This is a package for easily performing regression analysis in Python. All the heavy lifting is being done by Pandas and Statsmodels; this is just an interface that should be familiar to anyone who has used Stata, with some funny implementation details that make the output a bit more like Stata output (i.e. the fixed-effects implementation has an "intercept" term).
##Examples
from regressions import RDataFrame
# RDataFrame subclasses the pandas DataFrame and attaches regression methods such
# as regress and xtreg.
grunfeld = RDataFrame.from_csv('regressions/Data/grunfeld.csv')
grunfeld.head()
# FIRM YEAR I F C
# 0 1 1935 317.6 3078.5 2.8
# 1 1 1936 391.8 4661.7 52.6
# 2 1 1937 410.6 5387.1 156.9
# 3 1 1938 257.7 2792.2 209.2
# 4 1 1939 330.8 4313.2 203.4
# Linear regression
grunfeld.regress('I ~ F + C')
# OLS Regression Results
# ==============================================================================
# Dep. Variable: I R-squared: 0.812
# Model: OLS Adj. R-squared: 0.811
# Method: Least Squares F-statistic: 426.6
# Date: Mon, 23 Nov 2015 Prob (F-statistic): 2.58e-72
# Time: 10:22:00 Log-Likelihood: -1191.8
# No. Observations: 200 AIC: 2390.
# Df Residuals: 197 BIC: 2399.
# Df Model: 2
# Covariance Type: nonrobust
# ==============================================================================
# coef std err t P>|t| [95.0% Conf. Int.]
# ------------------------------------------------------------------------------
# Intercept -42.7144 9.512 -4.491 0.000 -61.472 -23.957
# F 0.1156 0.006 19.803 0.000 0.104 0.127
# C 0.2307 0.025 9.055 0.000 0.180 0.281
# ==============================================================================
# Omnibus: 27.240 Durbin-Watson: 0.358
# Prob(Omnibus): 0.000 Jarque-Bera (JB): 89.572
# Skew: 0.469 Prob(JB): 3.55e-20
# Kurtosis: 6.141 Cond. No. 2.46e+03
# ==============================================================================
#
# Warnings:
# [1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
# [2] The condition number is large, 2.46e+03. This might indicate that there are
# strong multicollinearity or other numerical problems.
#Cluster standard errors
grunfeld.regress('I ~ F + C', cluster='FIRM')
# OLS Regression Results
# ==============================================================================
# Dep. Variable: I R-squared: 0.812
# Model: OLS Adj. R-squared: 0.811
# Method: Least Squares F-statistic: 51.59
# Date: Mon, 23 Nov 2015 Prob (F-statistic): 1.17e-05
# Time: 10:23:50 Log-Likelihood: -1191.8
# No. Observations: 200 AIC: 2390.
# Df Residuals: 197 BIC: 2399.
# Df Model: 2
# Covariance Type: cluster
# ==============================================================================
# coef std err z P>|z| [95.0% Conf. Int.]
# ------------------------------------------------------------------------------
# Intercept -42.7144 20.425 -2.091 0.037 -82.747 -2.682
# F 0.1156 0.016 7.271 0.000 0.084 0.147
# C 0.2307 0.085 2.715 0.007 0.064 0.397
# ==============================================================================
# Omnibus: 27.240 Durbin-Watson: 0.358
# Prob(Omnibus): 0.000 Jarque-Bera (JB): 89.572
# Skew: 0.469 Prob(JB): 3.55e-20
# Kurtosis: 6.141 Cond. No. 2.46e+03
# ==============================================================================
# Warnings:
# [1] Standard Errors are robust tocluster correlation (cluster)
# [2] The condition number is large, 2.46e+03. This might indicate that there are
# strong multicollinearity or other numerical problems.
# Tell regressions.py that this is a panel
grunfeld.xtset(i='FIRM', t='YEAR')
# Balanced Panel.
# attribute value
# i FIRM
# t YEAR
# n 10
# T 20
# N 200
# Width 5
# Fixed-effects regression
grunfeld.xtreg('I ~ F + C', 'fe')
# OLS Regression Results
# ==============================================================================
# Dep. Variable: I R-squared: 0.767
# Model: OLS Adj. R-squared: 0.753
# Method: Least Squares F-statistic: 309.0
# Date: Mon, 23 Nov 2015 Prob (F-statistic): 3.75e-60
# Time: 10:36:17 Log-Likelihood: -1070.8
# No. Observations: 200 AIC: 2148.
# Df Residuals: 188 BIC: 2157.
# Df Model: 2
# Covariance Type: nonrobust
# ==============================================================================
# coef std err t P>|t| [95.0% Conf. Int.]
# ------------------------------------------------------------------------------
# Intercept -58.7439 12.454 -4.717 0.000 -83.311 -34.177
# F 0.1101 0.012 9.288 0.000 0.087 0.134
# C 0.3101 0.017 17.867 0.000 0.276 0.344
# ==============================================================================
# Omnibus: 29.604 Durbin-Watson: 1.079
# Prob(Omnibus): 0.000 Jarque-Bera (JB): 162.947
# Skew: 0.283 Prob(JB): 4.13e-36
# Kurtosis: 7.386 Cond. No. 3.91e+03
# ==============================================================================
#
# Warnings:
# [1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
# [2] The condition number is large, 3.91e+03. This might indicate that there are
# strong multicollinearity or other numerical problems.
#Fixed effects with "robust" (HC1) standard errors. Defaults to clustering on the 'i' variable.
grunfeld.xtreg('I ~ F + C', 'fe', 'robust')
# OLS Regression Results
# ==============================================================================
# Dep. Variable: I R-squared: 0.767
# Model: OLS Adj. R-squared: 0.753
# Method: Least Squares F-statistic: 28.31
# Date: Mon, 23 Nov 2015 Prob (F-statistic): 0.000131
# Time: 10:41:20 Log-Likelihood: -1070.8
# No. Observations: 200 AIC: 2148.
# Df Residuals: 188 BIC: 2157.
# Df Model: 2
# Covariance Type: cluster
# ==============================================================================
# coef std err z P>|z| [95.0% Conf. Int.]
# ------------------------------------------------------------------------------
# Intercept -58.7439 27.603 -2.128 0.033 -112.845 -4.643
# F 0.1101 0.015 7.248 0.000 0.080 0.140
# C 0.3101 0.053 5.878 0.000 0.207 0.413
# ==============================================================================
# Omnibus: 29.604 Durbin-Watson: 1.079
# Prob(Omnibus): 0.000 Jarque-Bera (JB): 162.947
# Skew: 0.283 Prob(JB): 4.13e-36
# Kurtosis: 7.386 Cond. No. 3.91e+03
# ==============================================================================
The RegressionTable object allows you to pretty-print the results from a list of different Regressions. It is similar to Stata's outreg.
r1 = grunfeld.regress('I ~ F')
r2 = grunfeld.regress('I ~ C')
r3 = grunfeld.regress('I ~ F + C')
rtable = RegressionTable([r1, r2, r3])
print(rtable.table()) # by default, output is in the 'pipe' format which is readable by github markdown
# | | (1) | (2) | (3) |
# |:----------|:---------|:---------|:-----------|
# | C | | 0.477*** | 0.231*** |
# | | | (0.038) | (0.025) |
# | F | 0.141*** | | 0.116*** |
# | | (0.006) | | (0.006) |
# | Intercept | -6.976 | 14.236 | -42.714*** |
# | | (10.273) | (15.639) | (9.512) |
# Headers are numbers by default, but the models can be renamed
rtable.model_names = ['Model 1', 'Model 2', 'Model 3']
print(rtable.table())
# | | Model 1 | Model 2 | Model 3 |
# |:----------|:----------|:----------|:-----------|
# | C | | 0.477*** | 0.231*** |
# | | | (0.038) | (0.025) |
# | F | 0.141*** | | 0.116*** |
# | | (0.006) | | (0.006) |
# | Intercept | -6.976 | 14.236 | -42.714*** |
# | | (10.273) | (15.639) | (9.512) |
# coefficients can be renamed as well
rtable.coefficient_names = {'C': 'capital', 'F': 'value'} # ommitting variables here omits them from the table
print(rtable.table())
# | | Model 1 | Model 2 | Model 3 |
# |:--------|:----------|:----------|:----------|
# | capital | | 0.477*** | 0.231*** |
# | | | (0.038) | (0.025) |
# | value | 0.141*** | | 0.116*** |
# | | (0.006) | | (0.006) |The default output is readable by markdown
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| capital | 0.477*** | 0.231*** | |
| (0.038) | (0.025) | ||
| value | 0.141*** | 0.116*** | |
| (0.006) | (0.006) |
but other options are available:
print(rtable.table(tablefmt='html'))
# <table>
# <tr><th> </th><th>Model 1 </th><th>Model 2 </th><th>Model 3 </th></tr>
# <tr><td>capital</td><td> </td><td>0.477*** </td><td>0.231*** </td></tr>
# <tr><td> </td><td> </td><td>(0.038) </td><td>(0.025) </td></tr>
# <tr><td>value </td><td>0.141*** </td><td> </td><td>0.116*** </td></tr>
# <tr><td> </td><td>(0.006) </td><td> </td><td>(0.006) </td></tr>
# </table>
print(rtable.table(tablefmt='latex'))
# \begin{tabular}{llll}
# \hline
# & Model 1 & Model 2 & Model 3 \\
# \hline
# capital & & 0.477*** & 0.231*** \\
# & & (0.038) & (0.025) \\
# value & 0.141*** & & 0.116*** \\
# & (0.006) & & (0.006) \\
# \hline
# \end{tabular}Tables are printed by the https://pypi.python.org/pypi/tabulate package. See the tabulate documentation for more details about table formats.