  sentiment_L23  -0.000000\n",
       "25  sentiment_L24  -0.000000\n",
       "26  sentiment_L25  -0.000000\n",
       "27  sentiment_L26  -0.000000\n",
       "28  sentiment_L27  -0.000000\n",
       "29  sentiment_L28  -0.000000\n",
       "30  sentiment_L29   0.000000\n",
       "31  sentiment_L30  -0.000000\n",
       "32  sentiment_L31   0.000000\n",
       "33  sentiment_L32  -0.000000\n",
       "34  sentiment_L33   0.000000\n",
       "35  sentiment_L34  -0.000000\n",
       "36  sentiment_L35   0.000000"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(list(zip(list(X_train.columns),lasso_fit.named_steps.lasso.coef_)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Results\n",
    "Sentiment with a 19-month lag is the most significant result"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.12 ('base')",
   "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": "3.9.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "3f0653bd30555679cc60e5853b2204587bc8607f0666d413a468e937b04317b4"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}