{ "cells": [ { "cell_type": "markdown", "id": "7794a688", "metadata": {}, "source": [ "(regmodels)=\n", "\n", "# Linear and Generalized Linear Models\n", "\n", "Let's simulate some data for illustrations." ] }, { "cell_type": "code", "execution_count": 1, "id": "fe26ef75", "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import statsmodels.api as sm\n", "\n", "nobs = 200\n", "ncov = 5\n", "np.random.seed(123)\n", "x = np.random.random((nobs, ncov)) # Uniform over [0, 1)\n", "beta = np.repeat(1, ncov)\n", "y = 2 + np.dot(x, beta) + np.random.normal(size = nobs)" ] }, { "cell_type": "code", "execution_count": 2, "id": "cf94ac0f", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(200,)" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "y.shape" ] }, { "cell_type": "code", "execution_count": 3, "id": "39b8c8e9", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(200, 5)" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "x.shape" ] }, { "cell_type": "markdown", "id": "299101b5", "metadata": {}, "source": [ "That is, the true linear regression model is\n", "```{math}\n", "y = 2 + x_1 + x_2 + x_3 + x_4 + x_5 + \\epsilon.\n", "```\n", "\n", "A regression model for the observed data can be fitted as" ] }, { "cell_type": "code", "execution_count": 4, "id": "49170256", "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
OLS Regression Results
Dep. Variable: y R-squared: 0.309
Model: OLS Adj. R-squared: 0.292
Method: Least Squares F-statistic: 17.38
Date: Fri, 29 Jul 2022 Prob (F-statistic): 3.31e-14
Time: 22:52:58 Log-Likelihood: -272.91
No. Observations: 200 AIC: 557.8
Df Residuals: 194 BIC: 577.6
Df Model: 5
Covariance Type: nonrobust
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err t P>|t| [0.025 0.975]
const 1.8754 0.282 6.656 0.000 1.320 2.431
x1 1.1703 0.248 4.723 0.000 0.682 1.659
x2 0.8988 0.235 3.825 0.000 0.435 1.362
x3 0.9784 0.238 4.114 0.000 0.509 1.448
x4 1.3418 0.250 5.367 0.000 0.849 1.835
x5 0.6027 0.239 2.519 0.013 0.131 1.075
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Omnibus: 0.810 Durbin-Watson: 1.978
Prob(Omnibus): 0.667 Jarque-Bera (JB): 0.903
Skew: -0.144 Prob(JB): 0.637
Kurtosis: 2.839 Cond. No. 8.31


Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified." ], "text/plain": [ "\n", "\"\"\"\n", " OLS Regression Results \n", "==============================================================================\n", "Dep. Variable: y R-squared: 0.309\n", "Model: OLS Adj. R-squared: 0.292\n", "Method: Least Squares F-statistic: 17.38\n", "Date: Fri, 29 Jul 2022 Prob (F-statistic): 3.31e-14\n", "Time: 22:52:58 Log-Likelihood: -272.91\n", "No. Observations: 200 AIC: 557.8\n", "Df Residuals: 194 BIC: 577.6\n", "Df Model: 5 \n", "Covariance Type: nonrobust \n", "==============================================================================\n", " coef std err t P>|t| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "const 1.8754 0.282 6.656 0.000 1.320 2.431\n", "x1 1.1703 0.248 4.723 0.000 0.682 1.659\n", "x2 0.8988 0.235 3.825 0.000 0.435 1.362\n", "x3 0.9784 0.238 4.114 0.000 0.509 1.448\n", "x4 1.3418 0.250 5.367 0.000 0.849 1.835\n", "x5 0.6027 0.239 2.519 0.013 0.131 1.075\n", "==============================================================================\n", "Omnibus: 0.810 Durbin-Watson: 1.978\n", "Prob(Omnibus): 0.667 Jarque-Bera (JB): 0.903\n", "Skew: -0.144 Prob(JB): 0.637\n", "Kurtosis: 2.839 Cond. No. 8.31\n", "==============================================================================\n", "\n", "Notes:\n", "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n", "\"\"\"" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "xmat = sm.add_constant(x)\n", "mymod = sm.OLS(y, xmat)\n", "myfit = mymod.fit()\n", "myfit.summary()" ] }, { "cell_type": "markdown", "id": "76b553d2", "metadata": {}, "source": [ "How are the regression coefficients interpreted? Intercept?\n", "\n", "Why does it make sense to center the covariates?\n", "\n", "Now we form a data frame with the variables" ] }, { "cell_type": "code", "execution_count": 5, "id": "22ee2337", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "RangeIndex: 200 entries, 0 to 199\n", "Data columns (total 6 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 y 200 non-null float64\n", " 1 x1 200 non-null float64\n", " 2 x2 200 non-null float64\n", " 3 x3 200 non-null float64\n", " 4 x4 200 non-null float64\n", " 5 x5 200 non-null float64\n", "dtypes: float64(6)\n", "memory usage: 9.5 KB\n" ] } ], "source": [ "import pandas as pd\n", "df = np.concatenate((y.reshape((nobs, 1)), x), axis = 1)\n", "df = pd.DataFrame(data = df,\n", " columns = [\"y\"] + [\"x\" + str(i) for i in range(1,\n", " ncov + 1)])\n", "df.info()" ] }, { "cell_type": "markdown", "id": "7dc7babc", "metadata": {}, "source": [ "Let's use a formula to specify the regression model as in R, and fit\n", "a robust linear model (`rlm`) instead of OLS." ] }, { "cell_type": "code", "execution_count": 6, "id": "785aaa14", "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Robust linear Model Regression Results
Dep. Variable: y No. Observations: 200
Model: RLM Df Residuals: 194
Method: IRLS Df Model: 5
Norm: HuberT
Scale Est.: mad
Cov Type: H1
Date: Fri, 29 Jul 2022
Time: 22:52:58
No. Iterations: 16
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err z P>|z| [0.025 0.975]
Intercept 1.8353 0.294 6.246 0.000 1.259 2.411
x1 1.1254 0.258 4.355 0.000 0.619 1.632
x2 0.9664 0.245 3.944 0.000 0.486 1.447
x3 0.9995 0.248 4.029 0.000 0.513 1.486
x4 1.3275 0.261 5.091 0.000 0.816 1.839
x5 0.6768 0.250 2.712 0.007 0.188 1.166


If the model instance has been used for another fit with different fit parameters, then the fit options might not be the correct ones anymore ." ], "text/plain": [ "\n", "\"\"\"\n", " Robust linear Model Regression Results \n", "==============================================================================\n", "Dep. Variable: y No. Observations: 200\n", "Model: RLM Df Residuals: 194\n", "Method: IRLS Df Model: 5\n", "Norm: HuberT \n", "Scale Est.: mad \n", "Cov Type: H1 \n", "Date: Fri, 29 Jul 2022 \n", "Time: 22:52:58 \n", "No. Iterations: 16 \n", "==============================================================================\n", " coef std err z P>|z| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "Intercept 1.8353 0.294 6.246 0.000 1.259 2.411\n", "x1 1.1254 0.258 4.355 0.000 0.619 1.632\n", "x2 0.9664 0.245 3.944 0.000 0.486 1.447\n", "x3 0.9995 0.248 4.029 0.000 0.513 1.486\n", "x4 1.3275 0.261 5.091 0.000 0.816 1.839\n", "x5 0.6768 0.250 2.712 0.007 0.188 1.166\n", "==============================================================================\n", "\n", "If the model instance has been used for another fit with different fit parameters, then the fit options might not be the correct ones anymore .\n", "\"\"\"" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import statsmodels.formula.api as smf\n", "mymod = smf.rlm(formula = \"y ~ x1 + x2 + x3 + x4 + x5\", data = df)\n", "myfit = mymod.fit()\n", "myfit.summary()" ] }, { "cell_type": "markdown", "id": "ba550d55", "metadata": {}, "source": [ "For model diagnostics, one can check residual plots." ] }, { "cell_type": "code", "execution_count": 7, "id": "40bbdb08", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "eval_env: 1\n" ] }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "filenames": { "image/png": "/home/runner/work/ids-s22/ids-s22/notes/_build/jupyter_execute/docs/models_11_1.png" }, "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "\n", "myOlsFit = smf.ols(formula = \"y ~ x1 + x2 + x3 + x4 + x5\", data = df).fit()\n", "fig = plt.figure(figsize = (12, 8))\n", "## residual versus x1; can do the same for other covariates\n", "fig = sm.graphics.plot_regress_exog(myOlsFit, 'x1', fig=fig)" ] }, { "cell_type": "markdown", "id": "8a52b322", "metadata": {}, "source": [ "See more on [residual diagnostics and specification\n", "tests](https://www.statsmodels.org/stable/stats.html#residual-diagnostics-and-specification-tests).\n", "\n", "\n", "# Generalized Linear Regression\n", "\n", "A linear regression model cannot be applied to presence/absence or\n", "count data. \n", "\n", "\n", "Binary or count data need to be modeled under a generlized\n", "framework. Consider a binary or count variable $Y$ with possible\n", "covariates $X$. A generalized model describes a transformation $g$\n", "of the conditional mean $E[Y | X]$ by a linear predictor\n", "$X^{\\top}\\beta$. That is\n", "\\begin{equation*}\n", "g( E[Y | X] ) = X^{\\top} \\beta.\n", "\\end{equation*}\n", "The transformation $g$ is known as the link function.\n", "\n", "\n", "For logistic regression with binary outcomes, the link function is\n", "the logit function\n", "\\begin{equation*}\n", "g(u) = \\log \\frac{u}{1 - u}, \\quad u \\in (0, 1).\n", "\\end{equation*}\n", "\n", "What is the interpretation of the regression coefficients in a\n", "logistic regression? Intercept?\n", "\n", "A logistic regression can be fit with `statsmodels.api.glm`.\n", "\n", "Let's generate some binary data first by dichotomizing existing variables." ] }, { "cell_type": "code", "execution_count": 8, "id": "6cb47365", "metadata": {}, "outputs": [], "source": [ "df['yb' ] = np.where(df['y' ] > 2.5, 1, 0)\n", "df['x1b'] = np.where(df['x1'] > 0.5, 1, 0)" ] }, { "cell_type": "markdown", "id": "887cae86", "metadata": {}, "source": [ "Fit a logistic regression for `y1b`." ] }, { "cell_type": "code", "execution_count": 9, "id": "5c74555f", "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Generalized Linear Model Regression Results
Dep. Variable: yb No. Observations: 200
Model: GLM Df Residuals: 194
Model Family: Binomial Df Model: 5
Link Function: Logit Scale: 1.0000
Method: IRLS Log-Likelihood: -30.654
Date: Fri, 29 Jul 2022 Deviance: 61.307
Time: 22:53:00 Pearson chi2: 217.
No. Iterations: 7 Pseudo R-squ. (CS): 0.05871
Covariance Type: nonrobust
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err z P>|z| [0.025 0.975]
Intercept -0.4354 1.309 -0.333 0.739 -3.002 2.131
x1b 1.2721 0.845 1.505 0.132 -0.384 2.928
x2 2.6897 1.418 1.897 0.058 -0.089 5.469
x3 -0.0537 1.270 -0.042 0.966 -2.542 2.435
x4 2.6576 1.438 1.848 0.065 -0.160 5.476
x5 1.8752 1.330 1.409 0.159 -0.732 4.483
" ], "text/plain": [ "\n", "\"\"\"\n", " Generalized Linear Model Regression Results \n", "==============================================================================\n", "Dep. Variable: yb No. Observations: 200\n", "Model: GLM Df Residuals: 194\n", "Model Family: Binomial Df Model: 5\n", "Link Function: Logit Scale: 1.0000\n", "Method: IRLS Log-Likelihood: -30.654\n", "Date: Fri, 29 Jul 2022 Deviance: 61.307\n", "Time: 22:53:00 Pearson chi2: 217.\n", "No. Iterations: 7 Pseudo R-squ. (CS): 0.05871\n", "Covariance Type: nonrobust \n", "==============================================================================\n", " coef std err z P>|z| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "Intercept -0.4354 1.309 -0.333 0.739 -3.002 2.131\n", "x1b 1.2721 0.845 1.505 0.132 -0.384 2.928\n", "x2 2.6897 1.418 1.897 0.058 -0.089 5.469\n", "x3 -0.0537 1.270 -0.042 0.966 -2.542 2.435\n", "x4 2.6576 1.438 1.848 0.065 -0.160 5.476\n", "x5 1.8752 1.330 1.409 0.159 -0.732 4.483\n", "==============================================================================\n", "\"\"\"" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "mylogistic = smf.glm(formula = 'yb ~ x1b + x2 + x3 + x4 + x5', data = df,\n", " family = sm.families.Binomial())\n", "mylfit = mylogistic.fit()\n", "mylfit.summary()" ] }, { "cell_type": "markdown", "id": "1e4dde08", "metadata": {}, "source": [ "If we treat `y1b` as count data, a Poisson regression can be fitted." ] }, { "cell_type": "code", "execution_count": 10, "id": "0986e206", "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
Generalized Linear Model Regression Results
Dep. Variable: yb No. Observations: 200
Model: GLM Df Residuals: 194
Model Family: Poisson Df Model: 5
Link Function: Log Scale: 1.0000
Method: IRLS Log-Likelihood: -199.55
Date: Fri, 29 Jul 2022 Deviance: 17.091
Time: 22:53:00 Pearson chi2: 8.99
No. Iterations: 4 Pseudo R-squ. (CS): 0.002487
Covariance Type: nonrobust
\n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "\n", " \n", "\n", "
coef std err z P>|z| [0.025 0.975]
Intercept -0.2045 0.288 -0.710 0.477 -0.769 0.360
x1b 0.0445 0.146 0.304 0.761 -0.242 0.331
x2 0.0965 0.249 0.387 0.699 -0.392 0.585
x3 -0.0088 0.254 -0.035 0.972 -0.507 0.490
x4 0.1072 0.267 0.402 0.688 -0.416 0.630
x5 0.0750 0.253 0.296 0.767 -0.421 0.571
" ], "text/plain": [ "\n", "\"\"\"\n", " Generalized Linear Model Regression Results \n", "==============================================================================\n", "Dep. Variable: yb No. Observations: 200\n", "Model: GLM Df Residuals: 194\n", "Model Family: Poisson Df Model: 5\n", "Link Function: Log Scale: 1.0000\n", "Method: IRLS Log-Likelihood: -199.55\n", "Date: Fri, 29 Jul 2022 Deviance: 17.091\n", "Time: 22:53:00 Pearson chi2: 8.99\n", "No. Iterations: 4 Pseudo R-squ. (CS): 0.002487\n", "Covariance Type: nonrobust \n", "==============================================================================\n", " coef std err z P>|z| [0.025 0.975]\n", "------------------------------------------------------------------------------\n", "Intercept -0.2045 0.288 -0.710 0.477 -0.769 0.360\n", "x1b 0.0445 0.146 0.304 0.761 -0.242 0.331\n", "x2 0.0965 0.249 0.387 0.699 -0.392 0.585\n", "x3 -0.0088 0.254 -0.035 0.972 -0.507 0.490\n", "x4 0.1072 0.267 0.402 0.688 -0.416 0.630\n", "x5 0.0750 0.253 0.296 0.767 -0.421 0.571\n", "==============================================================================\n", "\"\"\"" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "myPois = smf.glm(formula = 'yb ~ x1b + x2 + x3 + x4 + x5', data = df,\n", " family = sm.families.Poisson())\n", "mypfit = myPois.fit()\n", "mypfit.summary()" ] }, { "cell_type": "code", "execution_count": null, "id": "596c87b5", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "jupytext": { "text_representation": { "extension": ".md", "format_name": "myst", "format_version": 0.13, "jupytext_version": "1.11.5" } }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.13" }, "source_map": [ 12, 20, 32, 36, 38, 47, 52, 60, 67, 72, 77, 81, 88, 124, 127, 131, 136, 140, 147 ] }, "nbformat": 4, "nbformat_minor": 5 }