RDD_data_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d5bc112b",
   "metadata": {},
   "source": [
    "# Performing RDD analysis\n",
    "\n",
    "- Performing t-tests and analyzing peaks\n",
    "- Performing RDD without and with covariates "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6bf511c7",
   "metadata": {},
   "source": [
    "# Packages to import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f5720322",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from datetime import date\n",
    "import datetime\n",
    "import statsmodels.formula.api as smf\n",
    "import statsmodels.api as sm\n",
    "from tqdm.auto import tqdm\n",
    "from scipy.stats import ttest_ind\n",
    "import seaborn as sns\n",
    "from sklearn.linear_model import LinearRegression"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10b1dde5",
   "metadata": {},
   "source": [
    "# Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "0d9842b3",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def event_occured(date, cutoff_date):\n",
    "    if(date>=np.datetime64(cutoff_date)):\n",
    "        return 1\n",
    "    else:\n",
    "        return 0\n",
    "\n",
    "def days_from_event(x, cutoff_date):\n",
    "    d1 = x\n",
    "    d0 = datetime.datetime.strptime(cutoff_date,'%Y-%m-%d').date()\n",
    "    delta = d1 - d0\n",
    "    return delta.days"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33de67e1",
   "metadata": {},
   "source": [
    "# Data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6690493",
   "metadata": {},
   "outputs": [],
   "source": [
    "tweets = pd.read_pickle('/Users/deryadurmush/Desktop/Jobs /balenciaga/preprocessed_tweets.pkl')\n",
    "df_en = tweets[tweets['lang']=='en']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bcb57166",
   "metadata": {},
   "source": [
    "# Detecting jumps with t-test \n",
    "\n",
    "The RDD window usually happens to be a 1-day one which makes us go from days to hours in terms of time intervals. This is an unreliable time interval since different news sources have press-releases at different times of the day. Thus, we were forced to set the intervals equal to 1 day.\n",
    "\n",
    "To solve the issues above, we first detect the jumps using the t-tests. We construct the 2-sample t-test for every two days in a row and compare the brand perception metrics for these two days. In case we find that the t-statistic is significant, we mark the second day as the day with a jump. Figure 4 shows the graph of the t-statistic changes in time. The orange and the green lines represent the cut-offs, under which the t-statistic is insignificant. We set the cut-off to be symmetric and equal to 20 in order to filter 14 most outstanding PR moves or failures."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a3464da5",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_en['date'] = df_en['created_at'].apply(lambda x: x.date())\n",
    "dates = sorted(df_en['date'].unique().tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "73ab3d6f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "260f7e2b7ba848df9f4eb85b12c23d48",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/161 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## compound is a the sentiment score\n",
    "res = []\n",
    "\n",
    "for i in tqdm(range(len(dates)-1)):\n",
    "    date1 = dates[i]\n",
    "    date2 = dates[i+1]\n",
    "    \n",
    "    df1 = df_en[df_en['date'] == date1]\n",
    "    df2 = df_en[df_en['date'] == date2]\n",
    "    \n",
    "    res.append([date2, ttest_ind(df1['compound'], \n",
    "                                df2['compound'], equal_var=False)])\n",
    "    \n",
    "res_ = pd.DataFrame(res)\n",
    "res_.columns = ['jump_date', 'test_results']\n",
    "res_['t-stat'] = res_['test_results'].apply(lambda x: x[0])\n",
    "res_['p-val'] = res_['test_results'].apply(lambda x: x[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "ecec6861",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='jump_date', ylabel='t-stat'>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "res_['line1'] = 20\n",
    "res_['line2'] = -20\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "\n",
    "sns.lineplot(data=res_, x='jump_date', y='t-stat')\n",
    "sns.lineplot(data=res_, x='jump_date', y='line1')\n",
    "sns.lineplot(data=res_, x='jump_date', y='line2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "4b38c36c",
   "metadata": {},
   "outputs": [],
   "source": [
    "jump_dates = [datetime.date(2022, 11, 1)] + \\\n",
    "            res_[(res_['t-stat']>20)|(res_['t-stat']<-20)]['jump_date'].unique().tolist() + \\\n",
    "                [datetime.date(2023, 4, 10)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "ff99f75b",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_en['created_at_date'] = df_en['created_at'].dt.date"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2364314",
   "metadata": {},
   "source": [
    "# Analyzing a linear regression model for each sentiment score for each interval between events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "a7595e92",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 0\n",
      "0 1\n",
      "1 0\n",
      "1 1\n",
      "2 0\n",
      "2 1\n",
      "3 0\n",
      "3 1\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x1440 with 8 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 720x432 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(4, 2, figsize=(10,20))\n",
    "counter1 = 0\n",
    "counter2 = 0\n",
    "\n",
    "targets = ['compound', 'pos', 'neg', 'Angry', 'Sad', 'Surprise', 'Fear', 'Happy']\n",
    "\n",
    "for target in targets:\n",
    "\n",
    "    slopes = []\n",
    "\n",
    "    for i in range(1, len(jump_dates)):\n",
    "        date0 = jump_dates[i - 1]\n",
    "        date1 = jump_dates[i]\n",
    "\n",
    "        fin_df = df_en[(df_en['created_at_date'] < date1)&(df_en['created_at_date'] >= date0)]\n",
    "        dates = pd.DataFrame({'created_at_date':sorted(fin_df['created_at_date'].unique())}).reset_index()\n",
    "        fin_df = fin_df.merge(dates, on = 'created_at_date')\n",
    "        lr = LinearRegression()\n",
    "        lr.fit(fin_df[['index']], fin_df[target])\n",
    "\n",
    "        slopes.append([date0, date1, lr.coef_[0], lr.intercept_])\n",
    "\n",
    "    slopes_ = pd.DataFrame(slopes)\n",
    "    slopes_.columns = ['start', 'end', 'slope', 'intercept']\n",
    "\n",
    "    dates = pd.DataFrame({'created_at_date':sorted(df_en['created_at_date'].unique())})\n",
    "\n",
    "    for_graph = []\n",
    "\n",
    "    for i in range(dates.shape[0]):\n",
    "        for j in range(slopes_.shape[0]):\n",
    "            if dates['created_at_date'][i] < slopes_['end'][j] and dates['created_at_date'][i] >= slopes_['start'][j]:\n",
    "                diff = int(str(np.datetime64(dates['created_at_date'][i]) - \\\n",
    "                               np.datetime64(slopes_['start'][j])).replace(' days', ''))\n",
    "                y = slopes_['slope'][j] * diff + slopes_['intercept'][j]\n",
    "                for_graph.append([dates['created_at_date'][i], y])\n",
    "\n",
    "    for_graph = pd.DataFrame(for_graph)\n",
    "    for_graph.columns = ['date', target]\n",
    "\n",
    "    plt.figure(figsize=(10,6))\n",
    "\n",
    "    sns.lineplot(data=for_graph, x='date', y=target, ax = axs[counter1][counter2])\n",
    "    axs[counter1][counter2].set_title(target)\n",
    "    print(counter1, counter2)\n",
    "    \n",
    "    counter2 += 1\n",
    "    \n",
    "    if counter2 == 2:\n",
    "        counter1 += 1\n",
    "        counter2 = 0\n",
    "        \n",
    "# plt.show()\n",
    "fig.savefig('Interval Analysis.jpg')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e039805",
   "metadata": {},
   "source": [
    "# RDD design\n",
    "\n",
    "For the cases where we are able to construct a window greater than 2 days, we use the RDD design to get more information not only about the difference in means, but also about the differences in the regressions’ slope coefficients with an estimate of statistical significance. We cannot take the amount of days equal to 2 or less, because we get into the case of matrix singularity in terms of days, while the hour time interval is unreliable.\n",
    "\n",
    "Below we are calculating slope coefficients for each emotion:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "656e45ef",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.015\n",
      "Model:                            OLS   Adj. R-squared:                  0.015\n",
      "Method:                 Least Squares   F-statistic:                     451.0\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          6.80e-291\n",
      "Time:                        18:28:41   Log-Likelihood:                -20173.\n",
      "No. Observations:               89696   AIC:                         4.035e+04\n",
      "Df Residuals:                   89692   BIC:                         4.039e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1055      0.005     20.177      0.000       0.095       0.116\n",
      "days_from_event                 0.0124      0.002      6.888      0.000       0.009       0.016\n",
      "after_event                     0.1252      0.006     19.584      0.000       0.113       0.138\n",
      "after_event:days_from_event    -0.0412      0.002    -19.920      0.000      -0.045      -0.037\n",
      "==============================================================================\n",
      "Omnibus:                     2183.524   Durbin-Watson:                   1.841\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             4526.279\n",
      "Skew:                          -0.141   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.064   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.005\n",
      "Model:                            OLS   Adj. R-squared:                  0.005\n",
      "Method:                 Least Squares   F-statistic:                     139.0\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):           7.09e-90\n",
      "Time:                        18:28:41   Log-Likelihood:             1.1383e+05\n",
      "No. Observations:               89696   AIC:                        -2.277e+05\n",
      "Df Residuals:                   89692   BIC:                        -2.276e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0039      0.001      3.324      0.001       0.002       0.006\n",
      "days_from_event                -0.0076      0.000    -18.810      0.000      -0.008      -0.007\n",
      "after_event                     0.0213      0.001     14.821      0.000       0.018       0.024\n",
      "after_event:days_from_event     0.0063      0.000     13.500      0.000       0.005       0.007\n",
      "==============================================================================\n",
      "Omnibus:                    80486.795   Durbin-Watson:                   1.894\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2432315.629\n",
      "Skew:                           4.434   Prob(JB):                         0.00\n",
      "Kurtosis:                      26.920   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.014\n",
      "Model:                            OLS   Adj. R-squared:                  0.014\n",
      "Method:                 Least Squares   F-statistic:                     424.2\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          1.07e-273\n",
      "Time:                        18:28:41   Log-Likelihood:                 86598.\n",
      "No. Observations:               89696   AIC:                        -1.732e+05\n",
      "Df Residuals:                   89692   BIC:                        -1.732e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0498      0.002     31.307      0.000       0.047       0.053\n",
      "days_from_event                -0.0008      0.001     -1.543      0.123      -0.002       0.000\n",
      "after_event                     0.0475      0.002     24.417      0.000       0.044       0.051\n",
      "after_event:days_from_event    -0.0095      0.001    -15.091      0.000      -0.011      -0.008\n",
      "==============================================================================\n",
      "Omnibus:                    31652.921   Durbin-Watson:                   1.745\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           108188.295\n",
      "Skew:                           1.806   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.987   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.002\n",
      "Model:                            OLS   Adj. R-squared:                  0.002\n",
      "Method:                 Least Squares   F-statistic:                     68.97\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):           1.52e-44\n",
      "Time:                        18:28:41   Log-Likelihood:                 91703.\n",
      "No. Observations:               89696   AIC:                        -1.834e+05\n",
      "Df Residuals:                   89692   BIC:                        -1.834e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.0050      0.002     -3.342      0.001      -0.008      -0.002\n",
      "days_from_event                -0.0073      0.001    -14.031      0.000      -0.008      -0.006\n",
      "after_event                     0.0190      0.002     10.357      0.000       0.015       0.023\n",
      "after_event:days_from_event     0.0067      0.001     11.249      0.000       0.006       0.008\n",
      "==============================================================================\n",
      "Omnibus:                   123236.064   Durbin-Watson:                   1.924\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):         22648283.278\n",
      "Skew:                           8.260   Prob(JB):                         0.00\n",
      "Kurtosis:                      79.073   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.101\n",
      "Model:                            OLS   Adj. R-squared:                  0.101\n",
      "Method:                 Least Squares   F-statistic:                     3375.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:28:42   Log-Likelihood:                -2768.5\n",
      "No. Observations:               89696   AIC:                             5545.\n",
      "Df Residuals:                   89692   BIC:                             5583.\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2744      0.004     63.734      0.000       0.266       0.283\n",
      "days_from_event                 0.0260      0.001     17.529      0.000       0.023       0.029\n",
      "after_event                    -0.2444      0.005    -46.421      0.000      -0.255      -0.234\n",
      "after_event:days_from_event     0.0498      0.002     29.248      0.000       0.047       0.053\n",
      "==============================================================================\n",
      "Omnibus:                     5364.576   Durbin-Watson:                   1.804\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             6864.356\n",
      "Skew:                           0.576   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.715   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.017\n",
      "Model:                            OLS   Adj. R-squared:                  0.017\n",
      "Method:                 Least Squares   F-statistic:                     517.2\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:28:42   Log-Likelihood:                -22801.\n",
      "No. Observations:               89696   AIC:                         4.561e+04\n",
      "Df Residuals:                   89692   BIC:                         4.565e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1138      0.005     21.134      0.000       0.103       0.124\n",
      "days_from_event                 0.0026      0.002      1.407      0.160      -0.001       0.006\n",
      "after_event                    -0.0514      0.007     -7.810      0.000      -0.064      -0.038\n",
      "after_event:days_from_event     0.0294      0.002     13.815      0.000       0.025       0.034\n",
      "==============================================================================\n",
      "Omnibus:                    29496.065   Durbin-Watson:                   1.906\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            72617.003\n",
      "Skew:                           1.887   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.278   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.009\n",
      "Model:                            OLS   Adj. R-squared:                  0.009\n",
      "Method:                 Least Squares   F-statistic:                     261.3\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          7.86e-169\n",
      "Time:                        18:28:42   Log-Likelihood:                -20745.\n",
      "No. Observations:               89696   AIC:                         4.150e+04\n",
      "Df Residuals:                   89692   BIC:                         4.153e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2841      0.005     54.002      0.000       0.274       0.294\n",
      "days_from_event                 0.0189      0.002     10.419      0.000       0.015       0.022\n",
      "after_event                     0.0091      0.006      1.419      0.156      -0.003       0.022\n",
      "after_event:days_from_event    -0.0434      0.002    -20.832      0.000      -0.047      -0.039\n",
      "==============================================================================\n",
      "Omnibus:                    13857.524   Durbin-Watson:                   1.780\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            21300.753\n",
      "Skew:                           1.180   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.360   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.091\n",
      "Model:                            OLS   Adj. R-squared:                  0.091\n",
      "Method:                 Least Squares   F-statistic:                     2993.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:28:42   Log-Likelihood:                 5993.2\n",
      "No. Observations:               89696   AIC:                        -1.198e+04\n",
      "Df Residuals:                   89692   BIC:                        -1.194e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0127      0.004      3.244      0.001       0.005       0.020\n",
      "days_from_event                -0.0153      0.001    -11.325      0.000      -0.018      -0.013\n",
      "after_event                    -0.0250      0.005     -5.227      0.000      -0.034      -0.016\n",
      "after_event:days_from_event     0.0688      0.002     44.534      0.000       0.066       0.072\n",
      "==============================================================================\n",
      "Omnibus:                    23093.834   Durbin-Watson:                   1.793\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            55425.854\n",
      "Skew:                           1.440   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.557   Cond. No.                         38.9\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.031\n",
      "Model:                            OLS   Adj. R-squared:                  0.031\n",
      "Method:                 Least Squares   F-statistic:                     4123.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:30:16   Log-Likelihood:            -1.8909e+05\n",
      "No. Observations:              380462   AIC:                         3.782e+05\n",
      "Df Residuals:                  380458   BIC:                         3.782e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.1311      0.003    -44.592      0.000      -0.137      -0.125\n",
      "days_from_event                 0.0335      0.002     19.238      0.000       0.030       0.037\n",
      "after_event                     0.0834      0.003     25.974      0.000       0.077       0.090\n",
      "after_event:days_from_event    -0.0204      0.003     -7.637      0.000      -0.026      -0.015\n",
      "==============================================================================\n",
      "Omnibus:                     8912.072   Durbin-Watson:                   1.951\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             4708.292\n",
      "Skew:                           0.058   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.468   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.007\n",
      "Model:                            OLS   Adj. R-squared:                  0.007\n",
      "Method:                 Least Squares   F-statistic:                     908.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:30:17   Log-Likelihood:             3.1877e+05\n",
      "No. Observations:              380462   AIC:                        -6.375e+05\n",
      "Df Residuals:                  380458   BIC:                        -6.375e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1188      0.001    153.568      0.000       0.117       0.120\n",
      "days_from_event                 0.0026      0.000      5.733      0.000       0.002       0.004\n",
      "after_event                    -0.0253      0.001    -29.964      0.000      -0.027      -0.024\n",
      "after_event:days_from_event     0.0134      0.001     19.057      0.000       0.012       0.015\n",
      "==============================================================================\n",
      "Omnibus:                    52025.134   Durbin-Watson:                   1.961\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            84530.356\n",
      "Skew:                           0.945   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.327   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.025\n",
      "Model:                            OLS   Adj. R-squared:                  0.025\n",
      "Method:                 Least Squares   F-statistic:                     3284.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:30:17   Log-Likelihood:             4.1361e+05\n",
      "No. Observations:              380462   AIC:                        -8.272e+05\n",
      "Df Residuals:                  380458   BIC:                        -8.272e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0736      0.001    122.013      0.000       0.072       0.075\n",
      "days_from_event                 0.0140      0.000     39.171      0.000       0.013       0.015\n",
      "after_event                     0.0001      0.001      0.161      0.872      -0.001       0.001\n",
      "after_event:days_from_event    -0.0101      0.001    -18.379      0.000      -0.011      -0.009\n",
      "==============================================================================\n",
      "Omnibus:                    88292.226   Durbin-Watson:                   1.976\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           191415.969\n",
      "Skew:                           1.349   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.190   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.004\n",
      "Model:                            OLS   Adj. R-squared:                  0.004\n",
      "Method:                 Least Squares   F-statistic:                     452.9\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          9.24e-294\n",
      "Time:                        18:30:17   Log-Likelihood:             1.6582e+05\n",
      "No. Observations:              380462   AIC:                        -3.316e+05\n",
      "Df Residuals:                  380458   BIC:                        -3.316e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0719      0.001     62.190      0.000       0.070       0.074\n",
      "days_from_event                 0.0127      0.001     18.541      0.000       0.011       0.014\n",
      "after_event                    -0.0301      0.001    -23.791      0.000      -0.033      -0.028\n",
      "after_event:days_from_event    -0.0247      0.001    -23.529      0.000      -0.027      -0.023\n",
      "==============================================================================\n",
      "Omnibus:                   330247.231   Durbin-Watson:                   1.982\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          7892727.075\n",
      "Skew:                           4.316   Prob(JB):                         0.00\n",
      "Kurtosis:                      23.576   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.016\n",
      "Model:                            OLS   Adj. R-squared:                  0.016\n",
      "Method:                 Least Squares   F-statistic:                     2118.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:30:17   Log-Likelihood:                -74601.\n",
      "No. Observations:              380462   AIC:                         1.492e+05\n",
      "Df Residuals:                  380458   BIC:                         1.493e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1341      0.002     61.653      0.000       0.130       0.138\n",
      "days_from_event                -0.0305      0.001    -23.660      0.000      -0.033      -0.028\n",
      "after_event                     0.0883      0.002     37.165      0.000       0.084       0.093\n",
      "after_event:days_from_event     0.0895      0.002     45.237      0.000       0.086       0.093\n",
      "==============================================================================\n",
      "Omnibus:                    60855.189   Durbin-Watson:                   1.953\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            94432.393\n",
      "Skew:                           1.193   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.511   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.002\n",
      "Model:                            OLS   Adj. R-squared:                  0.002\n",
      "Method:                 Least Squares   F-statistic:                     192.7\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          7.31e-125\n",
      "Time:                        18:30:17   Log-Likelihood:                -90634.\n",
      "No. Observations:              380462   AIC:                         1.813e+05\n",
      "Df Residuals:                  380458   BIC:                         1.813e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1605      0.002     70.739      0.000       0.156       0.165\n",
      "days_from_event                -0.0311      0.001    -23.160      0.000      -0.034      -0.028\n",
      "after_event                     0.0437      0.002     17.639      0.000       0.039       0.049\n",
      "after_event:days_from_event     0.0401      0.002     19.460      0.000       0.036       0.044\n",
      "==============================================================================\n",
      "Omnibus:                    84779.673   Durbin-Watson:                   1.967\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           154289.622\n",
      "Skew:                           1.459   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.106   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.002\n",
      "Model:                            OLS   Adj. R-squared:                  0.002\n",
      "Method:                 Least Squares   F-statistic:                     202.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          2.50e-131\n",
      "Time:                        18:30:18   Log-Likelihood:            -1.3560e+05\n",
      "No. Observations:              380462   AIC:                         2.712e+05\n",
      "Df Residuals:                  380458   BIC:                         2.713e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2775      0.003    108.651      0.000       0.272       0.282\n",
      "days_from_event                -0.0132      0.002     -8.708      0.000      -0.016      -0.010\n",
      "after_event                    -0.0055      0.003     -1.960      0.050      -0.011   -1.15e-06\n",
      "after_event:days_from_event     0.0146      0.002      6.300      0.000       0.010       0.019\n",
      "==============================================================================\n",
      "Omnibus:                    43990.550   Durbin-Watson:                   1.957\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            58167.174\n",
      "Skew:                           0.942   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.649   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.008\n",
      "Model:                            OLS   Adj. R-squared:                  0.008\n",
      "Method:                 Least Squares   F-statistic:                     1020.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:30:18   Log-Likelihood:             1.6900e+05\n",
      "No. Observations:              380462   AIC:                        -3.380e+05\n",
      "Df Residuals:                  380458   BIC:                        -3.379e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0391      0.001     34.063      0.000       0.037       0.041\n",
      "days_from_event                -0.0134      0.001    -19.671      0.000      -0.015      -0.012\n",
      "after_event                     0.0027      0.001      2.186      0.029       0.000       0.005\n",
      "after_event:days_from_event    -0.0009      0.001     -0.910      0.363      -0.003       0.001\n",
      "==============================================================================\n",
      "Omnibus:                   288732.034   Durbin-Watson:                   1.995\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          4536375.199\n",
      "Skew:                           3.684   Prob(JB):                         0.00\n",
      "Kurtosis:                      18.228   Cond. No.                         11.3\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.007\n",
      "Model:                            OLS   Adj. R-squared:                  0.007\n",
      "Method:                 Least Squares   F-statistic:                     299.9\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          4.74e-194\n",
      "Time:                        18:31:03   Log-Likelihood:                -72221.\n",
      "No. Observations:              134489   AIC:                         1.444e+05\n",
      "Df Residuals:                  134485   BIC:                         1.445e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.0229      0.003     -7.676      0.000      -0.029      -0.017\n",
      "days_from_event                -0.0002      0.000     -1.900      0.057      -0.001    7.76e-06\n",
      "after_event                     0.0863      0.005     18.850      0.000       0.077       0.095\n",
      "after_event:days_from_event    -0.0004      0.000     -2.274      0.023      -0.001   -6.17e-05\n",
      "==============================================================================\n",
      "Omnibus:                      516.717   Durbin-Watson:                   1.629\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              394.671\n",
      "Skew:                           0.031   Prob(JB):                     1.99e-86\n",
      "Kurtosis:                       2.742   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     607.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:31:03   Log-Likelihood:             1.1451e+05\n",
      "No. Observations:              134489   AIC:                        -2.290e+05\n",
      "Df Residuals:                  134485   BIC:                        -2.290e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0871      0.001    116.947      0.000       0.086       0.089\n",
      "days_from_event                 0.0007   3.24e-05     20.169      0.000       0.001       0.001\n",
      "after_event                    -0.0310      0.001    -27.136      0.000      -0.033      -0.029\n",
      "after_event:days_from_event    -0.0008    4.9e-05    -16.457      0.000      -0.001      -0.001\n",
      "==============================================================================\n",
      "Omnibus:                    43390.096   Durbin-Watson:                   1.669\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           122184.606\n",
      "Skew:                           1.735   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.124   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.003\n",
      "Model:                            OLS   Adj. R-squared:                  0.003\n",
      "Method:                 Least Squares   F-statistic:                     139.9\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):           1.68e-90\n",
      "Time:                        18:31:03   Log-Likelihood:             1.1705e+05\n",
      "No. Observations:              134489   AIC:                        -2.341e+05\n",
      "Df Residuals:                  134485   BIC:                        -2.340e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0707      0.001     96.771      0.000       0.069       0.072\n",
      "days_from_event                 0.0003   3.18e-05      8.189      0.000       0.000       0.000\n",
      "after_event                     0.0088      0.001      7.830      0.000       0.007       0.011\n",
      "after_event:days_from_event    -0.0009   4.81e-05    -19.499      0.000      -0.001      -0.001\n",
      "==============================================================================\n",
      "Omnibus:                    38477.684   Durbin-Watson:                   1.683\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            93118.034\n",
      "Skew:                           1.608   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.504   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.006\n",
      "Model:                            OLS   Adj. R-squared:                  0.006\n",
      "Method:                 Least Squares   F-statistic:                     286.8\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          1.41e-185\n",
      "Time:                        18:31:03   Log-Likelihood:                 43686.\n",
      "No. Observations:              134489   AIC:                        -8.736e+04\n",
      "Df Residuals:                  134485   BIC:                        -8.732e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0298      0.001     23.646      0.000       0.027       0.032\n",
      "days_from_event                -0.0005   5.49e-05     -9.901      0.000      -0.001      -0.000\n",
      "after_event                    -0.0051      0.002     -2.642      0.008      -0.009      -0.001\n",
      "after_event:days_from_event     0.0020    8.3e-05     24.089      0.000       0.002       0.002\n",
      "==============================================================================\n",
      "Omnibus:                   113927.203   Durbin-Watson:                   1.648\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2287709.615\n",
      "Skew:                           4.236   Prob(JB):                         0.00\n",
      "Kurtosis:                      21.343   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.005\n",
      "Model:                            OLS   Adj. R-squared:                  0.005\n",
      "Method:                 Least Squares   F-statistic:                     209.2\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          2.29e-135\n",
      "Time:                        18:31:03   Log-Likelihood:                -26314.\n",
      "No. Observations:              134489   AIC:                         5.264e+04\n",
      "Df Residuals:                  134485   BIC:                         5.267e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1331      0.002     62.693      0.000       0.129       0.137\n",
      "days_from_event                -0.0014   9.24e-05    -15.135      0.000      -0.002      -0.001\n",
      "after_event                     0.0416      0.003     12.780      0.000       0.035       0.048\n",
      "after_event:days_from_event    -0.0002      0.000     -1.759      0.079      -0.001     2.8e-05\n",
      "==============================================================================\n",
      "Omnibus:                    44079.202   Durbin-Watson:                   1.739\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           109107.696\n",
      "Skew:                           1.872   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.334   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.024\n",
      "Model:                            OLS   Adj. R-squared:                  0.024\n",
      "Method:                 Least Squares   F-statistic:                     1107.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:31:03   Log-Likelihood:                -23897.\n",
      "No. Observations:              134489   AIC:                         4.780e+04\n",
      "Df Residuals:                  134485   BIC:                         4.784e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1070      0.002     51.330      0.000       0.103       0.111\n",
      "days_from_event                -0.0005   9.07e-05     -5.404      0.000      -0.001      -0.000\n",
      "after_event                    -0.0395      0.003    -12.353      0.000      -0.046      -0.033\n",
      "after_event:days_from_event     0.0054      0.000     39.149      0.000       0.005       0.006\n",
      "==============================================================================\n",
      "Omnibus:                    47460.487   Durbin-Watson:                   1.647\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           128764.905\n",
      "Skew:                           1.961   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.757   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.009\n",
      "Model:                            OLS   Adj. R-squared:                  0.009\n",
      "Method:                 Least Squares   F-statistic:                     386.4\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          6.60e-250\n",
      "Time:                        18:31:03   Log-Likelihood:                -50066.\n",
      "No. Observations:              134489   AIC:                         1.001e+05\n",
      "Df Residuals:                  134485   BIC:                         1.002e+05\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2200      0.003     86.852      0.000       0.215       0.225\n",
      "days_from_event                 0.0008      0.000      7.587      0.000       0.001       0.001\n",
      "after_event                     0.0490      0.004     12.631      0.000       0.041       0.057\n",
      "after_event:days_from_event    -0.0009      0.000     -5.509      0.000      -0.001      -0.001\n",
      "==============================================================================\n",
      "Omnibus:                    20990.196   Durbin-Watson:                   1.596\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            32838.888\n",
      "Skew:                           1.210   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.037   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.010\n",
      "Model:                            OLS   Adj. R-squared:                  0.010\n",
      "Method:                 Least Squares   F-statistic:                     435.4\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          1.44e-281\n",
      "Time:                        18:31:03   Log-Likelihood:                 29.447\n",
      "No. Observations:              134489   AIC:                            -50.89\n",
      "Df Residuals:                  134485   BIC:                            -11.66\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1346      0.002     77.116      0.000       0.131       0.138\n",
      "days_from_event                 0.0023    7.6e-05     30.223      0.000       0.002       0.002\n",
      "after_event                    -0.0323      0.003    -12.074      0.000      -0.038      -0.027\n",
      "after_event:days_from_event    -0.0017      0.000    -15.047      0.000      -0.002      -0.002\n",
      "==============================================================================\n",
      "Omnibus:                    71435.702   Durbin-Watson:                   1.637\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           392627.930\n",
      "Skew:                           2.654   Prob(JB):                         0.00\n",
      "Kurtosis:                       9.473   Cond. No.                         117.\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.017\n",
      "Model:                            OLS   Adj. R-squared:                  0.017\n",
      "Method:                 Least Squares   F-statistic:                     298.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          3.50e-192\n",
      "Time:                        18:31:42   Log-Likelihood:                -24582.\n",
      "No. Observations:               50341   AIC:                         4.917e+04\n",
      "Df Residuals:                   50337   BIC:                         4.921e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1428      0.006     24.508      0.000       0.131       0.154\n",
      "days_from_event                 0.0124      0.001     12.542      0.000       0.011       0.014\n",
      "after_event                    -0.1641      0.007    -22.027      0.000      -0.179      -0.150\n",
      "after_event:days_from_event    -0.0118      0.001     -8.980      0.000      -0.014      -0.009\n",
      "==============================================================================\n",
      "Omnibus:                       80.925   Durbin-Watson:                   1.687\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):               79.017\n",
      "Skew:                           0.083   Prob(JB):                     6.95e-18\n",
      "Kurtosis:                       2.901   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     215.3\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          8.99e-139\n",
      "Time:                        18:31:42   Log-Likelihood:                 41410.\n",
      "No. Observations:               50341   AIC:                        -8.281e+04\n",
      "Df Residuals:                   50337   BIC:                        -8.278e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0654      0.002     41.665      0.000       0.062       0.069\n",
      "days_from_event                 0.0018      0.000      6.592      0.000       0.001       0.002\n",
      "after_event                     0.0258      0.002     12.831      0.000       0.022       0.030\n",
      "after_event:days_from_event    -0.0065      0.000    -18.315      0.000      -0.007      -0.006\n",
      "==============================================================================\n",
      "Omnibus:                    17726.663   Durbin-Watson:                   1.770\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            55266.171\n",
      "Skew:                           1.852   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.554   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.029\n",
      "Model:                            OLS   Adj. R-squared:                  0.029\n",
      "Method:                 Least Squares   F-statistic:                     507.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:31:42   Log-Likelihood:                 42806.\n",
      "No. Observations:               50341   AIC:                        -8.560e+04\n",
      "Df Residuals:                   50337   BIC:                        -8.557e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1156      0.002     75.660      0.000       0.113       0.119\n",
      "days_from_event                 0.0068      0.000     26.250      0.000       0.006       0.007\n",
      "after_event                    -0.0442      0.002    -22.646      0.000      -0.048      -0.040\n",
      "after_event:days_from_event    -0.0100      0.000    -29.105      0.000      -0.011      -0.009\n",
      "==============================================================================\n",
      "Omnibus:                    16603.545   Durbin-Watson:                   1.798\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            47950.028\n",
      "Skew:                           1.762   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.231   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.006\n",
      "Model:                            OLS   Adj. R-squared:                  0.006\n",
      "Method:                 Least Squares   F-statistic:                     95.26\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):           1.78e-61\n",
      "Time:                        18:31:42   Log-Likelihood:                 17715.\n",
      "No. Observations:               50341   AIC:                        -3.542e+04\n",
      "Df Residuals:                   50337   BIC:                        -3.539e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0057      0.003      2.279      0.023       0.001       0.011\n",
      "days_from_event                -0.0070      0.000    -16.425      0.000      -0.008      -0.006\n",
      "after_event                     0.0299      0.003      9.292      0.000       0.024       0.036\n",
      "after_event:days_from_event     0.0072      0.001     12.741      0.000       0.006       0.008\n",
      "==============================================================================\n",
      "Omnibus:                    47126.908   Durbin-Watson:                   1.748\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          1301890.390\n",
      "Skew:                           4.802   Prob(JB):                         0.00\n",
      "Kurtosis:                      25.988   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.010\n",
      "Model:                            OLS   Adj. R-squared:                  0.010\n",
      "Method:                 Least Squares   F-statistic:                     174.5\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          1.55e-112\n",
      "Time:                        18:31:42   Log-Likelihood:                -6781.1\n",
      "No. Observations:               50341   AIC:                         1.357e+04\n",
      "Df Residuals:                   50337   BIC:                         1.361e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2117      0.004     51.752      0.000       0.204       0.220\n",
      "days_from_event                 0.0083      0.001     11.930      0.000       0.007       0.010\n",
      "after_event                    -0.0857      0.005    -16.388      0.000      -0.096      -0.075\n",
      "after_event:days_from_event    -0.0096      0.001    -10.393      0.000      -0.011      -0.008\n",
      "==============================================================================\n",
      "Omnibus:                    18993.112   Durbin-Watson:                   1.790\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            57306.145\n",
      "Skew:                           2.035   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.278   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.134\n",
      "Model:                            OLS   Adj. R-squared:                  0.134\n",
      "Method:                 Least Squares   F-statistic:                     2602.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:31:42   Log-Likelihood:                -11971.\n",
      "No. Observations:               50341   AIC:                         2.395e+04\n",
      "Df Residuals:                   50337   BIC:                         2.398e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0007      0.005      0.149      0.881      -0.008       0.010\n",
      "days_from_event                -0.0504      0.001    -65.275      0.000      -0.052      -0.049\n",
      "after_event                     0.0932      0.006     16.074      0.000       0.082       0.105\n",
      "after_event:days_from_event     0.0520      0.001     50.775      0.000       0.050       0.054\n",
      "==============================================================================\n",
      "Omnibus:                    12492.804   Durbin-Watson:                   1.682\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            25709.407\n",
      "Skew:                           1.488   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.844   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.065\n",
      "Model:                            OLS   Adj. R-squared:                  0.065\n",
      "Method:                 Least Squares   F-statistic:                     1165.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:31:43   Log-Likelihood:                -22152.\n",
      "No. Observations:               50341   AIC:                         4.431e+04\n",
      "Df Residuals:                   50337   BIC:                         4.435e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.4179      0.006     75.291      0.000       0.407       0.429\n",
      "days_from_event                 0.0343      0.001     36.271      0.000       0.032       0.036\n",
      "after_event                     0.0191      0.007      2.692      0.007       0.005       0.033\n",
      "after_event:days_from_event    -0.0724      0.001    -57.698      0.000      -0.075      -0.070\n",
      "==============================================================================\n",
      "Omnibus:                     6674.744   Durbin-Watson:                   1.715\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             9551.744\n",
      "Skew:                           1.059   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.740   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.015\n",
      "Model:                            OLS   Adj. R-squared:                  0.015\n",
      "Method:                 Least Squares   F-statistic:                     257.7\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          5.47e-166\n",
      "Time:                        18:31:43   Log-Likelihood:                 7605.3\n",
      "No. Observations:               50341   AIC:                        -1.520e+04\n",
      "Df Residuals:                   50337   BIC:                        -1.517e+04\n",
      "Df Model:                           3                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1443      0.003     46.947      0.000       0.138       0.150\n",
      "days_from_event                 0.0071      0.001     13.527      0.000       0.006       0.008\n",
      "after_event                    -0.0856      0.004    -21.780      0.000      -0.093      -0.078\n",
      "after_event:days_from_event    -0.0064      0.001     -9.168      0.000      -0.008      -0.005\n",
      "==============================================================================\n",
      "Omnibus:                    30850.675   Durbin-Watson:                   1.796\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           251577.301\n",
      "Skew:                           2.981   Prob(JB):                         0.00\n",
      "Kurtosis:                      12.187   Cond. No.                         33.1\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "indices = [1, 6, 13, 14]\n",
    "\n",
    "for cutoff_index in indices:\n",
    "    fin_df = df_en.copy()\n",
    "\n",
    "    date0 = jump_dates[cutoff_index - 1]\n",
    "    date1 = jump_dates[cutoff_index]\n",
    "    date2 = jump_dates[cutoff_index + 1]\n",
    "    \n",
    "    days1 = int(str(np.datetime64(date1) - np.datetime64(date0)).replace(' days', ''))\n",
    "    days2 = int(str(np.datetime64(date2) - np.datetime64(date1)).replace(' days', ''))\n",
    "\n",
    "    window = min(days1, days2)\n",
    "\n",
    "    fin_df = fin_df[(fin_df['created_at_date']>np.datetime64(date0))&\n",
    "                    (fin_df['created_at_date']<np.datetime64(date2))]\n",
    "\n",
    "    fin_df['after_event'] = fin_df.apply(lambda x: event_occured(x['created_at_date'], str(date1)), axis = 1)\n",
    "    fin_df['days_from_event'] = fin_df.apply(lambda x: days_from_event(x['created_at_date'], str(date1)), axis = 1)\n",
    "    \n",
    "    fin_df['after_event'] = fin_df.apply(lambda x: event_occured(x['created_at_date'], str(date1)), axis = 1)\n",
    "    fin_df['days_from_event'] = fin_df.apply(lambda x: days_from_event(x['created_at_date'], str(date1)), axis = 1)\n",
    "    \n",
    "    reg_df = fin_df[(fin_df['days_from_event'] >= -window)\n",
    "                &(fin_df['days_from_event'] <= window)]\n",
    "\n",
    "    for target in ['compound', 'neg', 'pos', 'Angry', 'Sad', 'Surprise', 'Fear', 'Happy']:\n",
    "\n",
    "        result = smf.ols(formula = f\"{target} ~ days_from_event + after_event + after_event * days_from_event\", \n",
    "                         data = reg_df).fit()\n",
    "        print(result.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "60d9c4a0",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.196\n",
      "Model:                            OLS   Adj. R-squared:                  0.196\n",
      "Method:                 Least Squares   F-statistic:                     3128.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:26   Log-Likelihood:                -11048.\n",
      "No. Observations:               89696   AIC:                         2.211e+04\n",
      "Df Residuals:                   89688   BIC:                         2.219e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0827      0.005     17.498      0.000       0.073       0.092\n",
      "days_from_event                 0.0081      0.002      4.998      0.000       0.005       0.011\n",
      "after_event                     0.1929      0.006     33.296      0.000       0.182       0.204\n",
      "after_event:days_from_event    -0.0856      0.002    -45.167      0.000      -0.089      -0.082\n",
      "retweet_count                1.238e-05    8.7e-08    142.206      0.000    1.22e-05    1.25e-05\n",
      "reply_count                  7.607e-06      0.000      0.037      0.970      -0.000       0.000\n",
      "like_count                  -4.166e-06   3.37e-06     -1.238      0.216   -1.08e-05    2.43e-06\n",
      "quote_count                  4.753e-05   6.98e-05      0.681      0.496   -8.92e-05       0.000\n",
      "==============================================================================\n",
      "Omnibus:                     7538.029   Durbin-Watson:                   1.776\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            24301.368\n",
      "Skew:                           0.424   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.405   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.053\n",
      "Model:                            OLS   Adj. R-squared:                  0.053\n",
      "Method:                 Least Squares   F-statistic:                     716.8\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:26   Log-Likelihood:             1.1607e+05\n",
      "No. Observations:               89696   AIC:                        -2.321e+05\n",
      "Df Residuals:                   89688   BIC:                        -2.320e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0065      0.001      5.700      0.000       0.004       0.009\n",
      "days_from_event                -0.0071      0.000    -18.019      0.000      -0.008      -0.006\n",
      "after_event                     0.0135      0.001      9.592      0.000       0.011       0.016\n",
      "after_event:days_from_event     0.0114      0.000     24.740      0.000       0.010       0.012\n",
      "retweet_count               -1.424e-06   2.11e-08    -67.474      0.000   -1.47e-06   -1.38e-06\n",
      "reply_count                     0.0002   4.95e-05      4.021      0.000       0.000       0.000\n",
      "like_count                  -2.127e-06   8.16e-07     -2.607      0.009   -3.73e-06   -5.28e-07\n",
      "quote_count                 -5.905e-05   1.69e-05     -3.491      0.000   -9.22e-05   -2.59e-05\n",
      "==============================================================================\n",
      "Omnibus:                    78378.706   Durbin-Watson:                   1.897\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2292470.056\n",
      "Skew:                           4.261   Prob(JB):                         0.00\n",
      "Kurtosis:                      26.255   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.071\n",
      "Model:                            OLS   Adj. R-squared:                  0.071\n",
      "Method:                 Least Squares   F-statistic:                     976.3\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:26   Log-Likelihood:                 89259.\n",
      "No. Observations:               89696   AIC:                        -1.785e+05\n",
      "Df Residuals:                   89688   BIC:                        -1.784e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0459      0.002     29.719      0.000       0.043       0.049\n",
      "days_from_event                -0.0016      0.001     -2.947      0.003      -0.003      -0.001\n",
      "after_event                     0.0590      0.002     31.146      0.000       0.055       0.063\n",
      "after_event:days_from_event    -0.0171      0.001    -27.534      0.000      -0.018      -0.016\n",
      "retweet_count                2.106e-06   2.85e-08     74.022      0.000    2.05e-06    2.16e-06\n",
      "reply_count                     0.0002   6.68e-05      2.659      0.008    4.67e-05       0.000\n",
      "like_count                  -2.583e-06    1.1e-06     -2.349      0.019   -4.74e-06   -4.28e-07\n",
      "quote_count                 -4.577e-05   2.28e-05     -2.007      0.045   -9.05e-05   -1.07e-06\n",
      "==============================================================================\n",
      "Omnibus:                    41029.010   Durbin-Watson:                   1.701\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           206032.624\n",
      "Skew:                           2.233   Prob(JB):                         0.00\n",
      "Kurtosis:                       8.931   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     173.0\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          1.89e-255\n",
      "Time:                        18:33:26   Log-Likelihood:                 92201.\n",
      "No. Observations:               89696   AIC:                        -1.844e+05\n",
      "Df Residuals:                   89688   BIC:                        -1.843e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.0035      0.001     -2.325      0.020      -0.006      -0.001\n",
      "days_from_event                -0.0070      0.001    -13.522      0.000      -0.008      -0.006\n",
      "after_event                     0.0144      0.002      7.864      0.000       0.011       0.018\n",
      "after_event:days_from_event     0.0097      0.001     16.139      0.000       0.008       0.011\n",
      "retweet_count               -8.387e-07   2.75e-08    -30.462      0.000   -8.93e-07   -7.85e-07\n",
      "reply_count                     0.0004   6.46e-05      5.478      0.000       0.000       0.000\n",
      "like_count                   -3.24e-06   1.06e-06     -3.044      0.002   -5.33e-06   -1.15e-06\n",
      "quote_count                 -8.114e-05   2.21e-05     -3.676      0.000      -0.000   -3.79e-05\n",
      "==============================================================================\n",
      "Omnibus:                   122622.419   Durbin-Watson:                   1.925\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):         22342152.588\n",
      "Skew:                           8.182   Prob(JB):                         0.00\n",
      "Kurtosis:                      78.567   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.251\n",
      "Model:                            OLS   Adj. R-squared:                  0.251\n",
      "Method:                 Least Squares   F-statistic:                     4287.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:26   Log-Likelihood:                 5379.9\n",
      "No. Observations:               89696   AIC:                        -1.074e+04\n",
      "Df Residuals:                   89688   BIC:                        -1.067e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2565      0.004     65.214      0.000       0.249       0.264\n",
      "days_from_event                 0.0227      0.001     16.731      0.000       0.020       0.025\n",
      "after_event                    -0.1914      0.005    -39.682      0.000      -0.201      -0.182\n",
      "after_event:days_from_event     0.0151      0.002      9.575      0.000       0.012       0.018\n",
      "retweet_count                9.688e-06   7.25e-08    133.662      0.000    9.55e-06    9.83e-06\n",
      "reply_count                     0.0003      0.000      1.960      0.050   -4.82e-08       0.001\n",
      "like_count                  -4.439e-06    2.8e-06     -1.584      0.113   -9.93e-06    1.05e-06\n",
      "quote_count                    -0.0001   5.81e-05     -2.105      0.035      -0.000   -8.44e-06\n",
      "==============================================================================\n",
      "Omnibus:                    21296.103   Durbin-Watson:                   1.798\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            51302.183\n",
      "Skew:                           1.324   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.592   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.113\n",
      "Model:                            OLS   Adj. R-squared:                  0.113\n",
      "Method:                 Least Squares   F-statistic:                     1636.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:27   Log-Likelihood:                -18181.\n",
      "No. Observations:               89696   AIC:                         3.638e+04\n",
      "Df Residuals:                   89688   BIC:                         3.645e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1309      0.005     25.589      0.000       0.121       0.141\n",
      "days_from_event                 0.0058      0.002      3.293      0.001       0.002       0.009\n",
      "after_event                    -0.1022      0.006    -16.296      0.000      -0.115      -0.090\n",
      "after_event:days_from_event     0.0628      0.002     30.599      0.000       0.059       0.067\n",
      "retweet_count               -9.298e-06   9.43e-08    -98.645      0.000   -9.48e-06   -9.11e-06\n",
      "reply_count                    -0.0005      0.000     -2.244      0.025      -0.001   -6.28e-05\n",
      "like_count                   5.095e-06   3.64e-06      1.398      0.162   -2.05e-06    1.22e-05\n",
      "quote_count                     0.0002   7.56e-05      2.241      0.025    2.12e-05       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    25862.867   Durbin-Watson:                   1.898\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            58338.977\n",
      "Skew:                           1.683   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.070   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.070\n",
      "Model:                            OLS   Adj. R-squared:                  0.070\n",
      "Method:                 Least Squares   F-statistic:                     967.9\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:27   Log-Likelihood:                -17869.\n",
      "No. Observations:               89696   AIC:                         3.575e+04\n",
      "Df Residuals:                   89688   BIC:                         3.583e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2974      0.005     58.346      0.000       0.287       0.307\n",
      "days_from_event                 0.0214      0.002     12.178      0.000       0.018       0.025\n",
      "after_event                    -0.0304      0.006     -4.871      0.000      -0.043      -0.018\n",
      "after_event:days_from_event    -0.0174      0.002     -8.527      0.000      -0.021      -0.013\n",
      "retweet_count               -7.238e-06   9.39e-08    -77.060      0.000   -7.42e-06   -7.05e-06\n",
      "reply_count                 -3.625e-05      0.000     -0.164      0.869      -0.000       0.000\n",
      "like_count                  -1.051e-07   3.63e-06     -0.029      0.977   -7.22e-06    7.01e-06\n",
      "quote_count                  2.571e-05   7.53e-05      0.342      0.733      -0.000       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    11251.139   Durbin-Watson:                   1.777\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            16052.105\n",
      "Skew:                           1.029   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.241   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.383\n",
      "Model:                            OLS   Adj. R-squared:                  0.383\n",
      "Method:                 Least Squares   F-statistic:                     7947.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:33:27   Log-Likelihood:                 23356.\n",
      "No. Observations:               89696   AIC:                        -4.670e+04\n",
      "Df Residuals:                   89688   BIC:                        -4.662e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.0098      0.003     -3.056      0.002      -0.016      -0.004\n",
      "days_from_event                -0.0195      0.001    -17.531      0.000      -0.022      -0.017\n",
      "after_event                     0.0418      0.004     10.594      0.000       0.034       0.050\n",
      "after_event:days_from_event     0.0250      0.001     19.389      0.000       0.023       0.028\n",
      "retweet_count                1.221e-05   5.93e-08    205.902      0.000    1.21e-05    1.23e-05\n",
      "reply_count                     0.0004      0.000      3.164      0.002       0.000       0.001\n",
      "like_count                  -6.226e-06   2.29e-06     -2.715      0.007   -1.07e-05   -1.73e-06\n",
      "quote_count                    -0.0001   4.76e-05     -2.954      0.003      -0.000   -4.73e-05\n",
      "==============================================================================\n",
      "Omnibus:                    57085.217   Durbin-Watson:                   1.800\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           639146.114\n",
      "Skew:                           2.984   Prob(JB):                         0.00\n",
      "Kurtosis:                      14.636   Cond. No.                     1.15e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.15e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.035\n",
      "Model:                            OLS   Adj. R-squared:                  0.035\n",
      "Method:                 Least Squares   F-statistic:                     1964.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:02   Log-Likelihood:            -1.8842e+05\n",
      "No. Observations:              380462   AIC:                         3.769e+05\n",
      "Df Residuals:                  380454   BIC:                         3.769e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.1348      0.003    -45.908      0.000      -0.141      -0.129\n",
      "days_from_event                 0.0370      0.002     21.268      0.000       0.034       0.040\n",
      "after_event                     0.0692      0.003     21.428      0.000       0.063       0.076\n",
      "after_event:days_from_event    -0.0394      0.003    -14.491      0.000      -0.045      -0.034\n",
      "retweet_count                1.765e-06   4.84e-08     36.469      0.000    1.67e-06    1.86e-06\n",
      "reply_count                     0.0001   6.57e-05      1.637      0.102   -2.12e-05       0.000\n",
      "like_count                   9.758e-07   2.53e-06      0.385      0.700   -3.99e-06    5.94e-06\n",
      "quote_count                    -0.0002   7.54e-05     -1.995      0.046      -0.000   -2.65e-06\n",
      "==============================================================================\n",
      "Omnibus:                     7133.868   Durbin-Watson:                   1.957\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             4131.096\n",
      "Skew:                           0.078   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.514   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.017\n",
      "Model:                            OLS   Adj. R-squared:                  0.017\n",
      "Method:                 Least Squares   F-statistic:                     929.4\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:03   Log-Likelihood:             3.2064e+05\n",
      "No. Observations:              380462   AIC:                        -6.413e+05\n",
      "Df Residuals:                  380454   BIC:                        -6.412e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1172      0.001    152.114      0.000       0.116       0.119\n",
      "days_from_event                 0.0042      0.000      9.158      0.000       0.003       0.005\n",
      "after_event                    -0.0316      0.001    -37.286      0.000      -0.033      -0.030\n",
      "after_event:days_from_event     0.0050      0.001      7.059      0.000       0.004       0.006\n",
      "retweet_count                 7.78e-07   1.27e-08     61.257      0.000    7.53e-07    8.03e-07\n",
      "reply_count                 -2.769e-06   1.72e-05     -0.161      0.872   -3.66e-05     3.1e-05\n",
      "like_count                   1.808e-07   6.64e-07      0.272      0.786   -1.12e-06    1.48e-06\n",
      "quote_count                  3.412e-06   1.98e-05      0.172      0.863   -3.54e-05    4.22e-05\n",
      "==============================================================================\n",
      "Omnibus:                    61500.965   Durbin-Watson:                   1.957\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           108875.540\n",
      "Skew:                           1.047   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.575   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.036\n",
      "Model:                            OLS   Adj. R-squared:                  0.036\n",
      "Method:                 Least Squares   F-statistic:                     2036.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:03   Log-Likelihood:             4.1574e+05\n",
      "No. Observations:              380462   AIC:                        -8.315e+05\n",
      "Df Residuals:                  380454   BIC:                        -8.314e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0722      0.001    120.352      0.000       0.071       0.073\n",
      "days_from_event                 0.0153      0.000     42.963      0.000       0.015       0.016\n",
      "after_event                    -0.0051      0.001     -7.736      0.000      -0.006      -0.004\n",
      "after_event:days_from_event    -0.0170      0.001    -30.669      0.000      -0.018      -0.016\n",
      "retweet_count                6.472e-07   9.89e-09     65.433      0.000    6.28e-07    6.67e-07\n",
      "reply_count                  2.742e-05   1.34e-05      2.042      0.041     1.1e-06    5.37e-05\n",
      "like_count                    2.67e-07   5.17e-07      0.516      0.606   -7.47e-07    1.28e-06\n",
      "quote_count                 -2.582e-05   1.54e-05     -1.675      0.094    -5.6e-05     4.4e-06\n",
      "==============================================================================\n",
      "Omnibus:                    98469.545   Durbin-Watson:                   1.981\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           238257.693\n",
      "Skew:                           1.444   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.587   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.015\n",
      "Model:                            OLS   Adj. R-squared:                  0.015\n",
      "Method:                 Least Squares   F-statistic:                     803.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:04   Log-Likelihood:             1.6794e+05\n",
      "No. Observations:              380462   AIC:                        -3.359e+05\n",
      "Df Residuals:                  380454   BIC:                        -3.358e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0745      0.001     64.739      0.000       0.072       0.077\n",
      "days_from_event                 0.0102      0.001     14.987      0.000       0.009       0.012\n",
      "after_event                    -0.0201      0.001    -15.878      0.000      -0.023      -0.018\n",
      "after_event:days_from_event    -0.0114      0.001    -10.732      0.000      -0.014      -0.009\n",
      "retweet_count               -1.236e-06    1.9e-08    -65.169      0.000   -1.27e-06    -1.2e-06\n",
      "reply_count                  4.044e-05   2.58e-05      1.570      0.116   -1.01e-05    9.09e-05\n",
      "like_count                  -7.841e-07   9.92e-07     -0.790      0.430   -2.73e-06    1.16e-06\n",
      "quote_count                 -9.811e-06   2.96e-05     -0.332      0.740   -6.78e-05    4.81e-05\n",
      "==============================================================================\n",
      "Omnibus:                   325776.608   Durbin-Watson:                   1.985\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          7545589.439\n",
      "Skew:                           4.238   Prob(JB):                         0.00\n",
      "Kurtosis:                      23.103   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.060\n",
      "Model:                            OLS   Adj. R-squared:                  0.060\n",
      "Method:                 Least Squares   F-statistic:                     3441.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:04   Log-Likelihood:                -66074.\n",
      "No. Observations:              380462   AIC:                         1.322e+05\n",
      "Df Residuals:                  380454   BIC:                         1.323e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1244      0.002     58.448      0.000       0.120       0.129\n",
      "days_from_event                -0.0212      0.001    -16.816      0.000      -0.024      -0.019\n",
      "after_event                     0.0510      0.002     21.786      0.000       0.046       0.056\n",
      "after_event:days_from_event     0.0396      0.002     20.102      0.000       0.036       0.043\n",
      "retweet_count                4.635e-06   3.51e-08    132.064      0.000    4.57e-06     4.7e-06\n",
      "reply_count                  7.077e-05   4.77e-05      1.485      0.137   -2.26e-05       0.000\n",
      "like_count                  -1.568e-06   1.84e-06     -0.854      0.393   -5.17e-06    2.03e-06\n",
      "quote_count                  1.668e-05   5.47e-05      0.305      0.760   -9.05e-05       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    80317.115   Durbin-Watson:                   1.971\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           144869.663\n",
      "Skew:                           1.362   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.311   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.004\n",
      "Model:                            OLS   Adj. R-squared:                  0.004\n",
      "Method:                 Least Squares   F-statistic:                     194.3\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          5.91e-289\n",
      "Time:                        18:35:05   Log-Likelihood:                -90244.\n",
      "No. Observations:              380462   AIC:                         1.805e+05\n",
      "Df Residuals:                  380454   BIC:                         1.806e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1583      0.002     69.808      0.000       0.154       0.163\n",
      "days_from_event                -0.0290      0.001    -21.597      0.000      -0.032      -0.026\n",
      "after_event                     0.0353      0.002     14.162      0.000       0.030       0.040\n",
      "after_event:days_from_event     0.0289      0.002     13.778      0.000       0.025       0.033\n",
      "retweet_count                1.042e-06   3.74e-08     27.876      0.000    9.69e-07    1.12e-06\n",
      "reply_count                 -7.434e-05   5.08e-05     -1.464      0.143      -0.000    2.52e-05\n",
      "like_count                   2.831e-06   1.96e-06      1.447      0.148      -1e-06    6.67e-06\n",
      "quote_count                  1.818e-06   5.83e-05      0.031      0.975      -0.000       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    86787.170   Durbin-Watson:                   1.967\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           160085.930\n",
      "Skew:                           1.482   Prob(JB):                         0.00\n",
      "Kurtosis:                       4.145   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.006\n",
      "Model:                            OLS   Adj. R-squared:                  0.006\n",
      "Method:                 Least Squares   F-statistic:                     336.1\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:06   Log-Likelihood:            -1.3473e+05\n",
      "No. Observations:              380462   AIC:                         2.695e+05\n",
      "Df Residuals:                  380454   BIC:                         2.696e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2811      0.003    110.270      0.000       0.276       0.286\n",
      "days_from_event                -0.0167      0.002    -11.036      0.000      -0.020      -0.014\n",
      "after_event                     0.0087      0.003      3.089      0.002       0.003       0.014\n",
      "after_event:days_from_event     0.0335      0.002     14.192      0.000       0.029       0.038\n",
      "retweet_count               -1.754e-06    4.2e-08    -41.729      0.000   -1.84e-06   -1.67e-06\n",
      "reply_count                   3.12e-05   5.71e-05      0.547      0.585   -8.07e-05       0.000\n",
      "like_count                  -1.657e-06    2.2e-06     -0.754      0.451   -5.97e-06    2.65e-06\n",
      "quote_count                  2.866e-05   6.55e-05      0.437      0.662   -9.98e-05       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    42121.993   Durbin-Watson:                   1.959\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            53589.607\n",
      "Skew:                           0.898   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.602   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     725.4\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:06   Log-Likelihood:             1.7000e+05\n",
      "No. Observations:              380462   AIC:                        -3.400e+05\n",
      "Df Residuals:                  380454   BIC:                        -3.399e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0408      0.001     35.677      0.000       0.039       0.043\n",
      "days_from_event                -0.0150      0.001    -22.179      0.000      -0.016      -0.014\n",
      "after_event                     0.0095      0.001      7.577      0.000       0.007       0.012\n",
      "after_event:days_from_event     0.0081      0.001      7.676      0.000       0.006       0.010\n",
      "retweet_count               -8.443e-07   1.89e-08    -44.741      0.000   -8.81e-07   -8.07e-07\n",
      "reply_count                  2.288e-06   2.56e-05      0.089      0.929   -4.79e-05    5.25e-05\n",
      "like_count                  -2.591e-07   9.87e-07     -0.263      0.793   -2.19e-06    1.68e-06\n",
      "quote_count                 -6.101e-06   2.94e-05     -0.207      0.836   -6.37e-05    5.15e-05\n",
      "==============================================================================\n",
      "Omnibus:                   286162.759   Durbin-Watson:                   1.997\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          4386001.958\n",
      "Skew:                           3.646   Prob(JB):                         0.00\n",
      "Kurtosis:                      17.950   Cond. No.                     1.26e+05\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.26e+05. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.025\n",
      "Model:                            OLS   Adj. R-squared:                  0.025\n",
      "Method:                 Least Squares   F-statistic:                     500.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:50   Log-Likelihood:                -70939.\n",
      "No. Observations:              134489   AIC:                         1.419e+05\n",
      "Df Residuals:                  134481   BIC:                         1.420e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.0040      0.003     -1.328      0.184      -0.010       0.002\n",
      "days_from_event                -0.0004      0.000     -3.300      0.001      -0.001      -0.000\n",
      "after_event                     0.0765      0.005     16.867      0.000       0.068       0.085\n",
      "after_event:days_from_event    -0.0001      0.000     -0.749      0.454      -0.001       0.000\n",
      "retweet_count               -1.997e-05   3.93e-07    -50.841      0.000   -2.07e-05   -1.92e-05\n",
      "reply_count                     0.0001      0.000      0.364      0.716      -0.000       0.001\n",
      "like_count                    1.37e-06   4.78e-06      0.286      0.775   -8.01e-06    1.07e-05\n",
      "quote_count                    -0.0004      0.000     -0.993      0.321      -0.001       0.000\n",
      "==============================================================================\n",
      "Omnibus:                      206.745   Durbin-Watson:                   1.647\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              171.906\n",
      "Skew:                           0.004   Prob(JB):                     4.69e-38\n",
      "Kurtosis:                       2.825   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.018\n",
      "Model:                            OLS   Adj. R-squared:                  0.018\n",
      "Method:                 Least Squares   F-statistic:                     349.5\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:50   Log-Likelihood:             1.1482e+05\n",
      "No. Observations:              134489   AIC:                        -2.296e+05\n",
      "Df Residuals:                  134481   BIC:                        -2.295e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0848      0.001    113.149      0.000       0.083       0.086\n",
      "days_from_event                 0.0007   3.24e-05     20.876      0.000       0.001       0.001\n",
      "after_event                    -0.0298      0.001    -26.130      0.000      -0.032      -0.028\n",
      "after_event:days_from_event    -0.0008   4.89e-05    -17.242      0.000      -0.001      -0.001\n",
      "retweet_count                2.433e-06   9.87e-08     24.646      0.000    2.24e-06    2.63e-06\n",
      "reply_count                     0.0001   7.34e-05      1.962      0.050    1.61e-07       0.000\n",
      "like_count                   1.095e-06    1.2e-06      0.911      0.362   -1.26e-06    3.45e-06\n",
      "quote_count                 -3.429e-05      0.000     -0.304      0.761      -0.000       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    44339.484   Durbin-Watson:                   1.672\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           129004.321\n",
      "Skew:                           1.758   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.265   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.023\n",
      "Model:                            OLS   Adj. R-squared:                  0.023\n",
      "Method:                 Least Squares   F-statistic:                     455.2\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:50   Log-Likelihood:             1.1841e+05\n",
      "No. Observations:              134489   AIC:                        -2.368e+05\n",
      "Df Residuals:                  134481   BIC:                        -2.367e+05\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0755      0.001    103.498      0.000       0.074       0.077\n",
      "days_from_event                 0.0002   3.15e-05      6.845      0.000       0.000       0.000\n",
      "after_event                     0.0063      0.001      5.690      0.000       0.004       0.008\n",
      "after_event:days_from_event    -0.0009   4.76e-05    -18.103      0.000      -0.001      -0.001\n",
      "retweet_count                -5.04e-06   9.61e-08    -52.443      0.000   -5.23e-06   -4.85e-06\n",
      "reply_count                     0.0002   7.15e-05      2.220      0.026    1.86e-05       0.000\n",
      "like_count                   6.607e-07   1.17e-06      0.565      0.572   -1.63e-06    2.95e-06\n",
      "quote_count                    -0.0001      0.000     -1.177      0.239      -0.000    8.61e-05\n",
      "==============================================================================\n",
      "Omnibus:                    38476.180   Durbin-Watson:                   1.707\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            93785.807\n",
      "Skew:                           1.603   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.540   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.009\n",
      "Model:                            OLS   Adj. R-squared:                  0.009\n",
      "Method:                 Least Squares   F-statistic:                     172.3\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          4.12e-255\n",
      "Time:                        18:35:50   Log-Likelihood:                 43858.\n",
      "No. Observations:              134489   AIC:                        -8.770e+04\n",
      "Df Residuals:                  134481   BIC:                        -8.762e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0327      0.001     25.789      0.000       0.030       0.035\n",
      "days_from_event                -0.0006   5.49e-05    -10.409      0.000      -0.001      -0.000\n",
      "after_event                    -0.0066      0.002     -3.426      0.001      -0.010      -0.003\n",
      "after_event:days_from_event     0.0020   8.29e-05     24.665      0.000       0.002       0.002\n",
      "retweet_count               -3.095e-06   1.67e-07    -18.496      0.000   -3.42e-06   -2.77e-06\n",
      "reply_count                     0.0001      0.000      1.054      0.292      -0.000       0.000\n",
      "like_count                  -4.333e-07   2.04e-06     -0.213      0.832   -4.43e-06    3.56e-06\n",
      "quote_count                    -0.0001      0.000     -0.621      0.535      -0.000       0.000\n",
      "==============================================================================\n",
      "Omnibus:                   113712.048   Durbin-Watson:                   1.650\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2275063.140\n",
      "Skew:                           4.226   Prob(JB):                         0.00\n",
      "Kurtosis:                      21.291   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.008\n",
      "Model:                            OLS   Adj. R-squared:                  0.008\n",
      "Method:                 Least Squares   F-statistic:                     146.6\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          1.78e-216\n",
      "Time:                        18:35:51   Log-Likelihood:                -26116.\n",
      "No. Observations:              134489   AIC:                         5.225e+04\n",
      "Df Residuals:                  134481   BIC:                         5.233e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1382      0.002     64.698      0.000       0.134       0.142\n",
      "days_from_event                -0.0014   9.23e-05    -15.673      0.000      -0.002      -0.001\n",
      "after_event                     0.0389      0.003     11.956      0.000       0.032       0.045\n",
      "after_event:days_from_event    -0.0002      0.000     -1.182      0.237      -0.000       0.000\n",
      "retweet_count               -5.509e-06   2.82e-07    -19.571      0.000   -6.06e-06   -4.96e-06\n",
      "reply_count                     0.0004      0.000      1.962      0.050    3.74e-07       0.001\n",
      "like_count                  -4.773e-06   3.43e-06     -1.392      0.164   -1.15e-05    1.95e-06\n",
      "quote_count                     0.0003      0.000      0.895      0.371      -0.000       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    44047.964   Durbin-Watson:                   1.747\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           109012.613\n",
      "Skew:                           1.871   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.335   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.024\n",
      "Model:                            OLS   Adj. R-squared:                  0.024\n",
      "Method:                 Least Squares   F-statistic:                     475.5\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:51   Log-Likelihood:                -23894.\n",
      "No. Observations:              134489   AIC:                         4.780e+04\n",
      "Df Residuals:                  134481   BIC:                         4.788e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1064      0.002     50.649      0.000       0.102       0.111\n",
      "days_from_event                -0.0005   9.08e-05     -5.348      0.000      -0.001      -0.000\n",
      "after_event                    -0.0392      0.003    -12.255      0.000      -0.045      -0.033\n",
      "after_event:days_from_event     0.0054      0.000     39.068      0.000       0.005       0.006\n",
      "retweet_count                 5.58e-07   2.77e-07      2.015      0.044    1.52e-08     1.1e-06\n",
      "reply_count                     0.0003      0.000      1.307      0.191      -0.000       0.001\n",
      "like_count                   2.889e-06   3.37e-06      0.857      0.392   -3.72e-06     9.5e-06\n",
      "quote_count                    -0.0004      0.000     -1.307      0.191      -0.001       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    47455.608   Durbin-Watson:                   1.647\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           128734.283\n",
      "Skew:                           1.961   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.756   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     252.2\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:51   Log-Likelihood:                -49766.\n",
      "No. Observations:              134489   AIC:                         9.955e+04\n",
      "Df Residuals:                  134481   BIC:                         9.963e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2277      0.003     89.389      0.000       0.223       0.233\n",
      "days_from_event                 0.0008      0.000      6.942      0.000       0.001       0.001\n",
      "after_event                     0.0450      0.004     11.614      0.000       0.037       0.053\n",
      "after_event:days_from_event    -0.0008      0.000     -4.783      0.000      -0.001      -0.000\n",
      "retweet_count               -8.206e-06   3.36e-07    -24.450      0.000   -8.86e-06   -7.55e-06\n",
      "reply_count                     0.0004      0.000      1.437      0.151      -0.000       0.001\n",
      "like_count                  -1.613e-06   4.09e-06     -0.395      0.693   -9.62e-06     6.4e-06\n",
      "quote_count                    -0.0002      0.000     -0.457      0.648      -0.001       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    20648.431   Durbin-Watson:                   1.601\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            32087.831\n",
      "Skew:                           1.196   Prob(JB):                         0.00\n",
      "Kurtosis:                       3.004   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     247.8\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:35:51   Log-Likelihood:                 241.36\n",
      "No. Observations:              134489   AIC:                            -466.7\n",
      "Df Residuals:                  134481   BIC:                            -388.2\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1391      0.002     79.202      0.000       0.136       0.143\n",
      "days_from_event                 0.0023   7.59e-05     29.700      0.000       0.002       0.002\n",
      "after_event                    -0.0346      0.003    -12.952      0.000      -0.040      -0.029\n",
      "after_event:days_from_event    -0.0017      0.000    -14.438      0.000      -0.002      -0.001\n",
      "retweet_count               -4.767e-06   2.31e-07    -20.598      0.000   -5.22e-06   -4.31e-06\n",
      "reply_count                 -6.599e-05      0.000     -0.383      0.701      -0.000       0.000\n",
      "like_count                  -1.179e-06   2.82e-06     -0.418      0.676    -6.7e-06    4.34e-06\n",
      "quote_count                     0.0001      0.000      0.409      0.682      -0.000       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    70930.619   Durbin-Watson:                   1.640\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           385055.158\n",
      "Skew:                           2.636   Prob(JB):                         0.00\n",
      "Kurtosis:                       9.397   Cond. No.                     1.35e+04\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.35e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               compound   R-squared:                       0.026\n",
      "Model:                            OLS   Adj. R-squared:                  0.026\n",
      "Method:                 Least Squares   F-statistic:                     190.1\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          2.13e-279\n",
      "Time:                        18:36:30   Log-Likelihood:                -24370.\n",
      "No. Observations:               50341   AIC:                         4.876e+04\n",
      "Df Residuals:                   50333   BIC:                         4.883e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1537      0.006     26.383      0.000       0.142       0.165\n",
      "days_from_event                 0.0106      0.001     10.669      0.000       0.009       0.013\n",
      "after_event                    -0.1617      0.007    -21.792      0.000      -0.176      -0.147\n",
      "after_event:days_from_event    -0.0099      0.001     -7.520      0.000      -0.012      -0.007\n",
      "retweet_count               -2.354e-05   1.14e-06    -20.647      0.000   -2.58e-05   -2.13e-05\n",
      "reply_count                    -0.0003      0.000     -0.833      0.405      -0.001       0.000\n",
      "like_count                  -3.357e-06   6.59e-06     -0.509      0.610   -1.63e-05    9.56e-06\n",
      "quote_count                     0.0005      0.001      0.794      0.427      -0.001       0.002\n",
      "==============================================================================\n",
      "Omnibus:                        9.698   Durbin-Watson:                   1.694\n",
      "Prob(Omnibus):                  0.008   Jarque-Bera (JB):                9.471\n",
      "Skew:                           0.021   Prob(JB):                      0.00878\n",
      "Kurtosis:                       2.948   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    neg   R-squared:                       0.013\n",
      "Model:                            OLS   Adj. R-squared:                  0.013\n",
      "Method:                 Least Squares   F-statistic:                     94.88\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          3.11e-138\n",
      "Time:                        18:36:30   Log-Likelihood:                 41419.\n",
      "No. Observations:               50341   AIC:                        -8.282e+04\n",
      "Df Residuals:                   50333   BIC:                        -8.275e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0659      0.002     41.789      0.000       0.063       0.069\n",
      "days_from_event                 0.0017      0.000      6.232      0.000       0.001       0.002\n",
      "after_event                     0.0259      0.002     12.898      0.000       0.022       0.030\n",
      "after_event:days_from_event    -0.0064      0.000    -18.014      0.000      -0.007      -0.006\n",
      "retweet_count               -1.123e-06   3.09e-07     -3.640      0.000   -1.73e-06   -5.18e-07\n",
      "reply_count                     0.0002      0.000      1.941      0.052   -2.06e-06       0.000\n",
      "like_count                   8.774e-07   1.78e-06      0.492      0.623   -2.62e-06    4.37e-06\n",
      "quote_count                    -0.0002      0.000     -1.344      0.179      -0.001       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    17614.499   Durbin-Watson:                   1.769\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            54470.610\n",
      "Skew:                           1.843   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.518   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    pos   R-squared:                       0.044\n",
      "Model:                            OLS   Adj. R-squared:                  0.043\n",
      "Method:                 Least Squares   F-statistic:                     327.4\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:36:30   Log-Likelihood:                 43177.\n",
      "No. Observations:               50341   AIC:                        -8.634e+04\n",
      "Df Residuals:                   50333   BIC:                        -8.627e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1193      0.002     78.357      0.000       0.116       0.122\n",
      "days_from_event                 0.0062      0.000     23.858      0.000       0.006       0.007\n",
      "after_event                    -0.0434      0.002    -22.373      0.000      -0.047      -0.040\n",
      "after_event:days_from_event    -0.0094      0.000    -27.299      0.000      -0.010      -0.009\n",
      "retweet_count               -8.137e-06   2.98e-07    -27.300      0.000   -8.72e-06   -7.55e-06\n",
      "reply_count                  2.767e-05      0.000      0.264      0.792      -0.000       0.000\n",
      "like_count                   3.169e-07   1.72e-06      0.184      0.854   -3.06e-06    3.69e-06\n",
      "quote_count                    -0.0001      0.000     -0.699      0.484      -0.000       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    16084.978   Durbin-Watson:                   1.799\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            45582.894\n",
      "Skew:                           1.711   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.165   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Angry   R-squared:                       0.011\n",
      "Model:                            OLS   Adj. R-squared:                  0.011\n",
      "Method:                 Least Squares   F-statistic:                     82.69\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          4.41e-120\n",
      "Time:                        18:36:30   Log-Likelihood:                 17861.\n",
      "No. Observations:               50341   AIC:                        -3.571e+04\n",
      "Df Residuals:                   50333   BIC:                        -3.563e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.0096      0.003      3.828      0.000       0.005       0.015\n",
      "days_from_event                -0.0077      0.000    -17.951      0.000      -0.009      -0.007\n",
      "after_event                     0.0307      0.003      9.586      0.000       0.024       0.037\n",
      "after_event:days_from_event     0.0079      0.001     13.962      0.000       0.007       0.009\n",
      "retweet_count               -8.408e-06   4.93e-07    -17.060      0.000   -9.37e-06   -7.44e-06\n",
      "reply_count                    -0.0001      0.000     -0.646      0.518      -0.000       0.000\n",
      "like_count                  -7.563e-07   2.85e-06     -0.266      0.791   -6.34e-06    4.82e-06\n",
      "quote_count                     0.0001      0.000      0.441      0.659      -0.000       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    46942.621   Durbin-Watson:                   1.758\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          1284168.316\n",
      "Skew:                           4.776   Prob(JB):                         0.00\n",
      "Kurtosis:                      25.825   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                    Sad   R-squared:                       0.011\n",
      "Model:                            OLS   Adj. R-squared:                  0.011\n",
      "Method:                 Least Squares   F-statistic:                     78.66\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          4.57e-114\n",
      "Time:                        18:36:30   Log-Likelihood:                -6767.6\n",
      "No. Observations:               50341   AIC:                         1.355e+04\n",
      "Df Residuals:                   50333   BIC:                         1.362e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.2126      0.004     51.779      0.000       0.205       0.221\n",
      "days_from_event                 0.0081      0.001     11.584      0.000       0.007       0.009\n",
      "after_event                    -0.0854      0.005    -16.323      0.000      -0.096      -0.075\n",
      "after_event:days_from_event    -0.0094      0.001    -10.144      0.000      -0.011      -0.008\n",
      "retweet_count               -2.576e-06   8.04e-07     -3.205      0.001   -4.15e-06      -1e-06\n",
      "reply_count                     0.0011      0.000      3.910      0.000       0.001       0.002\n",
      "like_count                  -7.362e-07   4.64e-06     -0.159      0.874   -9.84e-06    8.37e-06\n",
      "quote_count                    -0.0012      0.000     -2.669      0.008      -0.002      -0.000\n",
      "==============================================================================\n",
      "Omnibus:                    18852.703   Durbin-Watson:                   1.789\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            56370.835\n",
      "Skew:                           2.024   Prob(JB):                         0.00\n",
      "Kurtosis:                       6.239   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:               Surprise   R-squared:                       0.145\n",
      "Model:                            OLS   Adj. R-squared:                  0.145\n",
      "Method:                 Least Squares   F-statistic:                     1217.\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:36:30   Log-Likelihood:                -11664.\n",
      "No. Observations:               50341   AIC:                         2.334e+04\n",
      "Df Residuals:                   50333   BIC:                         2.341e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                      -0.0094      0.005     -2.085      0.037      -0.018      -0.001\n",
      "days_from_event                -0.0487      0.001    -63.143      0.000      -0.050      -0.047\n",
      "after_event                     0.0909      0.006     15.774      0.000       0.080       0.102\n",
      "after_event:days_from_event     0.0502      0.001     49.178      0.000       0.048       0.052\n",
      "retweet_count                2.199e-05   8.86e-07     24.818      0.000    2.03e-05    2.37e-05\n",
      "reply_count                  4.763e-05      0.000      0.153      0.878      -0.001       0.001\n",
      "like_count                   3.551e-06   5.12e-06      0.694      0.488   -6.48e-06    1.36e-05\n",
      "quote_count                    -0.0005      0.000     -1.080      0.280      -0.001       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    13372.956   Durbin-Watson:                   1.696\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            29364.957\n",
      "Skew:                           1.549   Prob(JB):                         0.00\n",
      "Kurtosis:                       5.097   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Fear   R-squared:                       0.073\n",
      "Model:                            OLS   Adj. R-squared:                  0.073\n",
      "Method:                 Least Squares   F-statistic:                     565.1\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):               0.00\n",
      "Time:                        18:36:31   Log-Likelihood:                -21938.\n",
      "No. Observations:               50341   AIC:                         4.389e+04\n",
      "Df Residuals:                   50333   BIC:                         4.396e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.4281      0.006     77.125      0.000       0.417       0.439\n",
      "days_from_event                 0.0325      0.001     34.393      0.000       0.031       0.034\n",
      "after_event                     0.0215      0.007      3.034      0.002       0.008       0.035\n",
      "after_event:days_from_event    -0.0705      0.001    -56.314      0.000      -0.073      -0.068\n",
      "retweet_count                -2.24e-05   1.09e-06    -20.618      0.000   -2.45e-05   -2.03e-05\n",
      "reply_count                     0.0001      0.000      0.307      0.759      -0.001       0.001\n",
      "like_count                  -7.404e-06   6.28e-06     -1.179      0.238   -1.97e-05     4.9e-06\n",
      "quote_count                     0.0010      0.001      1.604      0.109      -0.000       0.002\n",
      "==============================================================================\n",
      "Omnibus:                     6529.871   Durbin-Watson:                   1.728\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             9250.083\n",
      "Skew:                           1.041   Prob(JB):                         0.00\n",
      "Kurtosis:                       2.732   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  Happy   R-squared:                       0.015\n",
      "Model:                            OLS   Adj. R-squared:                  0.015\n",
      "Method:                 Least Squares   F-statistic:                     112.4\n",
      "Date:                Tue, 18 Apr 2023   Prob (F-statistic):          2.23e-164\n",
      "Time:                        18:36:31   Log-Likelihood:                 7612.3\n",
      "No. Observations:               50341   AIC:                        -1.521e+04\n",
      "Df Residuals:                   50333   BIC:                        -1.514e+04\n",
      "Df Model:                           7                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===============================================================================================\n",
      "                                  coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------\n",
      "Intercept                       0.1433      0.003     46.445      0.000       0.137       0.149\n",
      "days_from_event                 0.0073      0.001     13.797      0.000       0.006       0.008\n",
      "after_event                    -0.0858      0.004    -21.837      0.000      -0.094      -0.078\n",
      "after_event:days_from_event    -0.0065      0.001     -9.401      0.000      -0.008      -0.005\n",
      "retweet_count                2.125e-06   6.04e-07      3.518      0.000    9.41e-07    3.31e-06\n",
      "reply_count                 -9.393e-05      0.000     -0.442      0.658      -0.001       0.000\n",
      "like_count                   3.575e-06   3.49e-06      1.024      0.306   -3.27e-06    1.04e-05\n",
      "quote_count                    -0.0002      0.000     -0.696      0.486      -0.001       0.000\n",
      "==============================================================================\n",
      "Omnibus:                    30986.346   Durbin-Watson:                   1.797\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           255246.818\n",
      "Skew:                           2.994   Prob(JB):                         0.00\n",
      "Kurtosis:                      12.265   Cond. No.                     8.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 8.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    }
   ],
   "source": [
    "## Adding covariates\n",
    "indices = [1, 6, 13, 14]\n",
    "\n",
    "for cutoff_index in indices:\n",
    "    fin_df = df_en.copy()\n",
    "\n",
    "    date0 = jump_dates[cutoff_index - 1]\n",
    "    date1 = jump_dates[cutoff_index]\n",
    "    date2 = jump_dates[cutoff_index + 1]\n",
    "    \n",
    "    days1 = int(str(np.datetime64(date1) - np.datetime64(date0)).replace(' days', ''))\n",
    "    days2 = int(str(np.datetime64(date2) - np.datetime64(date1)).replace(' days', ''))\n",
    "\n",
    "    window = min(days1, days2)\n",
    "\n",
    "    fin_df = fin_df[(fin_df['created_at_date']>np.datetime64(date0))&\n",
    "                    (fin_df['created_at_date']<np.datetime64(date2))]\n",
    "\n",
    "    fin_df['after_event'] = fin_df.apply(lambda x: event_occured(x['created_at_date'], str(date1)), axis = 1)\n",
    "    fin_df['days_from_event'] = fin_df.apply(lambda x: days_from_event(x['created_at_date'], str(date1)), axis = 1)\n",
    "    \n",
    "    fin_df['after_event'] = fin_df.apply(lambda x: event_occured(x['created_at_date'], str(date1)), axis = 1)\n",
    "    fin_df['days_from_event'] = fin_df.apply(lambda x: days_from_event(x['created_at_date'], str(date1)), axis = 1)\n",
    "    \n",
    "    reg_df = fin_df[(fin_df['days_from_event'] >= -window)\n",
    "                &(fin_df['days_from_event'] <= window)]\n",
    "\n",
    "    for target in ['compound', 'neg', 'pos', 'Angry', 'Sad', 'Surprise', 'Fear', 'Happy']:\n",
    "\n",
    "        result = smf.ols(formula = f\"{target} ~ days_from_event + after_event + after_event * days_from_event + retweet_count + reply_count + like_count + quote_count\", \n",
    "                         data = reg_df).fit()\n",
    "        print(result.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "e2a77d14",
   "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>start</th>\n",
       "      <th>end</th>\n",
       "      <th>slope</th>\n",
       "      <th>intercept</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2022-11-01</td>\n",
       "      <td>2022-11-17</td>\n",
       "      <td>-0.004078</td>\n",
       "      <td>0.104056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2022-11-17</td>\n",
       "      <td>2022-11-22</td>\n",
       "      <td>0.053575</td>\n",
       "      <td>-0.012284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2022-11-22</td>\n",
       "      <td>2022-11-23</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.142909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2022-11-23</td>\n",
       "      <td>2022-11-27</td>\n",
       "      <td>-0.027823</td>\n",
       "      <td>0.143518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2022-11-27</td>\n",
       "      <td>2022-11-28</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.087245</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2022-11-28</td>\n",
       "      <td>2022-12-02</td>\n",
       "      <td>0.004347</td>\n",
       "      <td>0.043131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2022-12-02</td>\n",
       "      <td>2022-12-04</td>\n",
       "      <td>-0.014309</td>\n",
       "      <td>0.041802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2022-12-04</td>\n",
       "      <td>2022-12-05</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.039749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2022-12-05</td>\n",
       "      <td>2022-12-14</td>\n",
       "      <td>-0.017526</td>\n",
       "      <td>0.175185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2022-12-14</td>\n",
       "      <td>2022-12-16</td>\n",
       "      <td>-0.003608</td>\n",
       "      <td>0.047129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2022-12-16</td>\n",
       "      <td>2023-01-05</td>\n",
       "      <td>0.000744</td>\n",
       "      <td>0.080114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2023-01-05</td>\n",
       "      <td>2023-01-06</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.026008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2023-01-06</td>\n",
       "      <td>2023-02-14</td>\n",
       "      <td>0.002275</td>\n",
       "      <td>0.045579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2023-02-14</td>\n",
       "      <td>2023-03-31</td>\n",
       "      <td>0.000684</td>\n",
       "      <td>0.101143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2023-03-31</td>\n",
       "      <td>2023-04-10</td>\n",
       "      <td>0.000715</td>\n",
       "      <td>0.058667</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         start         end     slope  intercept\n",
       "0   2022-11-01  2022-11-17 -0.004078   0.104056\n",
       "1   2022-11-17  2022-11-22  0.053575  -0.012284\n",
       "2   2022-11-22  2022-11-23  0.000000   0.142909\n",
       "3   2022-11-23  2022-11-27 -0.027823   0.143518\n",
       "4   2022-11-27  2022-11-28  0.000000   0.087245\n",
       "5   2022-11-28  2022-12-02  0.004347   0.043131\n",
       "6   2022-12-02  2022-12-04 -0.014309   0.041802\n",
       "7   2022-12-04  2022-12-05  0.000000   0.039749\n",
       "8   2022-12-05  2022-12-14 -0.017526   0.175185\n",
       "9   2022-12-14  2022-12-16 -0.003608   0.047129\n",
       "10  2022-12-16  2023-01-05  0.000744   0.080114\n",
       "11  2023-01-05  2023-01-06  0.000000   0.026008\n",
       "12  2023-01-06  2023-02-14  0.002275   0.045579\n",
       "13  2023-02-14  2023-03-31  0.000684   0.101143\n",
       "14  2023-03-31  2023-04-10  0.000715   0.058667"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "slopes_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "24a9906e",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}