Skip to content
Permalink
master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

stat_learn_lin_regression/exercises_final.ipynb

Go to file
 
 
Cannot retrieve contributors at this time
2779 lines (2779 sloc) 567 KB
Raw Blame
Open in GitHub Desktop
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import os, sys\n",
"import csv\n",
"import matplotlib\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"import pandas as pd\n",
"import statsmodels.api as sm\n",
"import statsmodels.formula.api as smf # http://www.statsmodels.org/dev/example_formulas.html\n",
"from sklearn import datasets, linear_model\n",
"from sklearn.metrics import mean_squared_error, r2_score\n",
"from sklearn.preprocessing import PolynomialFeatures"
]
},
{
"cell_type": "code",
"execution_count": 2,
"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>Unnamed: 0</th>\n",
" <th>crim</th>\n",
" <th>zn</th>\n",
" <th>indus</th>\n",
" <th>chas</th>\n",
" <th>nox</th>\n",
" <th>rm</th>\n",
" <th>age</th>\n",
" <th>dis</th>\n",
" <th>rad</th>\n",
" <th>tax</th>\n",
" <th>ptratio</th>\n",
" <th>black</th>\n",
" <th>lstat</th>\n",
" <th>medv</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1</td>\n",
" <td>0.00632</td>\n",
" <td>18.0</td>\n",
" <td>2.31</td>\n",
" <td>0</td>\n",
" <td>0.538</td>\n",
" <td>6.575</td>\n",
" <td>65.2</td>\n",
" <td>4.0900</td>\n",
" <td>1</td>\n",
" <td>296</td>\n",
" <td>15.3</td>\n",
" <td>396.90</td>\n",
" <td>4.98</td>\n",
" <td>24.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>2</td>\n",
" <td>0.02731</td>\n",
" <td>0.0</td>\n",
" <td>7.07</td>\n",
" <td>0</td>\n",
" <td>0.469</td>\n",
" <td>6.421</td>\n",
" <td>78.9</td>\n",
" <td>4.9671</td>\n",
" <td>2</td>\n",
" <td>242</td>\n",
" <td>17.8</td>\n",
" <td>396.90</td>\n",
" <td>9.14</td>\n",
" <td>21.6</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>3</td>\n",
" <td>0.02729</td>\n",
" <td>0.0</td>\n",
" <td>7.07</td>\n",
" <td>0</td>\n",
" <td>0.469</td>\n",
" <td>7.185</td>\n",
" <td>61.1</td>\n",
" <td>4.9671</td>\n",
" <td>2</td>\n",
" <td>242</td>\n",
" <td>17.8</td>\n",
" <td>392.83</td>\n",
" <td>4.03</td>\n",
" <td>34.7</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4</td>\n",
" <td>0.03237</td>\n",
" <td>0.0</td>\n",
" <td>2.18</td>\n",
" <td>0</td>\n",
" <td>0.458</td>\n",
" <td>6.998</td>\n",
" <td>45.8</td>\n",
" <td>6.0622</td>\n",
" <td>3</td>\n",
" <td>222</td>\n",
" <td>18.7</td>\n",
" <td>394.63</td>\n",
" <td>2.94</td>\n",
" <td>33.4</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5</td>\n",
" <td>0.06905</td>\n",
" <td>0.0</td>\n",
" <td>2.18</td>\n",
" <td>0</td>\n",
" <td>0.458</td>\n",
" <td>7.147</td>\n",
" <td>54.2</td>\n",
" <td>6.0622</td>\n",
" <td>3</td>\n",
" <td>222</td>\n",
" <td>18.7</td>\n",
" <td>396.90</td>\n",
" <td>5.33</td>\n",
" <td>36.2</td>\n",
" </tr>\n",
" <tr>\n",
" <th>5</th>\n",
" <td>6</td>\n",
" <td>0.02985</td>\n",
" <td>0.0</td>\n",
" <td>2.18</td>\n",
" <td>0</td>\n",
" <td>0.458</td>\n",
" <td>6.430</td>\n",
" <td>58.7</td>\n",
" <td>6.0622</td>\n",
" <td>3</td>\n",
" <td>222</td>\n",
" <td>18.7</td>\n",
" <td>394.12</td>\n",
" <td>5.21</td>\n",
" <td>28.7</td>\n",
" </tr>\n",
" <tr>\n",
" <th>6</th>\n",
" <td>7</td>\n",
" <td>0.08829</td>\n",
" <td>12.5</td>\n",
" <td>7.87</td>\n",
" <td>0</td>\n",
" <td>0.524</td>\n",
" <td>6.012</td>\n",
" <td>66.6</td>\n",
" <td>5.5605</td>\n",
" <td>5</td>\n",
" <td>311</td>\n",
" <td>15.2</td>\n",
" <td>395.60</td>\n",
" <td>12.43</td>\n",
" <td>22.9</td>\n",
" </tr>\n",
" <tr>\n",
" <th>7</th>\n",
" <td>8</td>\n",
" <td>0.14455</td>\n",
" <td>12.5</td>\n",
" <td>7.87</td>\n",
" <td>0</td>\n",
" <td>0.524</td>\n",
" <td>6.172</td>\n",
" <td>96.1</td>\n",
" <td>5.9505</td>\n",
" <td>5</td>\n",
" <td>311</td>\n",
" <td>15.2</td>\n",
" <td>396.90</td>\n",
" <td>19.15</td>\n",
" <td>27.1</td>\n",
" </tr>\n",
" <tr>\n",
" <th>8</th>\n",
" <td>9</td>\n",
" <td>0.21124</td>\n",
" <td>12.5</td>\n",
" <td>7.87</td>\n",
" <td>0</td>\n",
" <td>0.524</td>\n",
" <td>5.631</td>\n",
" <td>100.0</td>\n",
" <td>6.0821</td>\n",
" <td>5</td>\n",
" <td>311</td>\n",
" <td>15.2</td>\n",
" <td>386.63</td>\n",
" <td>29.93</td>\n",
" <td>16.5</td>\n",
" </tr>\n",
" <tr>\n",
" <th>9</th>\n",
" <td>10</td>\n",
" <td>0.17004</td>\n",
" <td>12.5</td>\n",
" <td>7.87</td>\n",
" <td>0</td>\n",
" <td>0.524</td>\n",
" <td>6.004</td>\n",
" <td>85.9</td>\n",
" <td>6.5921</td>\n",
" <td>5</td>\n",
" <td>311</td>\n",
" <td>15.2</td>\n",
" <td>386.71</td>\n",
" <td>17.10</td>\n",
" <td>18.9</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Unnamed: 0 crim zn indus chas nox rm age dis rad \\\n",
"0 1 0.00632 18.0 2.31 0 0.538 6.575 65.2 4.0900 1 \n",
"1 2 0.02731 0.0 7.07 0 0.469 6.421 78.9 4.9671 2 \n",
"2 3 0.02729 0.0 7.07 0 0.469 7.185 61.1 4.9671 2 \n",
"3 4 0.03237 0.0 2.18 0 0.458 6.998 45.8 6.0622 3 \n",
"4 5 0.06905 0.0 2.18 0 0.458 7.147 54.2 6.0622 3 \n",
"5 6 0.02985 0.0 2.18 0 0.458 6.430 58.7 6.0622 3 \n",
"6 7 0.08829 12.5 7.87 0 0.524 6.012 66.6 5.5605 5 \n",
"7 8 0.14455 12.5 7.87 0 0.524 6.172 96.1 5.9505 5 \n",
"8 9 0.21124 12.5 7.87 0 0.524 5.631 100.0 6.0821 5 \n",
"9 10 0.17004 12.5 7.87 0 0.524 6.004 85.9 6.5921 5 \n",
"\n",
" tax ptratio black lstat medv \n",
"0 296 15.3 396.90 4.98 24.0 \n",
"1 242 17.8 396.90 9.14 21.6 \n",
"2 242 17.8 392.83 4.03 34.7 \n",
"3 222 18.7 394.63 2.94 33.4 \n",
"4 222 18.7 396.90 5.33 36.2 \n",
"5 222 18.7 394.12 5.21 28.7 \n",
"6 311 15.2 395.60 12.43 22.9 \n",
"7 311 15.2 396.90 19.15 27.1 \n",
"8 311 15.2 386.63 29.93 16.5 \n",
"9 311 15.2 386.71 17.10 18.9 "
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"'''following 3.6 Lab: Linear Regression (page 109)'''\n",
"\n",
"#read in Boston housing data set and getting it into Pandas DataFrame\n",
"fname = 'Boston.csv'\n",
"df = pd.read_csv(fname)\n",
"\n",
"df.head(10) #show first 10 lines of dataset"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>medv</td> <th> R-squared: </th> <td> 0.544</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.543</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 601.6</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>5.08e-88</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:24:43</td> <th> Log-Likelihood: </th> <td> -1641.5</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 506</td> <th> AIC: </th> <td> 3287.</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 504</td> <th> BIC: </th> <td> 3295.</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 1</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>const</th> <td> 34.5538</td> <td> 0.563</td> <td> 61.415</td> <td> 0.000</td> <td> 33.448</td> <td> 35.659</td>\n",
"</tr>\n",
"<tr>\n",
" <th>lstat</th> <td> -0.9500</td> <td> 0.039</td> <td> -24.528</td> <td> 0.000</td> <td> -1.026</td> <td> -0.874</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>137.043</td> <th> Durbin-Watson: </th> <td> 0.892</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 291.373</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.453</td> <th> Prob(JB): </th> <td>5.36e-64</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.319</td> <th> Cond. No. </th> <td> 29.7</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: medv R-squared: 0.544\n",
"Model: OLS Adj. R-squared: 0.543\n",
"Method: Least Squares F-statistic: 601.6\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 5.08e-88\n",
"Time: 08:24:43 Log-Likelihood: -1641.5\n",
"No. Observations: 506 AIC: 3287.\n",
"Df Residuals: 504 BIC: 3295.\n",
"Df Model: 1 \n",
"Covariance Type: nonrobust \n",
"==============================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"------------------------------------------------------------------------------\n",
"const 34.5538 0.563 61.415 0.000 33.448 35.659\n",
"lstat -0.9500 0.039 -24.528 0.000 -1.026 -0.874\n",
"==============================================================================\n",
"Omnibus: 137.043 Durbin-Watson: 0.892\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 291.373\n",
"Skew: 1.453 Prob(JB): 5.36e-64\n",
"Kurtosis: 5.319 Cond. No. 29.7\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"\"\"\""
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"#doing the linear regression using the package statsmodels \n",
"\n",
"'''Lab: 3.6.2 Simple Linear Regression (page 110)'''\n",
"\n",
"X = df[\"lstat\"] #independent variable (or predictor)\n",
"y = df[\"medv\"] #dependent variable (which we want to fit / predict)\n",
"\n",
"X0 = sm.add_constant(X) # we have to manually specify that we want an intercept (beta_0) in our model\n",
"model = sm.OLS(y, X0).fit() #doing the actual linear regression, using the OLS = ordinary least squared\n",
"\n",
"model.summary() #showing the results in a table"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Text(0,0.5,'medv')"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"Y_hat = model.predict(X0) #predicted Y accodring to linear fit\n",
"\n",
"fig1, ax1 = plt.subplots(1,1,figsize=(6,4))\n",
"ax1.scatter(X,y, color='b')\n",
"ax1.plot(X,Y_hat,color='r',linewidth=3)\n",
"ax1.set_xlabel('lstat')\n",
"ax1.set_ylabel('medv')"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"coefficients: \n",
" const 34.553841\n",
"lstat -0.950049\n",
"dtype: float64\n",
"\n",
"R^2: 0.544146297586\n",
"\n",
"pvalues: \n",
" const 3.743081e-236\n",
"lstat 5.081103e-88\n",
"dtype: float64\n"
]
}
],
"source": [
"# to access the important values\"\n",
"print(\"coefficients: \\n\", model.params)\n",
"print()\n",
"print(\"R^2: \", model.rsquared)\n",
"print()\n",
"print(\"pvalues: \\n\",model.pvalues)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>medv</td> <th> R-squared: </th> <td> 0.551</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.549</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 309.0</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>2.98e-88</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:24:43</td> <th> Log-Likelihood: </th> <td> -1637.5</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 506</td> <th> AIC: </th> <td> 3281.</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 503</td> <th> BIC: </th> <td> 3294.</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 2</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>const</th> <td> 33.2228</td> <td> 0.731</td> <td> 45.458</td> <td> 0.000</td> <td> 31.787</td> <td> 34.659</td>\n",
"</tr>\n",
"<tr>\n",
" <th>lstat</th> <td> -1.0321</td> <td> 0.048</td> <td> -21.416</td> <td> 0.000</td> <td> -1.127</td> <td> -0.937</td>\n",
"</tr>\n",
"<tr>\n",
" <th>age</th> <td> 0.0345</td> <td> 0.012</td> <td> 2.826</td> <td> 0.005</td> <td> 0.011</td> <td> 0.059</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>124.288</td> <th> Durbin-Watson: </th> <td> 0.945</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 244.026</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.362</td> <th> Prob(JB): </th> <td>1.02e-53</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.038</td> <th> Cond. No. </th> <td> 201.</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: medv R-squared: 0.551\n",
"Model: OLS Adj. R-squared: 0.549\n",
"Method: Least Squares F-statistic: 309.0\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 2.98e-88\n",
"Time: 08:24:43 Log-Likelihood: -1637.5\n",
"No. Observations: 506 AIC: 3281.\n",
"Df Residuals: 503 BIC: 3294.\n",
"Df Model: 2 \n",
"Covariance Type: nonrobust \n",
"==============================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"------------------------------------------------------------------------------\n",
"const 33.2228 0.731 45.458 0.000 31.787 34.659\n",
"lstat -1.0321 0.048 -21.416 0.000 -1.127 -0.937\n",
"age 0.0345 0.012 2.826 0.005 0.011 0.059\n",
"==============================================================================\n",
"Omnibus: 124.288 Durbin-Watson: 0.945\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 244.026\n",
"Skew: 1.362 Prob(JB): 1.02e-53\n",
"Kurtosis: 5.038 Cond. No. 201.\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"\"\"\""
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"'''Lab: 3.6.3 Multiple Linear Regression (page 113)'''\n",
"\n",
"#with two predictor variables\n",
"X = df[['lstat', 'age']]\n",
"y = df[\"medv\"]\n",
"X0 = sm.add_constant(X) # we have to manually specify that we want an intercept (beta_0) in our model\n",
"\n",
"model = sm.OLS(y, X0).fit()\n",
"model.summary()"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>medv</td> <th> R-squared: </th> <td> 0.741</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.734</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 108.1</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>6.72e-135</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:24:44</td> <th> Log-Likelihood: </th> <td> -1498.8</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 506</td> <th> AIC: </th> <td> 3026.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 492</td> <th> BIC: </th> <td> 3085.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 13</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>const</th> <td> 36.4595</td> <td> 5.103</td> <td> 7.144</td> <td> 0.000</td> <td> 26.432</td> <td> 46.487</td>\n",
"</tr>\n",
"<tr>\n",
" <th>crim</th> <td> -0.1080</td> <td> 0.033</td> <td> -3.287</td> <td> 0.001</td> <td> -0.173</td> <td> -0.043</td>\n",
"</tr>\n",
"<tr>\n",
" <th>zn</th> <td> 0.0464</td> <td> 0.014</td> <td> 3.382</td> <td> 0.001</td> <td> 0.019</td> <td> 0.073</td>\n",
"</tr>\n",
"<tr>\n",
" <th>indus</th> <td> 0.0206</td> <td> 0.061</td> <td> 0.334</td> <td> 0.738</td> <td> -0.100</td> <td> 0.141</td>\n",
"</tr>\n",
"<tr>\n",
" <th>chas</th> <td> 2.6867</td> <td> 0.862</td> <td> 3.118</td> <td> 0.002</td> <td> 0.994</td> <td> 4.380</td>\n",
"</tr>\n",
"<tr>\n",
" <th>nox</th> <td> -17.7666</td> <td> 3.820</td> <td> -4.651</td> <td> 0.000</td> <td> -25.272</td> <td> -10.262</td>\n",
"</tr>\n",
"<tr>\n",
" <th>rm</th> <td> 3.8099</td> <td> 0.418</td> <td> 9.116</td> <td> 0.000</td> <td> 2.989</td> <td> 4.631</td>\n",
"</tr>\n",
"<tr>\n",
" <th>age</th> <td> 0.0007</td> <td> 0.013</td> <td> 0.052</td> <td> 0.958</td> <td> -0.025</td> <td> 0.027</td>\n",
"</tr>\n",
"<tr>\n",
" <th>dis</th> <td> -1.4756</td> <td> 0.199</td> <td> -7.398</td> <td> 0.000</td> <td> -1.867</td> <td> -1.084</td>\n",
"</tr>\n",
"<tr>\n",
" <th>rad</th> <td> 0.3060</td> <td> 0.066</td> <td> 4.613</td> <td> 0.000</td> <td> 0.176</td> <td> 0.436</td>\n",
"</tr>\n",
"<tr>\n",
" <th>tax</th> <td> -0.0123</td> <td> 0.004</td> <td> -3.280</td> <td> 0.001</td> <td> -0.020</td> <td> -0.005</td>\n",
"</tr>\n",
"<tr>\n",
" <th>ptratio</th> <td> -0.9527</td> <td> 0.131</td> <td> -7.283</td> <td> 0.000</td> <td> -1.210</td> <td> -0.696</td>\n",
"</tr>\n",
"<tr>\n",
" <th>black</th> <td> 0.0093</td> <td> 0.003</td> <td> 3.467</td> <td> 0.001</td> <td> 0.004</td> <td> 0.015</td>\n",
"</tr>\n",
"<tr>\n",
" <th>lstat</th> <td> -0.5248</td> <td> 0.051</td> <td> -10.347</td> <td> 0.000</td> <td> -0.624</td> <td> -0.425</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>178.041</td> <th> Durbin-Watson: </th> <td> 1.078</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 783.126</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.521</td> <th> Prob(JB): </th> <td>8.84e-171</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 8.281</td> <th> Cond. No. </th> <td>1.51e+04</td> \n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 1.51e+04. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: medv R-squared: 0.741\n",
"Model: OLS Adj. R-squared: 0.734\n",
"Method: Least Squares F-statistic: 108.1\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 6.72e-135\n",
"Time: 08:24:44 Log-Likelihood: -1498.8\n",
"No. Observations: 506 AIC: 3026.\n",
"Df Residuals: 492 BIC: 3085.\n",
"Df Model: 13 \n",
"Covariance Type: nonrobust \n",
"==============================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"------------------------------------------------------------------------------\n",
"const 36.4595 5.103 7.144 0.000 26.432 46.487\n",
"crim -0.1080 0.033 -3.287 0.001 -0.173 -0.043\n",
"zn 0.0464 0.014 3.382 0.001 0.019 0.073\n",
"indus 0.0206 0.061 0.334 0.738 -0.100 0.141\n",
"chas 2.6867 0.862 3.118 0.002 0.994 4.380\n",
"nox -17.7666 3.820 -4.651 0.000 -25.272 -10.262\n",
"rm 3.8099 0.418 9.116 0.000 2.989 4.631\n",
"age 0.0007 0.013 0.052 0.958 -0.025 0.027\n",
"dis -1.4756 0.199 -7.398 0.000 -1.867 -1.084\n",
"rad 0.3060 0.066 4.613 0.000 0.176 0.436\n",
"tax -0.0123 0.004 -3.280 0.001 -0.020 -0.005\n",
"ptratio -0.9527 0.131 -7.283 0.000 -1.210 -0.696\n",
"black 0.0093 0.003 3.467 0.001 0.004 0.015\n",
"lstat -0.5248 0.051 -10.347 0.000 -0.624 -0.425\n",
"==============================================================================\n",
"Omnibus: 178.041 Durbin-Watson: 1.078\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 783.126\n",
"Skew: 1.521 Prob(JB): 8.84e-171\n",
"Kurtosis: 8.281 Cond. No. 1.51e+04\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 1.51e+04. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# multiple linear regression with all predictor variables \n",
"\n",
"def_new=df.drop([df.columns[0], 'medv'], axis=1) #drop first column (just name of indices) and predicted variable column \n",
"Xall = def_new\n",
"y = df[\"medv\"]\n",
"Xall0 = sm.add_constant(Xall) # we have to manually specify that we want an intercept (beta_0) in our model\n",
"model = sm.OLS(y, Xall0).fit()\n",
"model.summary()"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>medv</td> <th> R-squared: </th> <td> 0.556</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.553</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 209.3</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>4.86e-88</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:24:44</td> <th> Log-Likelihood: </th> <td> -1635.0</td>\n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 506</td> <th> AIC: </th> <td> 3278.</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 502</td> <th> BIC: </th> <td> 3295.</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 3</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> 36.0885</td> <td> 1.470</td> <td> 24.553</td> <td> 0.000</td> <td> 33.201</td> <td> 38.976</td>\n",
"</tr>\n",
"<tr>\n",
" <th>lstat</th> <td> -1.3921</td> <td> 0.167</td> <td> -8.313</td> <td> 0.000</td> <td> -1.721</td> <td> -1.063</td>\n",
"</tr>\n",
"<tr>\n",
" <th>age</th> <td> -0.0007</td> <td> 0.020</td> <td> -0.036</td> <td> 0.971</td> <td> -0.040</td> <td> 0.038</td>\n",
"</tr>\n",
"<tr>\n",
" <th>lstat:age</th> <td> 0.0042</td> <td> 0.002</td> <td> 2.244</td> <td> 0.025</td> <td> 0.001</td> <td> 0.008</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>135.601</td> <th> Durbin-Watson: </th> <td> 0.965</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 296.955</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.417</td> <th> Prob(JB): </th> <td>3.29e-65</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.461</td> <th> Cond. No. </th> <td>6.88e+03</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 6.88e+03. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: medv R-squared: 0.556\n",
"Model: OLS Adj. R-squared: 0.553\n",
"Method: Least Squares F-statistic: 209.3\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 4.86e-88\n",
"Time: 08:24:44 Log-Likelihood: -1635.0\n",
"No. Observations: 506 AIC: 3278.\n",
"Df Residuals: 502 BIC: 3295.\n",
"Df Model: 3 \n",
"Covariance Type: nonrobust \n",
"==============================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"------------------------------------------------------------------------------\n",
"Intercept 36.0885 1.470 24.553 0.000 33.201 38.976\n",
"lstat -1.3921 0.167 -8.313 0.000 -1.721 -1.063\n",
"age -0.0007 0.020 -0.036 0.971 -0.040 0.038\n",
"lstat:age 0.0042 0.002 2.244 0.025 0.001 0.008\n",
"==============================================================================\n",
"Omnibus: 135.601 Durbin-Watson: 0.965\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 296.955\n",
"Skew: 1.417 Prob(JB): 3.29e-65\n",
"Kurtosis: 5.461 Cond. No. 6.88e+03\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 6.88e+03. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"'''Lab: 3.6.4 Interaction terms (page 115)'''\n",
"\n",
"# multiple linear regression including interaction terms (here between lstat and age)\n",
"\n",
"# just for illustration I use smf.ols() which uses the formula syntax known from R, \n",
"# see http://www.statsmodels.org/devel/examples/notebooks/generated/formulas.html\n",
"# for comprehensive overview of how these formula work: http://patsy.readthedocs.io/en/latest/formulas.html#the-formula-language\n",
"\n",
"model_interac = smf.ols(formula='medv ~ lstat * age', data=df).fit() #important: because we use formula, we need smf instead of sm, and ols in lower case\n",
"# equivalent: model_interac = smf.ols(formula='medv ~ lstat + age + lstat:age', data=df).fit()\n",
"model_interac.summary()"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"scrolled": true
},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>medv</td> <th> R-squared: </th> <td> 0.641</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.639</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 448.5</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>1.56e-112</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:24:44</td> <th> Log-Likelihood: </th> <td> -1581.3</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 506</td> <th> AIC: </th> <td> 3169.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 503</td> <th> BIC: </th> <td> 3181.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 2</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> 42.8620</td> <td> 0.872</td> <td> 49.149</td> <td> 0.000</td> <td> 41.149</td> <td> 44.575</td>\n",
"</tr>\n",
"<tr>\n",
" <th>lstat</th> <td> -2.3328</td> <td> 0.124</td> <td> -18.843</td> <td> 0.000</td> <td> -2.576</td> <td> -2.090</td>\n",
"</tr>\n",
"<tr>\n",
" <th>I(lstat ** 2)</th> <td> 0.0435</td> <td> 0.004</td> <td> 11.628</td> <td> 0.000</td> <td> 0.036</td> <td> 0.051</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>107.006</td> <th> Durbin-Watson: </th> <td> 0.921</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 228.388</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 1.128</td> <th> Prob(JB): </th> <td>2.55e-50</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.397</td> <th> Cond. No. </th> <td>1.13e+03</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 1.13e+03. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: medv R-squared: 0.641\n",
"Model: OLS Adj. R-squared: 0.639\n",
"Method: Least Squares F-statistic: 448.5\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 1.56e-112\n",
"Time: 08:24:44 Log-Likelihood: -1581.3\n",
"No. Observations: 506 AIC: 3169.\n",
"Df Residuals: 503 BIC: 3181.\n",
"Df Model: 2 \n",
"Covariance Type: nonrobust \n",
"=================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"---------------------------------------------------------------------------------\n",
"Intercept 42.8620 0.872 49.149 0.000 41.149 44.575\n",
"lstat -2.3328 0.124 -18.843 0.000 -2.576 -2.090\n",
"I(lstat ** 2) 0.0435 0.004 11.628 0.000 0.036 0.051\n",
"==============================================================================\n",
"Omnibus: 107.006 Durbin-Watson: 0.921\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 228.388\n",
"Skew: 1.128 Prob(JB): 2.55e-50\n",
"Kurtosis: 5.397 Cond. No. 1.13e+03\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 1.13e+03. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"'''Lab: 3.6.5 Non-linear transfomartions of predictos (page 115)'''\n",
"\n",
"# e.g. including quadratic terms of predictors\n",
"\n",
"model_nonlin = smf.ols(formula='medv ~ lstat + I(lstat**2)', data=df).fit() #important: because we use formula, we need smf instead of sm, and ols in lower case\n",
"model_nonlin.summary()"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Text(0,0.5,'medv')"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"X = df[\"lstat\"] \n",
"Y_hat_nonlin = model_nonlin.predict(X)\n",
"\n",
"fig2, ax2 = plt.subplots(1,1,figsize=(6,4))\n",
"ax2.scatter(X,y, color='b')\n",
"ax2.scatter(X,Y_hat_nonlin,color='r')#,ls='none')#linewidth=3) \n",
"ax2.set_xlabel('lstat')\n",
"ax2.set_ylabel('medv')"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Coefficients [-0.95004935]\n",
"Intercept: 34.5538408794\n",
"Variance score: 0.54\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:14: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
" \n",
"/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:15: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
" from ipykernel import kernelapp as app\n"
]
},
{
"data": {
"text/plain": [
"[<matplotlib.lines.Line2D at 0x7f2c6e0b7518>]"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Now lets have a look at scikit learn, another python package for all kinds of statistical learning algorithms\n",
"# In order to directly compare it to statsmodels, I first do again the simple linear regression from Lab 3.6.2\n",
"\n",
"fname = 'Boston.csv'\n",
"df = pd.read_csv(fname)\n",
"\n",
"'''Lab: 3.6.2 Simple Linear Regression (page 110)'''\n",
"\n",
"X = df[\"lstat\"] #independent variable (or predictor)\n",
"y = df[\"medv\"] #dependent variable (which we want to fit / predict)\n",
"\n",
"\n",
"#sklearn does not really handle pandas dataframes (at least it did not work for me) => convert it to numpy array\n",
"X_np = pd.DataFrame.as_matrix(X)\n",
"y_np = pd.DataFrame.as_matrix(y)\n",
"\n",
"#necessary to convert from 1D to 2D array since sklearn needs 2D input for predictor matrix\n",
"X_np2 = X_np[:,np.newaxis]\n",
"\n",
"\n",
"regr = linear_model.LinearRegression()\n",
"regr.fit(X_np2, y_np) #performing actual linear regression\n",
"\n",
"y_hat = regr.predict(X_np2) #predicting values using fit\n",
"\n",
"# we dont get a nice table with sklearn but we can acces the results like this:\n",
"# (note that there is no implementation of STDs, p-values,..., so in my opinion statsmodels is better for linear regressions)\n",
"print('Coefficients', regr.coef_)\n",
"print('Intercept:', regr.intercept_)\n",
"print('Variance score: %.2f' % r2_score(y_np, y_hat))\n",
"\n",
"\n",
"# Plot outputs\n",
"plt.scatter(X_np2, y_np, color='blue')\n",
"plt.plot(X_np2, y_hat, color='red', linewidth=3)\n"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Coefficients [-0.15784473]\n",
"Intercept: 39.9358610212\n",
"Variance score: 0.61\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:20: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n",
"/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:21: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n"
]
},
{
"data": {
"text/plain": [
"Text(0,0.5,'mpg')"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"'''Exercise: 3.7.8 Simple Linear Regression (page 121)'''\n",
"\n",
"\n",
"fname = 'Auto.csv'\n",
"df = pd.read_csv(fname)\n",
"\n",
"#convert strings to float in case of e.g. '?' convert it to NaN\n",
"df['horsepower'] = df['horsepower'].apply(pd.to_numeric, errors='coerce') \n",
"\n",
"\n",
"# Drop NaN values, listing the converted columns explicitly, so NaN values in other columns aren't dropped\n",
"df = df.dropna(how='any',subset = ['horsepower'])\n",
"\n",
"\n",
"X = df[\"horsepower\"] #independent variable (or predictor)\n",
"Y = df[\"mpg\"] #dependent variable (which we want to fit / predict)\n",
"\n",
"\n",
"#sklearn does not really handle pandas dataframes (at leat it did not work for me) => convert to numpy array\n",
"X_np=pd.DataFrame.as_matrix(X)\n",
"Y_np =pd.DataFrame.as_matrix(Y)\n",
"\n",
"#necessary to convert from 1D to 2D array since sklearn needs 2D input for predictor matrix\n",
"X_np2 = X_np[:,np.newaxis]\n",
"\n",
"\n",
"regr = linear_model.LinearRegression()\n",
"regr.fit(X_np2, Y_np) #performing actual linear regression\n",
"\n",
"Y_hat = regr.predict(X_np2) #predicting values using fit\n",
"\n",
"print('Coefficients', regr.coef_)\n",
"print('Intercept:', regr.intercept_)\n",
"print('Variance score: %.2f' % r2_score(Y_np, Y_hat))\n",
"\n",
"# Plot outputs\n",
"plt.scatter(X_np2, Y_np, color='blue')\n",
"plt.plot(X_np2, Y_hat, color='red', linewidth=3) \n",
"plt.xlabel('horespower')\n",
"plt.ylabel('mpg')"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1440x1440 with 64 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"'''Exercise: 3.7.9 Multiple linear Regression (page 122)'''\n",
"\n",
"\n",
"fname = 'Auto.csv'\n",
"df = pd.read_csv(fname)\n",
"\n",
"#plot all binary variable combinations as scatter plots\n",
"f, axarr = plt.subplots(8,8,figsize=(20,20))\n",
"\n",
"col_to_use = ['mpg','cylinders','displacement','horsepower','weight','acceleration','year','origin']\n",
"df_new = df[col_to_use]\n",
"\n",
"c_row = 0\n",
"for row in df_new:\n",
" c_col = 0\n",
" for col in df_new:\n",
" var1 = df_new[col]\n",
" var2 = df_new[row]\n",
" axarr[c_row,c_col].scatter(var1,var2)\n",
" axarr[c_row,c_col].set_xlabel(col)\n",
" axarr[c_row,c_col].set_ylabel(row)\n",
" c_col += 1\n",
" c_row += 1\n",
"f.subplots_adjust(hspace=0.4, wspace=0.4)"
]
},
{
"cell_type": "code",
"execution_count": 14,
"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>mpg</th>\n",
" <th>cylinders</th>\n",
" <th>displacement</th>\n",
" <th>weight</th>\n",
" <th>acceleration</th>\n",
" <th>year</th>\n",
" <th>origin</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>mpg</th>\n",
" <td>1.000000</td>\n",
" <td>-0.776260</td>\n",
" <td>-0.804443</td>\n",
" <td>-0.831739</td>\n",
" <td>0.422297</td>\n",
" <td>0.581469</td>\n",
" <td>0.563698</td>\n",
" </tr>\n",
" <tr>\n",
" <th>cylinders</th>\n",
" <td>-0.776260</td>\n",
" <td>1.000000</td>\n",
" <td>0.950920</td>\n",
" <td>0.897017</td>\n",
" <td>-0.504061</td>\n",
" <td>-0.346717</td>\n",
" <td>-0.564972</td>\n",
" </tr>\n",
" <tr>\n",
" <th>displacement</th>\n",
" <td>-0.804443</td>\n",
" <td>0.950920</td>\n",
" <td>1.000000</td>\n",
" <td>0.933104</td>\n",
" <td>-0.544162</td>\n",
" <td>-0.369804</td>\n",
" <td>-0.610664</td>\n",
" </tr>\n",
" <tr>\n",
" <th>weight</th>\n",
" <td>-0.831739</td>\n",
" <td>0.897017</td>\n",
" <td>0.933104</td>\n",
" <td>1.000000</td>\n",
" <td>-0.419502</td>\n",
" <td>-0.307900</td>\n",
" <td>-0.581265</td>\n",
" </tr>\n",
" <tr>\n",
" <th>acceleration</th>\n",
" <td>0.422297</td>\n",
" <td>-0.504061</td>\n",
" <td>-0.544162</td>\n",
" <td>-0.419502</td>\n",
" <td>1.000000</td>\n",
" <td>0.282901</td>\n",
" <td>0.210084</td>\n",
" </tr>\n",
" <tr>\n",
" <th>year</th>\n",
" <td>0.581469</td>\n",
" <td>-0.346717</td>\n",
" <td>-0.369804</td>\n",
" <td>-0.307900</td>\n",
" <td>0.282901</td>\n",
" <td>1.000000</td>\n",
" <td>0.184314</td>\n",
" </tr>\n",
" <tr>\n",
" <th>origin</th>\n",
" <td>0.563698</td>\n",
" <td>-0.564972</td>\n",
" <td>-0.610664</td>\n",
" <td>-0.581265</td>\n",
" <td>0.210084</td>\n",
" <td>0.184314</td>\n",
" <td>1.000000</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" mpg cylinders displacement weight acceleration \\\n",
"mpg 1.000000 -0.776260 -0.804443 -0.831739 0.422297 \n",
"cylinders -0.776260 1.000000 0.950920 0.897017 -0.504061 \n",
"displacement -0.804443 0.950920 1.000000 0.933104 -0.544162 \n",
"weight -0.831739 0.897017 0.933104 1.000000 -0.419502 \n",
"acceleration 0.422297 -0.504061 -0.544162 -0.419502 1.000000 \n",
"year 0.581469 -0.346717 -0.369804 -0.307900 0.282901 \n",
"origin 0.563698 -0.564972 -0.610664 -0.581265 0.210084 \n",
"\n",
" year origin \n",
"mpg 0.581469 0.563698 \n",
"cylinders -0.346717 -0.564972 \n",
"displacement -0.369804 -0.610664 \n",
"weight -0.307900 -0.581265 \n",
"acceleration 0.282901 0.210084 \n",
"year 1.000000 0.184314 \n",
"origin 0.184314 1.000000 "
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"#correlation matrix to check for possible collinearities\n",
"df.corr()"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Coefficients [-0.49337632 0.01989564 -0.01695114 -0.00647404 0.08057584 0.75077268\n",
" 1.4261405 ]\n",
"Intercept: -17.218434622\n",
"Variance score: 0.82\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:20: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead.\n"
]
}
],
"source": [
"#doing multiple linear regression fit\n",
"\n",
"col_to_use_fit = ['cylinders','displacement','horsepower','weight','acceleration','year','origin']\n",
"\n",
"\n",
"#for using sklearn we have to use floats, however rows containing strings as '?' have to be deleted\n",
"\n",
"#convert strings to float and in case of e.g. '?' convert it to NaN\n",
"df[col_to_use_fit] = df[col_to_use_fit].apply(pd.to_numeric, errors='coerce') \n",
"\n",
"\n",
"# Drop NaN values, listing the converted columns explicitly, so NaN values in other columns aren't dropped\n",
"df = df.dropna(how='any',subset = col_to_use_fit)\n",
"\n",
"X = df[col_to_use_fit] #independent variables (or predictors)\n",
"Y = df[\"mpg\"] #dependent variable (which we want to fit / predict)\n",
"\n",
"\n",
"#sklearn does not really handle pandas dataframes (at least it did not work for me) => convert to numpy array\n",
"X_np=pd.DataFrame.as_matrix(X)\n",
"\n",
"\n",
"regr = linear_model.LinearRegression()\n",
"regr.fit(X_np, Y_np)\n",
"\n",
"Y_hat = regr.predict(X_np) #predicting values using fit\n",
"\n",
"print('Coefficients', regr.coef_)\n",
"print('Intercept:', regr.intercept_)\n",
"print('Variance score: %.2f' % r2_score(Y_np, Y_hat))\n"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Coefficients [ 6.98857616e+00 -4.78538689e-01 5.03433939e-01 4.13288566e-03\n",
" -5.85917321e+00 6.97430284e-01 -2.08955704e+01 -3.38326306e-03\n",
" 1.16133262e-02 3.57462990e-04 2.77871993e-01 -1.74125857e-01\n",
" 4.02168216e-01 -8.49062164e-05 2.47186800e-05 -3.47899523e-03\n",
" 5.93380246e-03 2.39811277e-02 -1.96842975e-05 -7.21273895e-03\n",
" -5.83750571e-03 2.23250718e-03 2.34619450e-04 -2.24523733e-04\n",
" -5.78847492e-04 5.56215079e-02 4.58316099e-01 1.39257020e-01]\n",
"Intercept: 35.4788874758\n",
"Variance score: 0.89\n"
]
}
],
"source": [
"# in order to get all interaction terms, obtain all combinations of predictors (of second order), \n",
"# see https://stackoverflow.com/questions/45828964/how-to-add-interaction-term-in-python-sklearn\n",
"\n",
"poly = PolynomialFeatures(interaction_only=True,include_bias = False)\n",
"X_interact = poly.fit_transform(X_np) #new feature space: [x1,x2,...xN,x1*x2,...x1*XN,x2*x3,...x2*xN,...,...x(N-1)*xN] \n",
"\n",
"regr_interact = linear_model.LinearRegression()\n",
"regr_interact.fit(X_interact, Y_np)\n",
"\n",
"Y_hat_interact = regr_interact.predict(X_interact) #predicting values using fit\n",
"\n",
"print('Coefficients', regr_interact.coef_)\n",
"print('Intercept:', regr_interact.intercept_)\n",
"print('Variance score: %.2f' % r2_score(Y_np, Y_hat_interact))"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>mpg</td> <th> R-squared: </th> <td> 0.821</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.818</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 252.4</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>2.04e-139</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:25:14</td> <th> Log-Likelihood: </th> <td> -1023.5</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 392</td> <th> AIC: </th> <td> 2063.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 384</td> <th> BIC: </th> <td> 2095.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 7</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>const</th> <td> -17.2184</td> <td> 4.644</td> <td> -3.707</td> <td> 0.000</td> <td> -26.350</td> <td> -8.087</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders</th> <td> -0.4934</td> <td> 0.323</td> <td> -1.526</td> <td> 0.128</td> <td> -1.129</td> <td> 0.142</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement</th> <td> 0.0199</td> <td> 0.008</td> <td> 2.647</td> <td> 0.008</td> <td> 0.005</td> <td> 0.035</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower</th> <td> -0.0170</td> <td> 0.014</td> <td> -1.230</td> <td> 0.220</td> <td> -0.044</td> <td> 0.010</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight</th> <td> -0.0065</td> <td> 0.001</td> <td> -9.929</td> <td> 0.000</td> <td> -0.008</td> <td> -0.005</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration</th> <td> 0.0806</td> <td> 0.099</td> <td> 0.815</td> <td> 0.415</td> <td> -0.114</td> <td> 0.275</td>\n",
"</tr>\n",
"<tr>\n",
" <th>year</th> <td> 0.7508</td> <td> 0.051</td> <td> 14.729</td> <td> 0.000</td> <td> 0.651</td> <td> 0.851</td>\n",
"</tr>\n",
"<tr>\n",
" <th>origin</th> <td> 1.4261</td> <td> 0.278</td> <td> 5.127</td> <td> 0.000</td> <td> 0.879</td> <td> 1.973</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>31.906</td> <th> Durbin-Watson: </th> <td> 1.309</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 53.100</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 0.529</td> <th> Prob(JB): </th> <td>2.95e-12</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 4.460</td> <th> Cond. No. </th> <td>8.59e+04</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 8.59e+04. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: mpg R-squared: 0.821\n",
"Model: OLS Adj. R-squared: 0.818\n",
"Method: Least Squares F-statistic: 252.4\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 2.04e-139\n",
"Time: 08:25:14 Log-Likelihood: -1023.5\n",
"No. Observations: 392 AIC: 2063.\n",
"Df Residuals: 384 BIC: 2095.\n",
"Df Model: 7 \n",
"Covariance Type: nonrobust \n",
"================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"--------------------------------------------------------------------------------\n",
"const -17.2184 4.644 -3.707 0.000 -26.350 -8.087\n",
"cylinders -0.4934 0.323 -1.526 0.128 -1.129 0.142\n",
"displacement 0.0199 0.008 2.647 0.008 0.005 0.035\n",
"horsepower -0.0170 0.014 -1.230 0.220 -0.044 0.010\n",
"weight -0.0065 0.001 -9.929 0.000 -0.008 -0.005\n",
"acceleration 0.0806 0.099 0.815 0.415 -0.114 0.275\n",
"year 0.7508 0.051 14.729 0.000 0.651 0.851\n",
"origin 1.4261 0.278 5.127 0.000 0.879 1.973\n",
"==============================================================================\n",
"Omnibus: 31.906 Durbin-Watson: 1.309\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 53.100\n",
"Skew: 0.529 Prob(JB): 2.95e-12\n",
"Kurtosis: 4.460 Cond. No. 8.59e+04\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 8.59e+04. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# since it is very difficult to interpret the above results I get back to using statsmodels with its nice output table\n",
"''' 3.7.9c: Mulitiple linear regression'''\n",
"\n",
"\n",
"fname = 'Auto.csv'\n",
"df = pd.read_csv(fname)\n",
"\n",
"col_to_use_fit = ['cylinders','displacement','horsepower','weight','acceleration','year','origin']\n",
"\n",
"# again rows containing strings as '?' have to be deleted\n",
"# convert strings to float and in case of e.g. '?' convert it to NaN\n",
"df[col_to_use_fit] = df[col_to_use_fit].apply(pd.to_numeric, errors='coerce') \n",
"\n",
"# Drop NaN values, listing the converted columns explicitly, so NaN values in other columns aren't dropped\n",
"df = df.dropna(how='any',subset = col_to_use_fit)\n",
"\n",
"\n",
"X = df[col_to_use_fit]\n",
"y = df[\"mpg\"]\n",
"X0 = sm.add_constant(X) # we have to manually specify that we want an intercept (beta_0) in our model\n",
"model = sm.OLS(y, X0).fit() # performing the linear regression\n",
"model.summary()\n"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 1440x432 with 14 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"''' 3.7.9d: Plotting'''\n",
"\n",
"fname = 'Auto.csv'\n",
"df = pd.read_csv(fname)\n",
"y = df[\"mpg\"]\n",
"\n",
"col_to_use_fit = ['cylinders','displacement','horsepower','weight','acceleration','year','origin']\n",
"\n",
"\n",
"\n",
"f, axarr = plt.subplots(2,len(col_to_use_fit),figsize=(20,6))\n",
"\n",
"for ind, param in enumerate(col_to_use_fit):\n",
" #print(ind,param)\n",
" \n",
" # convert strings to float and in case of e.g. '?' convert it to NaN\n",
" df[[param,'mpg']] = df[[param,'mpg']].apply(pd.to_numeric, errors='coerce') \n",
"\n",
" # Drop NaN values, listing the converted columns explicitly, so NaN values in other columns aren't dropped\n",
" df_used = df.dropna(how='any',subset = [param,'mpg'])\n",
" \n",
" y = df_used[\"mpg\"]\n",
"\n",
" \n",
" X = np.asarray(df_used[param])\n",
" X0 = sm.add_constant(X) # we have to manually specify that we want an intercept (beta_0) in our model\n",
" model = sm.OLS(y, X0).fit() # performing the linear regression\n",
" y_hat = model.fittedvalues\n",
" resids = model.resid\n",
"\n",
" axarr[0,ind].scatter(df_used[param],df_used['mpg'])\n",
" axarr[0,ind].plot(df_used[param],y_hat,color='red') \n",
" axarr[0,ind].set_xlabel(param)\n",
" axarr[0,ind].set_ylabel('mpg')\n",
" \n",
" #plot residuals\n",
" axarr[1,ind].scatter(y,resids, color='blue')\n",
" axarr[1,ind].set_xlabel('mpg')\n",
" axarr[1,ind].set_ylabel('residuals')\n",
" \n",
"f.subplots_adjust(hspace=0.4, wspace=0.4)\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>mpg</td> <th> R-squared: </th> <td> 0.931</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.906</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 38.51</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>1.50e-123</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:25:16</td> <th> Log-Likelihood: </th> <td> -838.24</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 392</td> <th> AIC: </th> <td> 1880.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 290</td> <th> BIC: </th> <td> 2286.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 101</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> -0.0172</td> <td> 0.395</td> <td> -0.043</td> <td> 0.965</td> <td> -0.795</td> <td> 0.761</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders</th> <td> -0.9694</td> <td> 3.083</td> <td> -0.314</td> <td> 0.753</td> <td> -7.037</td> <td> 5.098</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement</th> <td> 15.5886</td> <td> 26.002</td> <td> 0.600</td> <td> 0.549</td> <td> -35.589</td> <td> 66.766</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement</th> <td> 9.9093</td> <td> 7.405</td> <td> 1.338</td> <td> 0.182</td> <td> -4.664</td> <td> 24.483</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower</th> <td> 3.5012</td> <td> 23.820</td> <td> 0.147</td> <td> 0.883</td> <td> -43.381</td> <td> 50.383</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower</th> <td> 0.3706</td> <td> 8.546</td> <td> 0.043</td> <td> 0.965</td> <td> -16.450</td> <td> 17.191</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower</th> <td> -0.2285</td> <td> 0.316</td> <td> -0.722</td> <td> 0.471</td> <td> -0.851</td> <td> 0.394</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower</th> <td> -0.1061</td> <td> 0.094</td> <td> -1.128</td> <td> 0.260</td> <td> -0.291</td> <td> 0.079</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight</th> <td> -0.5497</td> <td> 0.851</td> <td> -0.646</td> <td> 0.519</td> <td> -2.224</td> <td> 1.124</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight</th> <td> -0.3716</td> <td> 0.284</td> <td> -1.309</td> <td> 0.192</td> <td> -0.931</td> <td> 0.187</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight</th> <td> -0.0100</td> <td> 0.022</td> <td> -0.455</td> <td> 0.649</td> <td> -0.053</td> <td> 0.033</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight</th> <td> 0.0016</td> <td> 0.004</td> <td> 0.389</td> <td> 0.698</td> <td> -0.007</td> <td> 0.010</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight</th> <td> 0.0024</td> <td> 0.011</td> <td> 0.224</td> <td> 0.823</td> <td> -0.019</td> <td> 0.024</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight</th> <td> 0.0040</td> <td> 0.004</td> <td> 1.084</td> <td> 0.279</td> <td> -0.003</td> <td> 0.011</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight</th> <td>-1.791e-05</td> <td> 0.000</td> <td> -0.082</td> <td> 0.935</td> <td> -0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight</th> <td> 2.426e-05</td> <td> 4.09e-05</td> <td> 0.593</td> <td> 0.553</td> <td>-5.62e-05</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration</th> <td> -24.8662</td> <td> 46.097</td> <td> -0.539</td> <td> 0.590</td> <td> -115.593</td> <td> 65.860</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:acceleration</th> <td> 2.7613</td> <td> 24.785</td> <td> 0.111</td> <td> 0.911</td> <td> -46.019</td> <td> 51.542</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:acceleration</th> <td> -1.0363</td> <td> 1.455</td> <td> -0.712</td> <td> 0.477</td> <td> -3.899</td> <td> 1.827</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:acceleration</th> <td> -0.7075</td> <td> 0.511</td> <td> -1.385</td> <td> 0.167</td> <td> -1.713</td> <td> 0.298</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:acceleration</th> <td> 0.0148</td> <td> 1.697</td> <td> 0.009</td> <td> 0.993</td> <td> -3.325</td> <td> 3.355</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:acceleration</th> <td> -0.1210</td> <td> 0.636</td> <td> -0.190</td> <td> 0.849</td> <td> -1.373</td> <td> 1.131</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:acceleration</th> <td> 0.0271</td> <td> 0.024</td> <td> 1.153</td> <td> 0.250</td> <td> -0.019</td> <td> 0.073</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:acceleration</th> <td> 0.0049</td> <td> 0.006</td> <td> 0.767</td> <td> 0.444</td> <td> -0.008</td> <td> 0.017</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:acceleration</th> <td> 0.0331</td> <td> 0.062</td> <td> 0.534</td> <td> 0.594</td> <td> -0.089</td> <td> 0.155</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:acceleration</th> <td> 0.0279</td> <td> 0.018</td> <td> 1.549</td> <td> 0.123</td> <td> -0.008</td> <td> 0.063</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:acceleration</th> <td> -0.0004</td> <td> 0.001</td> <td> -0.275</td> <td> 0.784</td> <td> -0.003</td> <td> 0.002</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:acceleration</th> <td> 0.0002</td> <td> 0.000</td> <td> 0.758</td> <td> 0.449</td> <td> -0.000</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:acceleration</th> <td> 7.027e-05</td> <td> 0.001</td> <td> 0.082</td> <td> 0.935</td> <td> -0.002</td> <td> 0.002</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:acceleration</th> <td> -0.0003</td> <td> 0.000</td> <td> -1.293</td> <td> 0.197</td> <td> -0.001</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:acceleration</th> <td> 5.366e-06</td> <td> 1.68e-05</td> <td> 0.319</td> <td> 0.750</td> <td>-2.77e-05</td> <td> 3.85e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:acceleration</th> <td>-2.931e-06</td> <td> 3.04e-06</td> <td> -0.965</td> <td> 0.335</td> <td>-8.91e-06</td> <td> 3.05e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>year</th> <td> -7.7724</td> <td> 5.452</td> <td> -1.426</td> <td> 0.155</td> <td> -18.503</td> <td> 2.958</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:year</th> <td> -2.2048</td> <td> 1.821</td> <td> -1.211</td> <td> 0.227</td> <td> -5.789</td> <td> 1.380</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:year</th> <td> -0.1627</td> <td> 0.344</td> <td> -0.473</td> <td> 0.636</td> <td> -0.839</td> <td> 0.514</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:year</th> <td> -0.1109</td> <td> 0.097</td> <td> -1.141</td> <td> 0.255</td> <td> -0.302</td> <td> 0.080</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:year</th> <td> 0.0345</td> <td> 0.322</td> <td> 0.107</td> <td> 0.915</td> <td> -0.600</td> <td> 0.669</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:year</th> <td> 0.0331</td> <td> 0.121</td> <td> 0.275</td> <td> 0.784</td> <td> -0.204</td> <td> 0.270</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:year</th> <td> 0.0018</td> <td> 0.004</td> <td> 0.411</td> <td> 0.681</td> <td> -0.007</td> <td> 0.010</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:year</th> <td> 0.0013</td> <td> 0.001</td> <td> 1.038</td> <td> 0.300</td> <td> -0.001</td> <td> 0.004</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:year</th> <td> 0.0100</td> <td> 0.015</td> <td> 0.674</td> <td> 0.501</td> <td> -0.019</td> <td> 0.039</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:year</th> <td> 0.0061</td> <td> 0.004</td> <td> 1.670</td> <td> 0.096</td> <td> -0.001</td> <td> 0.013</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:year</th> <td> 7.834e-05</td> <td> 0.000</td> <td> 0.297</td> <td> 0.767</td> <td> -0.000</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:year</th> <td>-2.064e-05</td> <td> 4.85e-05</td> <td> -0.425</td> <td> 0.671</td> <td> -0.000</td> <td> 7.48e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:year</th> <td>-2.282e-05</td> <td> 0.000</td> <td> -0.142</td> <td> 0.887</td> <td> -0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:year</th> <td>-8.244e-05</td> <td> 4.81e-05</td> <td> -1.716</td> <td> 0.087</td> <td> -0.000</td> <td> 1.21e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:year</th> <td> 7.251e-07</td> <td> 2.64e-06</td> <td> 0.274</td> <td> 0.784</td> <td>-4.48e-06</td> <td> 5.93e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:year</th> <td>-2.607e-07</td> <td> 4.95e-07</td> <td> -0.527</td> <td> 0.599</td> <td>-1.23e-06</td> <td> 7.13e-07</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration:year</th> <td> 0.8053</td> <td> 0.649</td> <td> 1.240</td> <td> 0.216</td> <td> -0.473</td> <td> 2.083</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:acceleration:year</th> <td> 0.1014</td> <td> 0.330</td> <td> 0.308</td> <td> 0.759</td> <td> -0.548</td> <td> 0.750</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:acceleration:year</th> <td> 0.0271</td> <td> 0.021</td> <td> 1.285</td> <td> 0.200</td> <td> -0.014</td> <td> 0.069</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:acceleration:year</th> <td> 0.0042</td> <td> 0.007</td> <td> 0.628</td> <td> 0.531</td> <td> -0.009</td> <td> 0.017</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:acceleration:year</th> <td> -0.0057</td> <td> 0.024</td> <td> -0.236</td> <td> 0.813</td> <td> -0.053</td> <td> 0.041</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:acceleration:year</th> <td> -0.0005</td> <td> 0.009</td> <td> -0.055</td> <td> 0.956</td> <td> -0.018</td> <td> 0.017</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:acceleration:year</th> <td> -0.0004</td> <td> 0.000</td> <td> -1.253</td> <td> 0.211</td> <td> -0.001</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:acceleration:year</th> <td>-2.812e-05</td> <td> 8.86e-05</td> <td> -0.318</td> <td> 0.751</td> <td> -0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:acceleration:year</th> <td> -0.0008</td> <td> 0.001</td> <td> -0.799</td> <td> 0.425</td> <td> -0.003</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:acceleration:year</th> <td> -0.0004</td> <td> 0.000</td> <td> -1.703</td> <td> 0.090</td> <td> -0.001</td> <td> 6.23e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:acceleration:year</th> <td> 3.71e-06</td> <td> 1.6e-05</td> <td> 0.231</td> <td> 0.817</td> <td>-2.78e-05</td> <td> 3.53e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:acceleration:year</th> <td>-1.154e-06</td> <td> 2.54e-06</td> <td> -0.455</td> <td> 0.650</td> <td>-6.15e-06</td> <td> 3.84e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:acceleration:year</th> <td> 1.038e-07</td> <td> 1.19e-05</td> <td> 0.009</td> <td> 0.993</td> <td>-2.33e-05</td> <td> 2.35e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:acceleration:year</th> <td> 5.657e-06</td> <td> 3.23e-06</td> <td> 1.751</td> <td> 0.081</td> <td>-7.03e-07</td> <td> 1.2e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:acceleration:year</th> <td>-6.701e-08</td> <td> 2.04e-07</td> <td> -0.328</td> <td> 0.743</td> <td>-4.69e-07</td> <td> 3.35e-07</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:acceleration:year</th> <td> 2.688e-08</td> <td> 3.69e-08</td> <td> 0.728</td> <td> 0.467</td> <td>-4.58e-08</td> <td> 9.96e-08</td>\n",
"</tr>\n",
"<tr>\n",
" <th>origin</th> <td> -1.6045</td> <td> 3.093</td> <td> -0.519</td> <td> 0.604</td> <td> -7.692</td> <td> 4.483</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:origin</th> <td> -0.7675</td> <td> 2.486</td> <td> -0.309</td> <td> 0.758</td> <td> -5.660</td> <td> 4.125</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:origin</th> <td> 10.2514</td> <td> 21.847</td> <td> 0.469</td> <td> 0.639</td> <td> -32.748</td> <td> 53.251</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:origin</th> <td> -11.6479</td> <td> 12.279</td> <td> -0.949</td> <td> 0.344</td> <td> -35.815</td> <td> 12.520</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:origin</th> <td> 3.0028</td> <td> 21.382</td> <td> 0.140</td> <td> 0.888</td> <td> -39.082</td> <td> 45.087</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:origin</th> <td> -1.6331</td> <td> 11.009</td> <td> -0.148</td> <td> 0.882</td> <td> -23.300</td> <td> 20.034</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:origin</th> <td> -0.1664</td> <td> 0.266</td> <td> -0.626</td> <td> 0.532</td> <td> -0.690</td> <td> 0.357</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:origin</th> <td> 0.1477</td> <td> 0.151</td> <td> 0.975</td> <td> 0.330</td> <td> -0.150</td> <td> 0.446</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:origin</th> <td> -0.3467</td> <td> 0.727</td> <td> -0.477</td> <td> 0.634</td> <td> -1.778</td> <td> 1.085</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:origin</th> <td> 0.4399</td> <td> 0.414</td> <td> 1.063</td> <td> 0.289</td> <td> -0.374</td> <td> 1.254</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:origin</th> <td> 0.0092</td> <td> 0.020</td> <td> 0.461</td> <td> 0.645</td> <td> -0.030</td> <td> 0.048</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:origin</th> <td> -0.0017</td> <td> 0.005</td> <td> -0.367</td> <td> 0.714</td> <td> -0.011</td> <td> 0.008</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:origin</th> <td> 0.0054</td> <td> 0.013</td> <td> 0.428</td> <td> 0.669</td> <td> -0.019</td> <td> 0.030</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:origin</th> <td> -0.0046</td> <td> 0.005</td> <td> -0.830</td> <td> 0.407</td> <td> -0.015</td> <td> 0.006</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:origin</th> <td> 6.377e-05</td> <td> 0.000</td> <td> 0.321</td> <td> 0.748</td> <td> -0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:origin</th> <td>-2.917e-05</td> <td> 4.8e-05</td> <td> -0.607</td> <td> 0.544</td> <td> -0.000</td> <td> 6.53e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration:origin</th> <td> -23.8434</td> <td> 45.005</td> <td> -0.530</td> <td> 0.597</td> <td> -112.422</td> <td> 64.735</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:acceleration:origin</th> <td> 6.8524</td> <td> 20.644</td> <td> 0.332</td> <td> 0.740</td> <td> -33.778</td> <td> 47.483</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:acceleration:origin</th> <td> -0.6654</td> <td> 1.190</td> <td> -0.559</td> <td> 0.576</td> <td> -3.007</td> <td> 1.676</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:acceleration:origin</th> <td> 0.7784</td> <td> 0.745</td> <td> 1.045</td> <td> 0.297</td> <td> -0.688</td> <td> 2.244</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:acceleration:origin</th> <td> 0.0619</td> <td> 1.563</td> <td> 0.040</td> <td> 0.968</td> <td> -3.015</td> <td> 3.139</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:acceleration:origin</th> <td> 0.0675</td> <td> 0.737</td> <td> 0.092</td> <td> 0.927</td> <td> -1.384</td> <td> 1.519</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:acceleration:origin</th> <td> 0.0011</td> <td> 0.017</td> <td> 0.063</td> <td> 0.950</td> <td> -0.033</td> <td> 0.035</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:acceleration:origin</th> <td> -0.0075</td> <td> 0.009</td> <td> -0.818</td> <td> 0.414</td> <td> -0.026</td> <td> 0.011</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:acceleration:origin</th> <td> 0.0212</td> <td> 0.055</td> <td> 0.385</td> <td> 0.701</td> <td> -0.087</td> <td> 0.129</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:acceleration:origin</th> <td> -0.0292</td> <td> 0.028</td> <td> -1.033</td> <td> 0.303</td> <td> -0.085</td> <td> 0.026</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:acceleration:origin</th> <td> 0.0005</td> <td> 0.001</td> <td> 0.395</td> <td> 0.693</td> <td> -0.002</td> <td> 0.003</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:acceleration:origin</th> <td> -0.0002</td> <td> 0.000</td> <td> -0.661</td> <td> 0.509</td> <td> -0.001</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:acceleration:origin</th> <td> -0.0006</td> <td> 0.001</td> <td> -0.636</td> <td> 0.526</td> <td> -0.002</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:acceleration:origin</th> <td> 0.0004</td> <td> 0.000</td> <td> 0.929</td> <td> 0.354</td> <td> -0.000</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:acceleration:origin</th> <td>-9.366e-06</td> <td> 1.57e-05</td> <td> -0.596</td> <td> 0.552</td> <td>-4.03e-05</td> <td> 2.16e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:acceleration:origin</th> <td> 3.338e-06</td> <td> 3.28e-06</td> <td> 1.018</td> <td> 0.309</td> <td>-3.11e-06</td> <td> 9.79e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>year:origin</th> <td> -6.1216</td> <td> 4.912</td> <td> -1.246</td> <td> 0.214</td> <td> -15.790</td> <td> 3.547</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:year:origin</th> <td> 4.3985</td> <td> 2.511</td> <td> 1.752</td> <td> 0.081</td> <td> -0.543</td> <td> 9.340</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:year:origin</th> <td> -0.1074</td> <td> 0.288</td> <td> -0.373</td> <td> 0.710</td> <td> -0.675</td> <td> 0.460</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:year:origin</th> <td> 0.1231</td> <td> 0.162</td> <td> 0.761</td> <td> 0.447</td> <td> -0.195</td> <td> 0.441</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:year:origin</th> <td> 0.0187</td> <td> 0.289</td> <td> 0.065</td> <td> 0.948</td> <td> -0.551</td> <td> 0.588</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:year:origin</th> <td> -0.0343</td> <td> 0.153</td> <td> -0.224</td> <td> 0.823</td> <td> -0.335</td> <td> 0.267</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:year:origin</th> <td> 0.0028</td> <td> 0.004</td> <td> 0.793</td> <td> 0.429</td> <td> -0.004</td> <td> 0.010</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:year:origin</th> <td> -0.0018</td> <td> 0.002</td> <td> -0.879</td> <td> 0.380</td> <td> -0.006</td> <td> 0.002</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:year:origin</th> <td> 0.0076</td> <td> 0.011</td> <td> 0.698</td> <td> 0.485</td> <td> -0.014</td> <td> 0.029</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:year:origin</th> <td> -0.0079</td> <td> 0.005</td> <td> -1.559</td> <td> 0.120</td> <td> -0.018</td> <td> 0.002</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:year:origin</th> <td>-9.258e-05</td> <td> 0.000</td> <td> -0.398</td> <td> 0.691</td> <td> -0.001</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:year:origin</th> <td> 2.574e-05</td> <td> 5.68e-05</td> <td> 0.454</td> <td> 0.650</td> <td> -8.6e-05</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:year:origin</th> <td> -0.0001</td> <td> 0.000</td> <td> -0.995</td> <td> 0.320</td> <td> -0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:year:origin</th> <td> 9.734e-05</td> <td> 6.81e-05</td> <td> 1.429</td> <td> 0.154</td> <td>-3.68e-05</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:year:origin</th> <td>-1.062e-06</td> <td> 2.32e-06</td> <td> -0.457</td> <td> 0.648</td> <td>-5.64e-06</td> <td> 3.51e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:year:origin</th> <td> 2.907e-07</td> <td> 5.94e-07</td> <td> 0.490</td> <td> 0.625</td> <td>-8.78e-07</td> <td> 1.46e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration:year:origin</th> <td> 0.6903</td> <td> 0.627</td> <td> 1.101</td> <td> 0.272</td> <td> -0.544</td> <td> 1.925</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:acceleration:year:origin</th> <td> -0.3571</td> <td> 0.283</td> <td> -1.263</td> <td> 0.208</td> <td> -0.914</td> <td> 0.200</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:acceleration:year:origin</th> <td> -0.0086</td> <td> 0.018</td> <td> -0.493</td> <td> 0.622</td> <td> -0.043</td> <td> 0.026</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:acceleration:year:origin</th> <td> -0.0045</td> <td> 0.010</td> <td> -0.456</td> <td> 0.648</td> <td> -0.024</td> <td> 0.015</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:acceleration:year:origin</th> <td> -0.0031</td> <td> 0.020</td> <td> -0.151</td> <td> 0.880</td> <td> -0.043</td> <td> 0.037</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:acceleration:year:origin</th> <td> 0.0022</td> <td> 0.010</td> <td> 0.218</td> <td> 0.827</td> <td> -0.018</td> <td> 0.022</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:acceleration:year:origin</th> <td> 6.583e-05</td> <td> 0.000</td> <td> 0.275</td> <td> 0.783</td> <td> -0.000</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:acceleration:year:origin</th> <td> 5.775e-05</td> <td> 0.000</td> <td> 0.464</td> <td> 0.643</td> <td> -0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight:acceleration:year:origin</th> <td> -0.0003</td> <td> 0.001</td> <td> -0.400</td> <td> 0.689</td> <td> -0.002</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:weight:acceleration:year:origin</th> <td> 0.0005</td> <td> 0.000</td> <td> 1.353</td> <td> 0.177</td> <td> -0.000</td> <td> 0.001</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:weight:acceleration:year:origin</th> <td> -3.57e-06</td> <td> 1.4e-05</td> <td> -0.254</td> <td> 0.799</td> <td>-3.12e-05</td> <td> 2.41e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:weight:acceleration:year:origin</th> <td> 9.149e-07</td> <td> 3e-06</td> <td> 0.305</td> <td> 0.760</td> <td>-4.98e-06</td> <td> 6.81e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower:weight:acceleration:year:origin</th> <td> 1.061e-05</td> <td> 1.08e-05</td> <td> 0.979</td> <td> 0.328</td> <td>-1.07e-05</td> <td> 3.2e-05</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:horsepower:weight:acceleration:year:origin</th> <td> -6.54e-06</td> <td> 4.75e-06</td> <td> -1.378</td> <td> 0.169</td> <td>-1.59e-05</td> <td> 2.8e-06</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement:horsepower:weight:acceleration:year:origin</th> <td> 1.05e-07</td> <td> 1.88e-07</td> <td> 0.559</td> <td> 0.577</td> <td>-2.65e-07</td> <td> 4.75e-07</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement:horsepower:weight:acceleration:year:origin</th> <td>-3.033e-08</td> <td> 4.04e-08</td> <td> -0.751</td> <td> 0.454</td> <td> -1.1e-07</td> <td> 4.92e-08</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>46.080</td> <th> Durbin-Watson: </th> <td> 1.751</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 130.524</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 0.539</td> <th> Prob(JB): </th> <td>4.54e-29</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.613</td> <th> Cond. No. </th> <td>3.52e+16</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 3.52e+16. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: mpg R-squared: 0.931\n",
"Model: OLS Adj. R-squared: 0.906\n",
"Method: Least Squares F-statistic: 38.51\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 1.50e-123\n",
"Time: 08:25:16 Log-Likelihood: -838.24\n",
"No. Observations: 392 AIC: 1880.\n",
"Df Residuals: 290 BIC: 2286.\n",
"Df Model: 101 \n",
"Covariance Type: nonrobust \n",
"=====================================================================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"-------------------------------------------------------------------------------------------------------------------------------------\n",
"Intercept -0.0172 0.395 -0.043 0.965 -0.795 0.761\n",
"cylinders -0.9694 3.083 -0.314 0.753 -7.037 5.098\n",
"displacement 15.5886 26.002 0.600 0.549 -35.589 66.766\n",
"cylinders:displacement 9.9093 7.405 1.338 0.182 -4.664 24.483\n",
"horsepower 3.5012 23.820 0.147 0.883 -43.381 50.383\n",
"cylinders:horsepower 0.3706 8.546 0.043 0.965 -16.450 17.191\n",
"displacement:horsepower -0.2285 0.316 -0.722 0.471 -0.851 0.394\n",
"cylinders:displacement:horsepower -0.1061 0.094 -1.128 0.260 -0.291 0.079\n",
"weight -0.5497 0.851 -0.646 0.519 -2.224 1.124\n",
"cylinders:weight -0.3716 0.284 -1.309 0.192 -0.931 0.187\n",
"displacement:weight -0.0100 0.022 -0.455 0.649 -0.053 0.033\n",
"cylinders:displacement:weight 0.0016 0.004 0.389 0.698 -0.007 0.010\n",
"horsepower:weight 0.0024 0.011 0.224 0.823 -0.019 0.024\n",
"cylinders:horsepower:weight 0.0040 0.004 1.084 0.279 -0.003 0.011\n",
"displacement:horsepower:weight -1.791e-05 0.000 -0.082 0.935 -0.000 0.000\n",
"cylinders:displacement:horsepower:weight 2.426e-05 4.09e-05 0.593 0.553 -5.62e-05 0.000\n",
"acceleration -24.8662 46.097 -0.539 0.590 -115.593 65.860\n",
"cylinders:acceleration 2.7613 24.785 0.111 0.911 -46.019 51.542\n",
"displacement:acceleration -1.0363 1.455 -0.712 0.477 -3.899 1.827\n",
"cylinders:displacement:acceleration -0.7075 0.511 -1.385 0.167 -1.713 0.298\n",
"horsepower:acceleration 0.0148 1.697 0.009 0.993 -3.325 3.355\n",
"cylinders:horsepower:acceleration -0.1210 0.636 -0.190 0.849 -1.373 1.131\n",
"displacement:horsepower:acceleration 0.0271 0.024 1.153 0.250 -0.019 0.073\n",
"cylinders:displacement:horsepower:acceleration 0.0049 0.006 0.767 0.444 -0.008 0.017\n",
"weight:acceleration 0.0331 0.062 0.534 0.594 -0.089 0.155\n",
"cylinders:weight:acceleration 0.0279 0.018 1.549 0.123 -0.008 0.063\n",
"displacement:weight:acceleration -0.0004 0.001 -0.275 0.784 -0.003 0.002\n",
"cylinders:displacement:weight:acceleration 0.0002 0.000 0.758 0.449 -0.000 0.001\n",
"horsepower:weight:acceleration 7.027e-05 0.001 0.082 0.935 -0.002 0.002\n",
"cylinders:horsepower:weight:acceleration -0.0003 0.000 -1.293 0.197 -0.001 0.000\n",
"displacement:horsepower:weight:acceleration 5.366e-06 1.68e-05 0.319 0.750 -2.77e-05 3.85e-05\n",
"cylinders:displacement:horsepower:weight:acceleration -2.931e-06 3.04e-06 -0.965 0.335 -8.91e-06 3.05e-06\n",
"year -7.7724 5.452 -1.426 0.155 -18.503 2.958\n",
"cylinders:year -2.2048 1.821 -1.211 0.227 -5.789 1.380\n",
"displacement:year -0.1627 0.344 -0.473 0.636 -0.839 0.514\n",
"cylinders:displacement:year -0.1109 0.097 -1.141 0.255 -0.302 0.080\n",
"horsepower:year 0.0345 0.322 0.107 0.915 -0.600 0.669\n",
"cylinders:horsepower:year 0.0331 0.121 0.275 0.784 -0.204 0.270\n",
"displacement:horsepower:year 0.0018 0.004 0.411 0.681 -0.007 0.010\n",
"cylinders:displacement:horsepower:year 0.0013 0.001 1.038 0.300 -0.001 0.004\n",
"weight:year 0.0100 0.015 0.674 0.501 -0.019 0.039\n",
"cylinders:weight:year 0.0061 0.004 1.670 0.096 -0.001 0.013\n",
"displacement:weight:year 7.834e-05 0.000 0.297 0.767 -0.000 0.001\n",
"cylinders:displacement:weight:year -2.064e-05 4.85e-05 -0.425 0.671 -0.000 7.48e-05\n",
"horsepower:weight:year -2.282e-05 0.000 -0.142 0.887 -0.000 0.000\n",
"cylinders:horsepower:weight:year -8.244e-05 4.81e-05 -1.716 0.087 -0.000 1.21e-05\n",
"displacement:horsepower:weight:year 7.251e-07 2.64e-06 0.274 0.784 -4.48e-06 5.93e-06\n",
"cylinders:displacement:horsepower:weight:year -2.607e-07 4.95e-07 -0.527 0.599 -1.23e-06 7.13e-07\n",
"acceleration:year 0.8053 0.649 1.240 0.216 -0.473 2.083\n",
"cylinders:acceleration:year 0.1014 0.330 0.308 0.759 -0.548 0.750\n",
"displacement:acceleration:year 0.0271 0.021 1.285 0.200 -0.014 0.069\n",
"cylinders:displacement:acceleration:year 0.0042 0.007 0.628 0.531 -0.009 0.017\n",
"horsepower:acceleration:year -0.0057 0.024 -0.236 0.813 -0.053 0.041\n",
"cylinders:horsepower:acceleration:year -0.0005 0.009 -0.055 0.956 -0.018 0.017\n",
"displacement:horsepower:acceleration:year -0.0004 0.000 -1.253 0.211 -0.001 0.000\n",
"cylinders:displacement:horsepower:acceleration:year -2.812e-05 8.86e-05 -0.318 0.751 -0.000 0.000\n",
"weight:acceleration:year -0.0008 0.001 -0.799 0.425 -0.003 0.001\n",
"cylinders:weight:acceleration:year -0.0004 0.000 -1.703 0.090 -0.001 6.23e-05\n",
"displacement:weight:acceleration:year 3.71e-06 1.6e-05 0.231 0.817 -2.78e-05 3.53e-05\n",
"cylinders:displacement:weight:acceleration:year -1.154e-06 2.54e-06 -0.455 0.650 -6.15e-06 3.84e-06\n",
"horsepower:weight:acceleration:year 1.038e-07 1.19e-05 0.009 0.993 -2.33e-05 2.35e-05\n",
"cylinders:horsepower:weight:acceleration:year 5.657e-06 3.23e-06 1.751 0.081 -7.03e-07 1.2e-05\n",
"displacement:horsepower:weight:acceleration:year -6.701e-08 2.04e-07 -0.328 0.743 -4.69e-07 3.35e-07\n",
"cylinders:displacement:horsepower:weight:acceleration:year 2.688e-08 3.69e-08 0.728 0.467 -4.58e-08 9.96e-08\n",
"origin -1.6045 3.093 -0.519 0.604 -7.692 4.483\n",
"cylinders:origin -0.7675 2.486 -0.309 0.758 -5.660 4.125\n",
"displacement:origin 10.2514 21.847 0.469 0.639 -32.748 53.251\n",
"cylinders:displacement:origin -11.6479 12.279 -0.949 0.344 -35.815 12.520\n",
"horsepower:origin 3.0028 21.382 0.140 0.888 -39.082 45.087\n",
"cylinders:horsepower:origin -1.6331 11.009 -0.148 0.882 -23.300 20.034\n",
"displacement:horsepower:origin -0.1664 0.266 -0.626 0.532 -0.690 0.357\n",
"cylinders:displacement:horsepower:origin 0.1477 0.151 0.975 0.330 -0.150 0.446\n",
"weight:origin -0.3467 0.727 -0.477 0.634 -1.778 1.085\n",
"cylinders:weight:origin 0.4399 0.414 1.063 0.289 -0.374 1.254\n",
"displacement:weight:origin 0.0092 0.020 0.461 0.645 -0.030 0.048\n",
"cylinders:displacement:weight:origin -0.0017 0.005 -0.367 0.714 -0.011 0.008\n",
"horsepower:weight:origin 0.0054 0.013 0.428 0.669 -0.019 0.030\n",
"cylinders:horsepower:weight:origin -0.0046 0.005 -0.830 0.407 -0.015 0.006\n",
"displacement:horsepower:weight:origin 6.377e-05 0.000 0.321 0.748 -0.000 0.000\n",
"cylinders:displacement:horsepower:weight:origin -2.917e-05 4.8e-05 -0.607 0.544 -0.000 6.53e-05\n",
"acceleration:origin -23.8434 45.005 -0.530 0.597 -112.422 64.735\n",
"cylinders:acceleration:origin 6.8524 20.644 0.332 0.740 -33.778 47.483\n",
"displacement:acceleration:origin -0.6654 1.190 -0.559 0.576 -3.007 1.676\n",
"cylinders:displacement:acceleration:origin 0.7784 0.745 1.045 0.297 -0.688 2.244\n",
"horsepower:acceleration:origin 0.0619 1.563 0.040 0.968 -3.015 3.139\n",
"cylinders:horsepower:acceleration:origin 0.0675 0.737 0.092 0.927 -1.384 1.519\n",
"displacement:horsepower:acceleration:origin 0.0011 0.017 0.063 0.950 -0.033 0.035\n",
"cylinders:displacement:horsepower:acceleration:origin -0.0075 0.009 -0.818 0.414 -0.026 0.011\n",
"weight:acceleration:origin 0.0212 0.055 0.385 0.701 -0.087 0.129\n",
"cylinders:weight:acceleration:origin -0.0292 0.028 -1.033 0.303 -0.085 0.026\n",
"displacement:weight:acceleration:origin 0.0005 0.001 0.395 0.693 -0.002 0.003\n",
"cylinders:displacement:weight:acceleration:origin -0.0002 0.000 -0.661 0.509 -0.001 0.000\n",
"horsepower:weight:acceleration:origin -0.0006 0.001 -0.636 0.526 -0.002 0.001\n",
"cylinders:horsepower:weight:acceleration:origin 0.0004 0.000 0.929 0.354 -0.000 0.001\n",
"displacement:horsepower:weight:acceleration:origin -9.366e-06 1.57e-05 -0.596 0.552 -4.03e-05 2.16e-05\n",
"cylinders:displacement:horsepower:weight:acceleration:origin 3.338e-06 3.28e-06 1.018 0.309 -3.11e-06 9.79e-06\n",
"year:origin -6.1216 4.912 -1.246 0.214 -15.790 3.547\n",
"cylinders:year:origin 4.3985 2.511 1.752 0.081 -0.543 9.340\n",
"displacement:year:origin -0.1074 0.288 -0.373 0.710 -0.675 0.460\n",
"cylinders:displacement:year:origin 0.1231 0.162 0.761 0.447 -0.195 0.441\n",
"horsepower:year:origin 0.0187 0.289 0.065 0.948 -0.551 0.588\n",
"cylinders:horsepower:year:origin -0.0343 0.153 -0.224 0.823 -0.335 0.267\n",
"displacement:horsepower:year:origin 0.0028 0.004 0.793 0.429 -0.004 0.010\n",
"cylinders:displacement:horsepower:year:origin -0.0018 0.002 -0.879 0.380 -0.006 0.002\n",
"weight:year:origin 0.0076 0.011 0.698 0.485 -0.014 0.029\n",
"cylinders:weight:year:origin -0.0079 0.005 -1.559 0.120 -0.018 0.002\n",
"displacement:weight:year:origin -9.258e-05 0.000 -0.398 0.691 -0.001 0.000\n",
"cylinders:displacement:weight:year:origin 2.574e-05 5.68e-05 0.454 0.650 -8.6e-05 0.000\n",
"horsepower:weight:year:origin -0.0001 0.000 -0.995 0.320 -0.000 0.000\n",
"cylinders:horsepower:weight:year:origin 9.734e-05 6.81e-05 1.429 0.154 -3.68e-05 0.000\n",
"displacement:horsepower:weight:year:origin -1.062e-06 2.32e-06 -0.457 0.648 -5.64e-06 3.51e-06\n",
"cylinders:displacement:horsepower:weight:year:origin 2.907e-07 5.94e-07 0.490 0.625 -8.78e-07 1.46e-06\n",
"acceleration:year:origin 0.6903 0.627 1.101 0.272 -0.544 1.925\n",
"cylinders:acceleration:year:origin -0.3571 0.283 -1.263 0.208 -0.914 0.200\n",
"displacement:acceleration:year:origin -0.0086 0.018 -0.493 0.622 -0.043 0.026\n",
"cylinders:displacement:acceleration:year:origin -0.0045 0.010 -0.456 0.648 -0.024 0.015\n",
"horsepower:acceleration:year:origin -0.0031 0.020 -0.151 0.880 -0.043 0.037\n",
"cylinders:horsepower:acceleration:year:origin 0.0022 0.010 0.218 0.827 -0.018 0.022\n",
"displacement:horsepower:acceleration:year:origin 6.583e-05 0.000 0.275 0.783 -0.000 0.001\n",
"cylinders:displacement:horsepower:acceleration:year:origin 5.775e-05 0.000 0.464 0.643 -0.000 0.000\n",
"weight:acceleration:year:origin -0.0003 0.001 -0.400 0.689 -0.002 0.001\n",
"cylinders:weight:acceleration:year:origin 0.0005 0.000 1.353 0.177 -0.000 0.001\n",
"displacement:weight:acceleration:year:origin -3.57e-06 1.4e-05 -0.254 0.799 -3.12e-05 2.41e-05\n",
"cylinders:displacement:weight:acceleration:year:origin 9.149e-07 3e-06 0.305 0.760 -4.98e-06 6.81e-06\n",
"horsepower:weight:acceleration:year:origin 1.061e-05 1.08e-05 0.979 0.328 -1.07e-05 3.2e-05\n",
"cylinders:horsepower:weight:acceleration:year:origin -6.54e-06 4.75e-06 -1.378 0.169 -1.59e-05 2.8e-06\n",
"displacement:horsepower:weight:acceleration:year:origin 1.05e-07 1.88e-07 0.559 0.577 -2.65e-07 4.75e-07\n",
"cylinders:displacement:horsepower:weight:acceleration:year:origin -3.033e-08 4.04e-08 -0.751 0.454 -1.1e-07 4.92e-08\n",
"==============================================================================\n",
"Omnibus: 46.080 Durbin-Watson: 1.751\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 130.524\n",
"Skew: 0.539 Prob(JB): 4.54e-29\n",
"Kurtosis: 5.613 Cond. No. 3.52e+16\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 3.52e+16. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"''' 3.7.9e: Including interaction terms'''\n",
"\n",
"model_interac = smf.ols(formula='mpg ~ cylinders * displacement * horsepower * weight * acceleration * year * origin', data=df).fit() #important: because we use formula, we need smf instead of sm, and ols in lower case\n",
"model_interac.summary()\n",
"\n",
"# as can be seen below we have may to many interaction terms if we include all possible (higher-order) interaction terms\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" number of coefficients: \n",
" 128\n"
]
}
],
"source": [
"print(\" number of coefficients: \\n\", len(model_interac.params))"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>mpg</td> <th> R-squared: </th> <td> 0.847</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.843</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 264.1</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>9.73e-151</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:25:16</td> <th> Log-Likelihood: </th> <td> -993.83</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 392</td> <th> AIC: </th> <td> 2006.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 383</td> <th> BIC: </th> <td> 2041.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 8</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> -2.7097</td> <td> 4.686</td> <td> -0.578</td> <td> 0.563</td> <td> -11.923</td> <td> 6.504</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders</th> <td> -2.6962</td> <td> 0.409</td> <td> -6.584</td> <td> 0.000</td> <td> -3.501</td> <td> -1.891</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement</th> <td> -0.0775</td> <td> 0.014</td> <td> -5.474</td> <td> 0.000</td> <td> -0.105</td> <td> -0.050</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders:displacement</th> <td> 0.0136</td> <td> 0.002</td> <td> 7.907</td> <td> 0.000</td> <td> 0.010</td> <td> 0.017</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower</th> <td> -0.0476</td> <td> 0.013</td> <td> -3.559</td> <td> 0.000</td> <td> -0.074</td> <td> -0.021</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight</th> <td> -0.0052</td> <td> 0.001</td> <td> -8.370</td> <td> 0.000</td> <td> -0.006</td> <td> -0.004</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration</th> <td> 0.0598</td> <td> 0.092</td> <td> 0.651</td> <td> 0.515</td> <td> -0.121</td> <td> 0.240</td>\n",
"</tr>\n",
"<tr>\n",
" <th>year</th> <td> 0.7595</td> <td> 0.047</td> <td> 16.044</td> <td> 0.000</td> <td> 0.666</td> <td> 0.853</td>\n",
"</tr>\n",
"<tr>\n",
" <th>origin</th> <td> 0.7087</td> <td> 0.274</td> <td> 2.590</td> <td> 0.010</td> <td> 0.171</td> <td> 1.247</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>35.211</td> <th> Durbin-Watson: </th> <td> 1.456</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 88.581</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 0.432</td> <th> Prob(JB): </th> <td>5.82e-20</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.162</td> <th> Cond. No. </th> <td>1.03e+05</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 1.03e+05. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: mpg R-squared: 0.847\n",
"Model: OLS Adj. R-squared: 0.843\n",
"Method: Least Squares F-statistic: 264.1\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 9.73e-151\n",
"Time: 08:25:16 Log-Likelihood: -993.83\n",
"No. Observations: 392 AIC: 2006.\n",
"Df Residuals: 383 BIC: 2041.\n",
"Df Model: 8 \n",
"Covariance Type: nonrobust \n",
"==========================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"------------------------------------------------------------------------------------------\n",
"Intercept -2.7097 4.686 -0.578 0.563 -11.923 6.504\n",
"cylinders -2.6962 0.409 -6.584 0.000 -3.501 -1.891\n",
"displacement -0.0775 0.014 -5.474 0.000 -0.105 -0.050\n",
"cylinders:displacement 0.0136 0.002 7.907 0.000 0.010 0.017\n",
"horsepower -0.0476 0.013 -3.559 0.000 -0.074 -0.021\n",
"weight -0.0052 0.001 -8.370 0.000 -0.006 -0.004\n",
"acceleration 0.0598 0.092 0.651 0.515 -0.121 0.240\n",
"year 0.7595 0.047 16.044 0.000 0.666 0.853\n",
"origin 0.7087 0.274 2.590 0.010 0.171 1.247\n",
"==============================================================================\n",
"Omnibus: 35.211 Durbin-Watson: 1.456\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 88.581\n",
"Skew: 0.432 Prob(JB): 5.82e-20\n",
"Kurtosis: 5.162 Cond. No. 1.03e+05\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 1.03e+05. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"''' 3.7.9e: Including chosen interaction terms: cylinder and displacement'''\n",
"\n",
"model_interac2 = smf.ols(formula='mpg ~ cylinders * displacement + horsepower + weight + acceleration + year + origin', data=df).fit() #important: because we use formula, we need smf instead of sm, and ols in lower case\n",
"model_interac2.summary()"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>mpg</td> <th> R-squared: </th> <td> 0.826</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.823</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 227.9</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>1.60e-140</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:25:16</td> <th> Log-Likelihood: </th> <td> -1018.0</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 392</td> <th> AIC: </th> <td> 2054.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 383</td> <th> BIC: </th> <td> 2090.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 8</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> 8.4915</td> <td> 9.044</td> <td> 0.939</td> <td> 0.348</td> <td> -9.290</td> <td> 26.273</td>\n",
"</tr>\n",
"<tr>\n",
" <th>cylinders</th> <td> -0.5042</td> <td> 0.319</td> <td> -1.579</td> <td> 0.115</td> <td> -1.132</td> <td> 0.123</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement</th> <td> 0.0157</td> <td> 0.008</td> <td> 2.081</td> <td> 0.038</td> <td> 0.001</td> <td> 0.030</td>\n",
"</tr>\n",
"<tr>\n",
" <th>horsepower</th> <td> -0.0140</td> <td> 0.014</td> <td> -1.025</td> <td> 0.306</td> <td> -0.041</td> <td> 0.013</td>\n",
"</tr>\n",
"<tr>\n",
" <th>weight</th> <td> -0.0064</td> <td> 0.001</td> <td> -9.851</td> <td> 0.000</td> <td> -0.008</td> <td> -0.005</td>\n",
"</tr>\n",
"<tr>\n",
" <th>acceleration</th> <td> 0.0918</td> <td> 0.098</td> <td> 0.941</td> <td> 0.348</td> <td> -0.100</td> <td> 0.284</td>\n",
"</tr>\n",
"<tr>\n",
" <th>year</th> <td> 0.4189</td> <td> 0.113</td> <td> 3.723</td> <td> 0.000</td> <td> 0.198</td> <td> 0.640</td>\n",
"</tr>\n",
"<tr>\n",
" <th>origin</th> <td> -14.0456</td> <td> 4.699</td> <td> -2.989</td> <td> 0.003</td> <td> -23.284</td> <td> -4.807</td>\n",
"</tr>\n",
"<tr>\n",
" <th>year:origin</th> <td> 0.1989</td> <td> 0.060</td> <td> 3.298</td> <td> 0.001</td> <td> 0.080</td> <td> 0.317</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>31.636</td> <th> Durbin-Watson: </th> <td> 1.346</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 50.301</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 0.542</td> <th> Prob(JB): </th> <td>1.19e-11</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 4.381</td> <th> Cond. No. </th> <td>1.87e+05</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 1.87e+05. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: mpg R-squared: 0.826\n",
"Model: OLS Adj. R-squared: 0.823\n",
"Method: Least Squares F-statistic: 227.9\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 1.60e-140\n",
"Time: 08:25:16 Log-Likelihood: -1018.0\n",
"No. Observations: 392 AIC: 2054.\n",
"Df Residuals: 383 BIC: 2090.\n",
"Df Model: 8 \n",
"Covariance Type: nonrobust \n",
"================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"--------------------------------------------------------------------------------\n",
"Intercept 8.4915 9.044 0.939 0.348 -9.290 26.273\n",
"cylinders -0.5042 0.319 -1.579 0.115 -1.132 0.123\n",
"displacement 0.0157 0.008 2.081 0.038 0.001 0.030\n",
"horsepower -0.0140 0.014 -1.025 0.306 -0.041 0.013\n",
"weight -0.0064 0.001 -9.851 0.000 -0.008 -0.005\n",
"acceleration 0.0918 0.098 0.941 0.348 -0.100 0.284\n",
"year 0.4189 0.113 3.723 0.000 0.198 0.640\n",
"origin -14.0456 4.699 -2.989 0.003 -23.284 -4.807\n",
"year:origin 0.1989 0.060 3.298 0.001 0.080 0.317\n",
"==============================================================================\n",
"Omnibus: 31.636 Durbin-Watson: 1.346\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 50.301\n",
"Skew: 0.542 Prob(JB): 1.19e-11\n",
"Kurtosis: 4.381 Cond. No. 1.87e+05\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 1.87e+05. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 22,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"''' 3.7.9e: Including chosen interaction term: year and origin'''\n",
"\n",
"model_interac2 = smf.ols(formula='mpg ~ cylinders + displacement + horsepower + weight + acceleration + year * origin', data=df).fit() #important: because we use formula, we need smf instead of sm, and ols in lower case\n",
"model_interac2.summary()\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"<table class=\"simpletable\">\n",
"<caption>OLS Regression Results</caption>\n",
"<tr>\n",
" <th>Dep. Variable:</th> <td>mpg</td> <th> R-squared: </th> <td> 0.689</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Model:</th> <td>OLS</td> <th> Adj. R-squared: </th> <td> 0.687</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Method:</th> <td>Least Squares</td> <th> F-statistic: </th> <td> 435.6</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Date:</th> <td>Tue, 17 Jul 2018</td> <th> Prob (F-statistic):</th> <td>1.55e-100</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Time:</th> <td>08:25:16</td> <th> Log-Likelihood: </th> <td> -1148.0</td> \n",
"</tr>\n",
"<tr>\n",
" <th>No. Observations:</th> <td> 397</td> <th> AIC: </th> <td> 2302.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Residuals:</th> <td> 394</td> <th> BIC: </th> <td> 2314.</td> \n",
"</tr>\n",
"<tr>\n",
" <th>Df Model:</th> <td> 2</td> <th> </th> <td> </td> \n",
"</tr>\n",
"<tr>\n",
" <th>Covariance Type:</th> <td>nonrobust</td> <th> </th> <td> </td> \n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <td></td> <th>coef</th> <th>std err</th> <th>t</th> <th>P>|t|</th> <th>[0.025</th> <th>0.975]</th> \n",
"</tr>\n",
"<tr>\n",
" <th>Intercept</th> <td> 42.2035</td> <td> 1.074</td> <td> 39.301</td> <td> 0.000</td> <td> 40.092</td> <td> 44.315</td>\n",
"</tr>\n",
"<tr>\n",
" <th>displacement</th> <td> -0.1393</td> <td> 0.011</td> <td> -12.539</td> <td> 0.000</td> <td> -0.161</td> <td> -0.117</td>\n",
"</tr>\n",
"<tr>\n",
" <th>I(displacement ** 2)</th> <td> 0.0002</td> <td> 2.36e-05</td> <td> 7.242</td> <td> 0.000</td> <td> 0.000</td> <td> 0.000</td>\n",
"</tr>\n",
"</table>\n",
"<table class=\"simpletable\">\n",
"<tr>\n",
" <th>Omnibus:</th> <td>42.434</td> <th> Durbin-Watson: </th> <td> 0.936</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Prob(Omnibus):</th> <td> 0.000</td> <th> Jarque-Bera (JB): </th> <td> 99.950</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Skew:</th> <td> 0.548</td> <th> Prob(JB): </th> <td>1.98e-22</td>\n",
"</tr>\n",
"<tr>\n",
" <th>Kurtosis:</th> <td> 5.200</td> <th> Cond. No. </th> <td>3.36e+05</td>\n",
"</tr>\n",
"</table><br/><br/>Warnings:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.<br/>[2] The condition number is large, 3.36e+05. This might indicate that there are<br/>strong multicollinearity or other numerical problems."
],
"text/plain": [
"<class 'statsmodels.iolib.summary.Summary'>\n",
"\"\"\"\n",
" OLS Regression Results \n",
"==============================================================================\n",
"Dep. Variable: mpg R-squared: 0.689\n",
"Model: OLS Adj. R-squared: 0.687\n",
"Method: Least Squares F-statistic: 435.6\n",
"Date: Tue, 17 Jul 2018 Prob (F-statistic): 1.55e-100\n",
"Time: 08:25:16 Log-Likelihood: -1148.0\n",
"No. Observations: 397 AIC: 2302.\n",
"Df Residuals: 394 BIC: 2314.\n",
"Df Model: 2 \n",
"Covariance Type: nonrobust \n",
"========================================================================================\n",
" coef std err t P>|t| [0.025 0.975]\n",
"----------------------------------------------------------------------------------------\n",
"Intercept 42.2035 1.074 39.301 0.000 40.092 44.315\n",
"displacement -0.1393 0.011 -12.539 0.000 -0.161 -0.117\n",
"I(displacement ** 2) 0.0002 2.36e-05 7.242 0.000 0.000 0.000\n",
"==============================================================================\n",
"Omnibus: 42.434 Durbin-Watson: 0.936\n",
"Prob(Omnibus): 0.000 Jarque-Bera (JB): 99.950\n",
"Skew: 0.548 Prob(JB): 1.98e-22\n",
"Kurtosis: 5.200 Cond. No. 3.36e+05\n",
"==============================================================================\n",
"\n",
"Warnings:\n",
"[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
"[2] The condition number is large, 3.36e+05. This might indicate that there are\n",
"strong multicollinearity or other numerical problems.\n",
"\"\"\""
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"''' 3.7.9f: Including nonlinear terms'''\n",
"\n",
"model_nonlin = smf.ols(formula='mpg ~ displacement + I(displacement**2)', data=df).fit() #important: because we use formula, we need smf instead of sm, and ols in lower case\n",
"model_nonlin.summary()"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Text(0,0.5,'mpg')"
]
},
"execution_count": 24,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"<Figure size 432x288 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Plot outputs\n",
"plt.scatter(df['displacement'],df['mpg'], color='blue')\n",
"y_hat = model_nonlin.fittedvalues\n",
"plt.scatter(df['displacement'], y_hat, color='red') \n",
"plt.xlabel('displacement')\n",
"plt.ylabel('mpg')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.4"
}
},
"nbformat": 4,
"nbformat_minor": 2
}