07_linear_models/04_statistical_inference_of_stock_returns_with_statsmodels.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Statistical inference of stock returns with linear regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports & Settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:01.108755Z",
     "start_time": "2021-04-15T20:03:01.105633Z"
    }
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:01.690820Z",
     "start_time": "2021-04-15T20:03:01.113262Z"
    }
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "from statsmodels.api import OLS, add_constant, graphics\n",
    "from statsmodels.graphics.tsaplots import plot_acf\n",
    "from scipy.stats import norm\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:01.693551Z",
     "start_time": "2021-04-15T20:03:01.691724Z"
    }
   },
   "outputs": [],
   "source": [
    "sns.set_style('whitegrid')\n",
    "idx = pd.IndexSlice"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:03.975320Z",
     "start_time": "2021-04-15T20:03:01.694643Z"
    }
   },
   "outputs": [],
   "source": [
    "with pd.HDFStore('data.h5') as store:\n",
    "    data = (store['model_data']\n",
    "            .dropna()\n",
    "            .drop(['open', 'close', 'low', 'high'], axis=1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select Investment Universe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:04.000126Z",
     "start_time": "2021-04-15T20:03:03.976535Z"
    }
   },
   "outputs": [],
   "source": [
    "data = data[data.dollar_vol_rank<100]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:04.019804Z",
     "start_time": "2021-04-15T20:03:04.000999Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "MultiIndex: 109675 entries, ('AAL', Timestamp('2013-07-25 00:00:00')) to ('ZTS', Timestamp('2014-12-04 00:00:00'))\n",
      "Data columns (total 65 columns):\n",
      " #   Column                 Non-Null Count   Dtype  \n",
      "---  ------                 --------------   -----  \n",
      " 0   volume                 109675 non-null  float64\n",
      " 1   dollar_vol             109675 non-null  float64\n",
      " 2   dollar_vol_1m          109675 non-null  float64\n",
      " 3   dollar_vol_rank        109675 non-null  float64\n",
      " 4   rsi                    109675 non-null  float64\n",
      " 5   bb_high                109675 non-null  float64\n",
      " 6   bb_low                 109675 non-null  float64\n",
      " 7   atr                    109675 non-null  float64\n",
      " 8   macd                   109675 non-null  float64\n",
      " 9   return_1d              109675 non-null  float64\n",
      " 10  return_5d              109675 non-null  float64\n",
      " 11  return_10d             109675 non-null  float64\n",
      " 12  return_21d             109675 non-null  float64\n",
      " 13  return_42d             109675 non-null  float64\n",
      " 14  return_63d             109675 non-null  float64\n",
      " 15  return_1d_lag1         109675 non-null  float64\n",
      " 16  return_5d_lag1         109675 non-null  float64\n",
      " 17  return_10d_lag1        109675 non-null  float64\n",
      " 18  return_21d_lag1        109675 non-null  float64\n",
      " 19  return_1d_lag2         109675 non-null  float64\n",
      " 20  return_5d_lag2         109675 non-null  float64\n",
      " 21  return_10d_lag2        109675 non-null  float64\n",
      " 22  return_21d_lag2        109675 non-null  float64\n",
      " 23  return_1d_lag3         109675 non-null  float64\n",
      " 24  return_5d_lag3         109675 non-null  float64\n",
      " 25  return_10d_lag3        109675 non-null  float64\n",
      " 26  return_21d_lag3        109675 non-null  float64\n",
      " 27  return_1d_lag4         109675 non-null  float64\n",
      " 28  return_5d_lag4         109675 non-null  float64\n",
      " 29  return_10d_lag4        109675 non-null  float64\n",
      " 30  return_21d_lag4        109675 non-null  float64\n",
      " 31  return_1d_lag5         109675 non-null  float64\n",
      " 32  return_5d_lag5         109675 non-null  float64\n",
      " 33  return_10d_lag5        109675 non-null  float64\n",
      " 34  return_21d_lag5        109675 non-null  float64\n",
      " 35  target_1d              109675 non-null  float64\n",
      " 36  target_5d              109675 non-null  float64\n",
      " 37  target_10d             109675 non-null  float64\n",
      " 38  target_21d             109675 non-null  float64\n",
      " 39  year_2014              109675 non-null  uint8  \n",
      " 40  year_2015              109675 non-null  uint8  \n",
      " 41  year_2016              109675 non-null  uint8  \n",
      " 42  year_2017              109675 non-null  uint8  \n",
      " 43  month_2                109675 non-null  uint8  \n",
      " 44  month_3                109675 non-null  uint8  \n",
      " 45  month_4                109675 non-null  uint8  \n",
      " 46  month_5                109675 non-null  uint8  \n",
      " 47  month_6                109675 non-null  uint8  \n",
      " 48  month_7                109675 non-null  uint8  \n",
      " 49  month_8                109675 non-null  uint8  \n",
      " 50  month_9                109675 non-null  uint8  \n",
      " 51  month_10               109675 non-null  uint8  \n",
      " 52  month_11               109675 non-null  uint8  \n",
      " 53  month_12               109675 non-null  uint8  \n",
      " 54  capital_goods          109675 non-null  uint8  \n",
      " 55  consumer_durables      109675 non-null  uint8  \n",
      " 56  consumer_non-durables  109675 non-null  uint8  \n",
      " 57  consumer_services      109675 non-null  uint8  \n",
      " 58  energy                 109675 non-null  uint8  \n",
      " 59  finance                109675 non-null  uint8  \n",
      " 60  health_care            109675 non-null  uint8  \n",
      " 61  miscellaneous          109675 non-null  uint8  \n",
      " 62  public_utilities       109675 non-null  uint8  \n",
      " 63  technology             109675 non-null  uint8  \n",
      " 64  transportation         109675 non-null  uint8  \n",
      "dtypes: float64(39), uint8(26)\n",
      "memory usage: 36.5+ MB\n"
     ]
    }
   ],
   "source": [
    "data.info(null_counts=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create Model Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:04.035121Z",
     "start_time": "2021-04-15T20:03:04.020754Z"
    }
   },
   "outputs": [],
   "source": [
    "y = data.filter(like='target')\n",
    "X = data.drop(y.columns, axis=1)\n",
    "X = X.drop(['dollar_vol', 'dollar_vol_rank', 'volume', 'consumer_durables'], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:04.392187Z",
     "start_time": "2021-04-15T20:03:04.036165Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.clustermap(y.corr(), cmap=sns.diverging_palette(h_neg=20, h_pos=220), center=0, annot=True, fmt='.2%');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:05.611740Z",
     "start_time": "2021-04-15T20:03:04.393449Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x1008 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.clustermap(X.corr(), cmap=sns.diverging_palette(h_neg=20, h_pos=220), center=0);\n",
    "plt.gcf().set_size_inches((14, 14))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:06.221840Z",
     "start_time": "2021-04-15T20:03:05.613055Z"
    }
   },
   "outputs": [],
   "source": [
    "corr_mat = X.corr().stack().reset_index()\n",
    "corr_mat.columns=['var1', 'var2', 'corr']\n",
    "corr_mat = corr_mat[corr_mat.var1!=corr_mat.var2].sort_values(by='corr', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:06.230006Z",
     "start_time": "2021-04-15T20:03:06.222689Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>var1</th>\n",
       "      <th>var2</th>\n",
       "      <th>corr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>581</th>\n",
       "      <td>return_42d</td>\n",
       "      <td>return_63d</td>\n",
       "      <td>0.835634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>637</th>\n",
       "      <td>return_63d</td>\n",
       "      <td>return_42d</td>\n",
       "      <td>0.835634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>286</th>\n",
       "      <td>macd</td>\n",
       "      <td>rsi</td>\n",
       "      <td>0.817113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>rsi</td>\n",
       "      <td>macd</td>\n",
       "      <td>0.817113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>518</th>\n",
       "      <td>return_21d</td>\n",
       "      <td>macd</td>\n",
       "      <td>0.793893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>515</th>\n",
       "      <td>return_21d</td>\n",
       "      <td>bb_high</td>\n",
       "      <td>-0.632777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>bb_high</td>\n",
       "      <td>return_10d</td>\n",
       "      <td>-0.693640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>458</th>\n",
       "      <td>return_10d</td>\n",
       "      <td>bb_high</td>\n",
       "      <td>-0.693640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>rsi</td>\n",
       "      <td>bb_high</td>\n",
       "      <td>-0.696555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>bb_high</td>\n",
       "      <td>rsi</td>\n",
       "      <td>-0.696555</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           var1        var2      corr\n",
       "581  return_42d  return_63d  0.835634\n",
       "637  return_63d  return_42d  0.835634\n",
       "286        macd         rsi  0.817113\n",
       "62          rsi        macd  0.817113\n",
       "518  return_21d        macd  0.793893\n",
       "515  return_21d     bb_high -0.632777\n",
       "122     bb_high  return_10d -0.693640\n",
       "458  return_10d     bb_high -0.693640\n",
       "59          rsi     bb_high -0.696555\n",
       "115     bb_high         rsi -0.696555"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_mat.head().append(corr_mat.tail())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:06.570956Z",
     "start_time": "2021-04-15T20:03:06.230912Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y.boxplot();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Linear Regression for Statistical Inference: OLS with statsmodels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ticker-wise standardization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`statsmodels` warns of high design matrix condition numbers. This can arise when the variables are not standardized and the Eigenvalues differ due to scaling. The following step avoids this warning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:15.994024Z",
     "start_time": "2021-04-15T20:03:06.571986Z"
    }
   },
   "outputs": [],
   "source": [
    "sectors = X.iloc[:, -10:]\n",
    "X = (X.drop(sectors.columns, axis=1)\n",
    "     .groupby(level='ticker')\n",
    "     .transform(lambda x: (x - x.mean()) / x.std())\n",
    "    .join(sectors)\n",
    "    .fillna(0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1-Day Returns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:16.222390Z",
     "start_time": "2021-04-15T20:03:15.995122Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:              target_1d   R-squared:                       0.010\n",
      "Model:                            OLS   Adj. R-squared:                  0.009\n",
      "Method:                 Least Squares   F-statistic:                     19.03\n",
      "Date:                Thu, 15 Apr 2021   Prob (F-statistic):          9.43e-189\n",
      "Time:                        15:03:16   Log-Likelihood:             2.8852e+05\n",
      "No. Observations:              109675   AIC:                        -5.769e+05\n",
      "Df Residuals:                  109617   BIC:                        -5.764e+05\n",
      "Df Model:                          57                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "=========================================================================================\n",
      "                            coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------\n",
      "const                    -0.0002      0.000     -0.793      0.428      -0.001       0.000\n",
      "dollar_vol_1m            -0.0004   6.88e-05     -5.214      0.000      -0.000      -0.000\n",
      "rsi                       0.0002      0.000      0.978      0.328      -0.000       0.001\n",
      "bb_high                   0.0002      0.000      0.927      0.354      -0.000       0.001\n",
      "bb_low                    0.0006      0.000      2.940      0.003       0.000       0.001\n",
      "atr                   -3.752e-05   7.44e-05     -0.504      0.614      -0.000       0.000\n",
      "macd                     -0.0004      0.000     -1.835      0.066      -0.001    2.96e-05\n",
      "return_1d                 0.0029      0.000      9.684      0.000       0.002       0.003\n",
      "return_5d                -0.0019      0.001     -2.015      0.044      -0.004   -5.08e-05\n",
      "return_10d               -0.0064      0.001     -6.443      0.000      -0.008      -0.004\n",
      "return_21d                0.0028      0.000      6.242      0.000       0.002       0.004\n",
      "return_42d               -0.0036      0.001     -5.995      0.000      -0.005      -0.002\n",
      "return_63d               -0.0019      0.000     -4.241      0.000      -0.003      -0.001\n",
      "return_1d_lag1            0.0027      0.000      8.998      0.000       0.002       0.003\n",
      "return_5d_lag1            0.0048      0.001      7.033      0.000       0.003       0.006\n",
      "return_10d_lag1          -0.0010      0.001     -0.896      0.370      -0.003       0.001\n",
      "return_21d_lag1           0.0029      0.000      7.667      0.000       0.002       0.004\n",
      "return_1d_lag2            0.0027      0.000      9.177      0.000       0.002       0.003\n",
      "return_5d_lag2            0.0009      0.001      1.141      0.254      -0.001       0.002\n",
      "return_10d_lag2        8.065e-05      0.001      0.097      0.923      -0.002       0.002\n",
      "return_21d_lag2           0.0003      0.000      0.857      0.392      -0.000       0.001\n",
      "return_1d_lag3            0.0027      0.000      8.973      0.000       0.002       0.003\n",
      "return_5d_lag3            0.0015      0.001      1.919      0.055   -3.15e-05       0.003\n",
      "return_10d_lag3           0.0005      0.000      3.457      0.001       0.000       0.001\n",
      "return_21d_lag3          -0.0002   5.55e-05     -3.739      0.000      -0.000   -9.88e-05\n",
      "return_1d_lag4            0.0030      0.000      9.976      0.000       0.002       0.004\n",
      "return_5d_lag4            0.0006      0.001      1.057      0.290      -0.001       0.002\n",
      "return_10d_lag4           0.0004      0.000      3.845      0.000       0.000       0.001\n",
      "return_21d_lag4        4.323e-05    5.5e-05      0.786      0.432   -6.46e-05       0.000\n",
      "return_1d_lag5        -8.924e-05   6.19e-05     -1.441      0.149      -0.000    3.21e-05\n",
      "return_5d_lag5            0.0001      0.001      0.241      0.810      -0.001       0.001\n",
      "return_10d_lag5           0.0007   9.95e-05      7.235      0.000       0.001       0.001\n",
      "return_21d_lag5        9.403e-05   5.48e-05      1.716      0.086   -1.34e-05       0.000\n",
      "year_2014                -0.0004   8.38e-05     -4.430      0.000      -0.001      -0.000\n",
      "year_2015                -0.0006   9.15e-05     -6.804      0.000      -0.001      -0.000\n",
      "year_2016                -0.0005   8.95e-05     -5.113      0.000      -0.001      -0.000\n",
      "year_2017                -0.0002   8.83e-05     -2.024      0.043      -0.000   -5.66e-06\n",
      "month_2                   0.0010   7.21e-05     14.079      0.000       0.001       0.001\n",
      "month_3                   0.0003   7.44e-05      4.637      0.000       0.000       0.000\n",
      "month_4                   0.0005   7.36e-05      7.053      0.000       0.000       0.001\n",
      "month_5                   0.0005   7.25e-05      6.741      0.000       0.000       0.001\n",
      "month_6                   0.0004   7.34e-05      5.696      0.000       0.000       0.001\n",
      "month_7                   0.0006   7.63e-05      8.415      0.000       0.000       0.001\n",
      "month_8                6.821e-05    7.7e-05      0.886      0.375   -8.26e-05       0.000\n",
      "month_9                   0.0004   7.59e-05      5.017      0.000       0.000       0.001\n",
      "month_10                  0.0007    7.8e-05      8.524      0.000       0.001       0.001\n",
      "month_11                  0.0006    7.6e-05      7.697      0.000       0.000       0.001\n",
      "month_12                  0.0004   7.42e-05      5.153      0.000       0.000       0.001\n",
      "capital_goods             0.0010      0.000      2.714      0.007       0.000       0.002\n",
      "consumer_non-durables     0.0007      0.000      1.793      0.073   -6.33e-05       0.001\n",
      "consumer_services         0.0008      0.000      2.234      0.025    9.29e-05       0.001\n",
      "energy                    0.0003      0.000      0.913      0.361      -0.000       0.001\n",
      "finance                   0.0009      0.000      2.533      0.011       0.000       0.002\n",
      "health_care               0.0007      0.000      2.065      0.039    3.51e-05       0.001\n",
      "miscellaneous             0.0011      0.000      2.565      0.010       0.000       0.002\n",
      "public_utilities          0.0003      0.000      0.749      0.454      -0.001       0.001\n",
      "technology                0.0011      0.000      3.285      0.001       0.000       0.002\n",
      "transportation            0.0010      0.000      2.426      0.015       0.000       0.002\n",
      "==============================================================================\n",
      "Omnibus:                    29139.970   Durbin-Watson:                   2.010\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2800989.218\n",
      "Skew:                          -0.105   Prob(JB):                         0.00\n",
      "Kurtosis:                      27.757   Cond. No.                         80.7\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "target = 'target_1d'\n",
    "model = OLS(endog=y[target], exog=add_constant(X))\n",
    "trained_model = model.fit()\n",
    "print(trained_model.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5-Day Returns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:16.455330Z",
     "start_time": "2021-04-15T20:03:16.223433Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:              target_5d   R-squared:                       0.031\n",
      "Model:                            OLS   Adj. R-squared:                  0.031\n",
      "Method:                 Least Squares   F-statistic:                     61.86\n",
      "Date:                Thu, 15 Apr 2021   Prob (F-statistic):               0.00\n",
      "Time:                        15:03:16   Log-Likelihood:             3.7883e+05\n",
      "No. Observations:              109675   AIC:                        -7.575e+05\n",
      "Df Residuals:                  109617   BIC:                        -7.570e+05\n",
      "Df Model:                          57                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "=========================================================================================\n",
      "                            coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------\n",
      "const                    -0.0005      0.000     -3.412      0.001      -0.001      -0.000\n",
      "dollar_vol_1m            -0.0003   3.02e-05     -8.790      0.000      -0.000      -0.000\n",
      "rsi                       0.0005   8.43e-05      5.477      0.000       0.000       0.001\n",
      "bb_high                   0.0010   9.88e-05     10.353      0.000       0.001       0.001\n",
      "bb_low                   -0.0004   9.46e-05     -4.550      0.000      -0.001      -0.000\n",
      "atr                     -9.8e-05   3.27e-05     -2.999      0.003      -0.000    -3.4e-05\n",
      "macd                     -0.0004      0.000     -3.880      0.000      -0.001      -0.000\n",
      "return_1d                 0.0010      0.000      7.468      0.000       0.001       0.001\n",
      "return_5d                 0.0012      0.000      2.947      0.003       0.000       0.002\n",
      "return_10d               -0.0038      0.000     -8.639      0.000      -0.005      -0.003\n",
      "return_21d                0.0021      0.000     10.714      0.000       0.002       0.003\n",
      "return_42d               -0.0024      0.000     -9.124      0.000      -0.003      -0.002\n",
      "return_63d               -0.0017      0.000     -8.800      0.000      -0.002      -0.001\n",
      "return_1d_lag1            0.0009      0.000      7.281      0.000       0.001       0.001\n",
      "return_5d_lag1            0.0035      0.000     11.632      0.000       0.003       0.004\n",
      "return_10d_lag1           0.0007      0.000      1.546      0.122      -0.000       0.002\n",
      "return_21d_lag1           0.0020      0.000     11.876      0.000       0.002       0.002\n",
      "return_1d_lag2            0.0010      0.000      7.493      0.000       0.001       0.001\n",
      "return_5d_lag2         -6.79e-06      0.000     -0.020      0.984      -0.001       0.001\n",
      "return_10d_lag2          -0.0003      0.000     -0.918      0.359      -0.001       0.000\n",
      "return_21d_lag2           0.0006      0.000      4.309      0.000       0.000       0.001\n",
      "return_1d_lag3            0.0010      0.000      7.844      0.000       0.001       0.001\n",
      "return_5d_lag3            0.0001      0.000      0.397      0.691      -0.001       0.001\n",
      "return_10d_lag3           0.0005    6.4e-05      8.171      0.000       0.000       0.001\n",
      "return_21d_lag3          -0.0002   2.44e-05     -7.933      0.000      -0.000      -0.000\n",
      "return_1d_lag4            0.0010      0.000      7.887      0.000       0.001       0.001\n",
      "return_5d_lag4            0.0007      0.000      2.443      0.015       0.000       0.001\n",
      "return_10d_lag4        4.558e-05   4.51e-05      1.012      0.312   -4.27e-05       0.000\n",
      "return_21d_lag4        8.989e-05   2.41e-05      3.724      0.000    4.26e-05       0.000\n",
      "return_1d_lag5         1.138e-05   2.72e-05      0.419      0.675   -4.19e-05    6.46e-05\n",
      "return_5d_lag5            0.0005      0.000      1.848      0.065   -2.91e-05       0.001\n",
      "return_10d_lag5           0.0004   4.37e-05      8.446      0.000       0.000       0.000\n",
      "return_21d_lag5           0.0002    2.4e-05      7.811      0.000       0.000       0.000\n",
      "year_2014                -0.0003   3.68e-05     -9.407      0.000      -0.000      -0.000\n",
      "year_2015                -0.0005   4.02e-05    -13.632      0.000      -0.001      -0.000\n",
      "year_2016                -0.0003   3.93e-05     -8.157      0.000      -0.000      -0.000\n",
      "year_2017                -0.0002   3.88e-05     -4.454      0.000      -0.000   -9.67e-05\n",
      "month_2                   0.0009   3.16e-05     27.541      0.000       0.001       0.001\n",
      "month_3                   0.0001   3.27e-05      3.949      0.000     6.5e-05       0.000\n",
      "month_4                   0.0003   3.23e-05      9.721      0.000       0.000       0.000\n",
      "month_5                   0.0005   3.18e-05     14.481      0.000       0.000       0.001\n",
      "month_6                   0.0002   3.22e-05      7.201      0.000       0.000       0.000\n",
      "month_7                   0.0004   3.35e-05     13.321      0.000       0.000       0.001\n",
      "month_8                3.574e-05   3.38e-05      1.058      0.290   -3.05e-05       0.000\n",
      "month_9                   0.0003   3.33e-05      7.745      0.000       0.000       0.000\n",
      "month_10                  0.0004   3.42e-05     12.425      0.000       0.000       0.000\n",
      "month_11                  0.0005   3.33e-05     13.788      0.000       0.000       0.001\n",
      "month_12                  0.0001   3.26e-05      4.388      0.000    7.91e-05       0.000\n",
      "capital_goods             0.0011      0.000      6.699      0.000       0.001       0.001\n",
      "consumer_non-durables     0.0009      0.000      5.179      0.000       0.001       0.001\n",
      "consumer_services         0.0009      0.000      6.038      0.000       0.001       0.001\n",
      "energy                    0.0004      0.000      2.488      0.013    8.11e-05       0.001\n",
      "finance                   0.0010      0.000      6.510      0.000       0.001       0.001\n",
      "health_care               0.0008      0.000      5.240      0.000       0.000       0.001\n",
      "miscellaneous             0.0012      0.000      6.354      0.000       0.001       0.002\n",
      "public_utilities          0.0003      0.000      1.864      0.062   -1.77e-05       0.001\n",
      "technology                0.0011      0.000      7.760      0.000       0.001       0.001\n",
      "transportation            0.0010      0.000      5.920      0.000       0.001       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    43640.257   Durbin-Watson:                   0.436\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          2362450.641\n",
      "Skew:                          -1.138   Prob(JB):                         0.00\n",
      "Kurtosis:                      25.623   Cond. No.                         80.7\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "target = 'target_5d'\n",
    "model = OLS(endog=y[target], exog=add_constant(X))\n",
    "trained_model = model.fit()\n",
    "print(trained_model.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Obtain the residuals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:16.512662Z",
     "start_time": "2021-04-15T20:03:16.456259Z"
    }
   },
   "outputs": [],
   "source": [
    "preds = trained_model.predict(add_constant(X))\n",
    "residuals = y[target] - preds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:17.799687Z",
     "start_time": "2021-04-15T20:03:16.513542Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(ncols=2, figsize=(14,4))\n",
    "sns.distplot(residuals, fit=norm, ax=axes[0], axlabel='Residuals', label='Residuals')\n",
    "axes[0].set_title('Residual Distribution')\n",
    "axes[0].legend()\n",
    "plot_acf(residuals, lags=10, zero=False, ax=axes[1], title='Residual Autocorrelation')\n",
    "axes[1].set_xlabel('Lags')\n",
    "sns.despine()\n",
    "fig.tight_layout();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 10-Day Returns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:18.025738Z",
     "start_time": "2021-04-15T20:03:17.800716Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:             target_10d   R-squared:                       0.043\n",
      "Model:                            OLS   Adj. R-squared:                  0.042\n",
      "Method:                 Least Squares   F-statistic:                     85.38\n",
      "Date:                Thu, 15 Apr 2021   Prob (F-statistic):               0.00\n",
      "Time:                        15:03:18   Log-Likelihood:             4.1892e+05\n",
      "No. Observations:              109675   AIC:                        -8.377e+05\n",
      "Df Residuals:                  109617   BIC:                        -8.372e+05\n",
      "Df Model:                          57                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "=========================================================================================\n",
      "                            coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------\n",
      "const                    -0.0004    9.5e-05     -3.922      0.000      -0.001      -0.000\n",
      "dollar_vol_1m            -0.0001    2.1e-05     -6.800      0.000      -0.000      -0.000\n",
      "rsi                       0.0002   5.85e-05      2.759      0.006    4.67e-05       0.000\n",
      "bb_high                   0.0006   6.85e-05      9.273      0.000       0.001       0.001\n",
      "bb_low                   -0.0002   6.56e-05     -2.847      0.004      -0.000   -5.82e-05\n",
      "atr                      -0.0002   2.27e-05     -7.328      0.000      -0.000      -0.000\n",
      "macd                     -0.0001   7.22e-05     -1.908      0.056      -0.000    3.76e-06\n",
      "return_1d                 0.0005   9.01e-05      6.046      0.000       0.000       0.001\n",
      "return_5d               5.98e-05      0.000      0.212      0.832      -0.000       0.001\n",
      "return_10d               -0.0008      0.000     -2.741      0.006      -0.001      -0.000\n",
      "return_21d                0.0017      0.000     11.902      0.000       0.001       0.002\n",
      "return_42d               -0.0025      0.000    -13.475      0.000      -0.003      -0.002\n",
      "return_63d               -0.0014      0.000    -10.322      0.000      -0.002      -0.001\n",
      "return_1d_lag1            0.0005   8.98e-05      5.980      0.000       0.000       0.001\n",
      "return_5d_lag1            0.0013      0.000      6.090      0.000       0.001       0.002\n",
      "return_10d_lag1           0.0009      0.000      2.615      0.009       0.000       0.002\n",
      "return_21d_lag1           0.0017      0.000     14.947      0.000       0.002       0.002\n",
      "return_1d_lag2            0.0005   8.99e-05      5.956      0.000       0.000       0.001\n",
      "return_5d_lag2         6.813e-05      0.000      0.294      0.769      -0.000       0.001\n",
      "return_10d_lag2           0.0002      0.000      0.690      0.490      -0.000       0.001\n",
      "return_21d_lag2           0.0002   9.01e-05      2.768      0.006    7.28e-05       0.000\n",
      "return_1d_lag3            0.0005   9.03e-05      6.066      0.000       0.000       0.001\n",
      "return_5d_lag3         9.299e-06      0.000      0.040      0.968      -0.000       0.000\n",
      "return_10d_lag3           0.0005   4.44e-05     12.279      0.000       0.000       0.001\n",
      "return_21d_lag3          -0.0001   1.69e-05     -6.699      0.000      -0.000   -8.02e-05\n",
      "return_1d_lag4            0.0006   9.08e-05      6.278      0.000       0.000       0.001\n",
      "return_5d_lag4            0.0002      0.000      1.250      0.211      -0.000       0.001\n",
      "return_10d_lag4           0.0002   3.13e-05      7.472      0.000       0.000       0.000\n",
      "return_21d_lag4        9.467e-05   1.67e-05      5.653      0.000    6.18e-05       0.000\n",
      "return_1d_lag5        -1.792e-05   1.89e-05     -0.950      0.342   -5.49e-05     1.9e-05\n",
      "return_5d_lag5            0.0002      0.000      1.064      0.287      -0.000       0.001\n",
      "return_10d_lag5           0.0003   3.03e-05     10.693      0.000       0.000       0.000\n",
      "return_21d_lag5           0.0001   1.67e-05      8.453      0.000       0.000       0.000\n",
      "year_2014                -0.0004   2.55e-05    -13.817      0.000      -0.000      -0.000\n",
      "year_2015                -0.0006   2.79e-05    -20.441      0.000      -0.001      -0.001\n",
      "year_2016                -0.0002   2.73e-05     -9.022      0.000      -0.000      -0.000\n",
      "year_2017                -0.0002   2.69e-05     -7.333      0.000      -0.000      -0.000\n",
      "month_2                   0.0006    2.2e-05     28.394      0.000       0.001       0.001\n",
      "month_3                4.204e-05   2.27e-05      1.855      0.064   -2.37e-06    8.65e-05\n",
      "month_4                   0.0002   2.24e-05      8.025      0.000       0.000       0.000\n",
      "month_5                   0.0003   2.21e-05     14.300      0.000       0.000       0.000\n",
      "month_6                   0.0002   2.24e-05      7.894      0.000       0.000       0.000\n",
      "month_7                   0.0002   2.32e-05      8.109      0.000       0.000       0.000\n",
      "month_8                -6.35e-05   2.34e-05     -2.709      0.007      -0.000   -1.76e-05\n",
      "month_9                4.795e-05   2.31e-05      2.075      0.038    2.66e-06    9.32e-05\n",
      "month_10                  0.0003   2.37e-05     14.479      0.000       0.000       0.000\n",
      "month_11                  0.0002   2.31e-05     10.264      0.000       0.000       0.000\n",
      "month_12              -3.354e-05   2.26e-05     -1.485      0.138   -7.78e-05    1.07e-05\n",
      "capital_goods             0.0010      0.000      8.686      0.000       0.001       0.001\n",
      "consumer_non-durables     0.0007      0.000      6.399      0.000       0.001       0.001\n",
      "consumer_services         0.0007      0.000      7.209      0.000       0.001       0.001\n",
      "energy                    0.0003      0.000      2.736      0.006    8.26e-05       0.001\n",
      "finance                   0.0009      0.000      8.318      0.000       0.001       0.001\n",
      "health_care               0.0006      0.000      6.168      0.000       0.000       0.001\n",
      "miscellaneous             0.0012      0.000      8.854      0.000       0.001       0.001\n",
      "public_utilities          0.0002      0.000      1.618      0.106   -4.39e-05       0.000\n",
      "technology                0.0010      0.000      9.984      0.000       0.001       0.001\n",
      "transportation            0.0010      0.000      7.928      0.000       0.001       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    38454.440   Durbin-Watson:                   0.233\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):          1111784.823\n",
      "Skew:                          -1.087   Prob(JB):                         0.00\n",
      "Kurtosis:                      18.446   Cond. No.                         80.7\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "target = 'target_10d'\n",
    "model = OLS(endog=y[target], exog=add_constant(X))\n",
    "trained_model = model.fit()\n",
    "print(trained_model.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Monthly Returns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-04-15T20:03:18.317266Z",
     "start_time": "2021-04-15T20:03:18.026607Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:             target_21d   R-squared:                       0.059\n",
      "Model:                            OLS   Adj. R-squared:                  0.058\n",
      "Method:                 Least Squares   F-statistic:                     119.7\n",
      "Date:                Thu, 15 Apr 2021   Prob (F-statistic):               0.00\n",
      "Time:                        15:03:18   Log-Likelihood:             4.6155e+05\n",
      "No. Observations:              109675   AIC:                        -9.230e+05\n",
      "Df Residuals:                  109617   BIC:                        -9.224e+05\n",
      "Df Model:                          57                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "=========================================================================================\n",
      "                            coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------\n",
      "const                    -0.0003   6.44e-05     -4.358      0.000      -0.000      -0.000\n",
      "dollar_vol_1m         -4.008e-05   1.42e-05     -2.820      0.005   -6.79e-05   -1.22e-05\n",
      "rsi                    2.026e-05   3.97e-05      0.511      0.609   -5.75e-05     9.8e-05\n",
      "bb_high                   0.0002   4.65e-05      3.684      0.000    8.01e-05       0.000\n",
      "bb_low                    0.0002   4.45e-05      4.367      0.000       0.000       0.000\n",
      "atr                      -0.0002   1.54e-05    -12.330      0.000      -0.000      -0.000\n",
      "macd                     -0.0003   4.89e-05     -5.137      0.000      -0.000      -0.000\n",
      "return_1d                 0.0003   6.11e-05      4.853      0.000       0.000       0.000\n",
      "return_5d                -0.0005      0.000     -2.575      0.010      -0.001      -0.000\n",
      "return_10d               -0.0003      0.000     -1.332      0.183      -0.001       0.000\n",
      "return_21d                0.0018   9.41e-05     18.701      0.000       0.002       0.002\n",
      "return_42d               -0.0018      0.000    -14.314      0.000      -0.002      -0.002\n",
      "return_63d               -0.0012   9.31e-05    -12.676      0.000      -0.001      -0.001\n",
      "return_1d_lag1            0.0003   6.09e-05      5.014      0.000       0.000       0.000\n",
      "return_5d_lag1            0.0003      0.000      2.387      0.017    6.01e-05       0.001\n",
      "return_10d_lag1        9.985e-05      0.000      0.445      0.656      -0.000       0.001\n",
      "return_21d_lag1           0.0017   7.93e-05     21.339      0.000       0.002       0.002\n",
      "return_1d_lag2            0.0003   6.09e-05      5.494      0.000       0.000       0.000\n",
      "return_5d_lag2         7.005e-05      0.000      0.446      0.655      -0.000       0.000\n",
      "return_10d_lag2           0.0001      0.000      0.716      0.474      -0.000       0.000\n",
      "return_21d_lag2           0.0005   6.11e-05      7.894      0.000       0.000       0.001\n",
      "return_1d_lag3            0.0004   6.12e-05      5.956      0.000       0.000       0.000\n",
      "return_5d_lag3            0.0001      0.000      0.696      0.486      -0.000       0.000\n",
      "return_10d_lag3           0.0001   3.01e-05      4.171      0.000    6.65e-05       0.000\n",
      "return_21d_lag3       -8.503e-05   1.15e-05     -7.415      0.000      -0.000   -6.26e-05\n",
      "return_1d_lag4            0.0004   6.16e-05      6.187      0.000       0.000       0.001\n",
      "return_5d_lag4         5.046e-05      0.000      0.401      0.688      -0.000       0.000\n",
      "return_10d_lag4        3.895e-05   2.12e-05      1.838      0.066   -2.59e-06    8.05e-05\n",
      "return_21d_lag4           0.0001   1.14e-05     13.100      0.000       0.000       0.000\n",
      "return_1d_lag5         3.351e-06   1.28e-05      0.262      0.793   -2.17e-05    2.84e-05\n",
      "return_5d_lag5        -9.804e-06      0.000     -0.080      0.936      -0.000       0.000\n",
      "return_10d_lag5        6.256e-06   2.05e-05      0.305      0.761    -3.4e-05    4.65e-05\n",
      "return_21d_lag5       -3.545e-05   1.13e-05     -3.134      0.002   -5.76e-05   -1.33e-05\n",
      "year_2014                -0.0003   1.73e-05    -17.088      0.000      -0.000      -0.000\n",
      "year_2015                -0.0006   1.89e-05    -29.827      0.000      -0.001      -0.001\n",
      "year_2016                -0.0002   1.85e-05     -9.707      0.000      -0.000      -0.000\n",
      "year_2017                -0.0002   1.82e-05    -10.028      0.000      -0.000      -0.000\n",
      "month_2                8.855e-05   1.49e-05      5.950      0.000    5.94e-05       0.000\n",
      "month_3                  -0.0001   1.54e-05     -9.594      0.000      -0.000      -0.000\n",
      "month_4                  -0.0001   1.52e-05     -7.960      0.000      -0.000   -9.11e-05\n",
      "month_5                2.356e-05    1.5e-05      1.574      0.116   -5.78e-06    5.29e-05\n",
      "month_6                -1.92e-05   1.52e-05     -1.266      0.206   -4.89e-05    1.05e-05\n",
      "month_7                  -0.0003   1.57e-05    -18.369      0.000      -0.000      -0.000\n",
      "month_8                  -0.0002   1.59e-05    -14.877      0.000      -0.000      -0.000\n",
      "month_9                  -0.0002   1.57e-05    -11.661      0.000      -0.000      -0.000\n",
      "month_10               6.676e-05   1.61e-05      4.147      0.000    3.52e-05    9.83e-05\n",
      "month_11                 -0.0001   1.57e-05     -7.393      0.000      -0.000   -8.52e-05\n",
      "month_12                 -0.0004   1.53e-05    -25.897      0.000      -0.000      -0.000\n",
      "capital_goods             0.0008   7.57e-05     11.198      0.000       0.001       0.001\n",
      "consumer_non-durables     0.0007   7.84e-05      8.618      0.000       0.001       0.001\n",
      "consumer_services         0.0006   7.01e-05      8.933      0.000       0.000       0.001\n",
      "energy                    0.0002   7.22e-05      2.652      0.008    4.99e-05       0.000\n",
      "finance                   0.0008   7.37e-05     10.794      0.000       0.001       0.001\n",
      "health_care               0.0005   6.89e-05      7.493      0.000       0.000       0.001\n",
      "miscellaneous             0.0012   9.02e-05     12.969      0.000       0.001       0.001\n",
      "public_utilities          0.0001   8.71e-05      1.241      0.215   -6.26e-05       0.000\n",
      "technology                0.0009   6.95e-05     13.417      0.000       0.001       0.001\n",
      "transportation            0.0009   8.24e-05     11.165      0.000       0.001       0.001\n",
      "==============================================================================\n",
      "Omnibus:                    40182.050   Durbin-Watson:                   0.129\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           848232.085\n",
      "Skew:                          -1.254   Prob(JB):                         0.00\n",
      "Kurtosis:                      16.391   Cond. No.                         80.7\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "target = 'target_21d'\n",
    "model = OLS(endog=y[target], exog=add_constant(X))\n",
    "trained_model = model.fit()\n",
    "print(trained_model.summary())"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "403px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}