Getting Started with MakeTables

This notebook contains interactive examples from the README to get you started with MakeTables.

Installation

First, make sure you have MakeTables installed:

pip install maketables

For development installation:

git clone https://github.com/dsliwka/maketables.git
cd maketables
pip install -e .

Descriptive Statistics Table

Let’s start with creating a descriptive statistics table using the classic auto dataset (Note that you can work with any pandas DataFrame but MakeTables also supports loading and saving dta files (Stata) as this facilitates working with variable labels).

import pandas as pd
import maketables as mt

# Load your data (here using a sample Stata dataset with the import_dta function that also stores variable labels)
df = mt.import_dta("https://www.stata-press.com/data/r18/auto.dta")

# Create descriptive statistics table
mt.DTable(df, vars=["mpg","weight","length"], bycol=["foreign"],
          caption="Descriptive Statistics")

Descriptive Statistics
	Domestic			Foreign
	N	Mean	Std. Dev.	N	Mean	Std. Dev.
Mileage (mpg)	52.00	19.83	4.74	22.00	24.77	6.61
Weight (lbs.)	52.00	3,317.12	695.36	22.00	2,315.91	433.00
Length (in.)	52.00	196.13	20.05	22.00	168.55	13.68

Regression Tables

Using PyFixest

Use the versatile PyFixest package.

import pyfixest as pf

# Fit your models here using pyfixest
est1 = pf.feols("mpg ~ weight", data=df)
est2 = pf.feols("mpg ~ weight * length", data=df)
est3 = pf.feols("mpg ~ weight * length | foreign", data=df)

# Make the table
mt.ETable([est1, est2, est3], caption = "Regression Results",)

Regression Results
	Mileage (mpg)
	(1)	(2)	(3)
coef
Weight (lbs.)	-0.006*** (0.001)	-0.004** (0.002)	-0.004 (0.001)
Length (in.)		-0.079 (0.056)	-0.082* (0.008)
Weight (lbs.) × Length (in.)		-0.000 (0.000)	-0.000 (0.000)
Intercept	39.440*** (1.614)	47.877*** (6.132)
fe
Car origin	-	-	x
stats
Observations	74	74	74
R2	0.652	0.661	0.674
Significance levels: * p < 0.1, ** p < 0.05, *** p < 0.01. Format of coefficient cell: Coefficient (Std. Error)

Using Statsmodels

MakeTables also works with Statsmodels for various types of regression models. Here is for instance a linear probability model and a probit.

import statsmodels.formula.api as smf

# Generate a dummy variable and label it
df["foreign_i"] = (df["foreign"] == "Foreign")*1
mt.set_var_labels(df, {"foreign_i": "Foreign (indicator)"})

# Fit your models 
est1 = smf.ols("foreign_i ~ weight + length + price", data=df).fit()
est2 = smf.probit("foreign_i ~ weight + length + price", data=df).fit(disp=0)

# Make the table
mt.ETable([est1, est2], model_stats=["N","r2","pseudo_r2"], 
          model_heads=["OLS","Probit"],
          caption="Regression Results (Statsmodels)")

Regression Results (Statsmodels)
	Foreign (indicator)
	OLS	Probit
	(1)	(2)
coef
Weight (lbs.)	-0.001*** (0.000)	-0.004*** (0.002)
Length (in.)	0.007 (0.005)	0.033 (0.045)
Price	0.000*** (0.000)	0.001*** (0.000)
Intercept	0.656 (0.583)	1.327 (4.992)
stats
Observations	74	74
R2	0.552	-
Pseudo R2	-	0.606
Significance levels: * p < 0.1, ** p < 0.05, *** p < 0.01. Format of coefficient cell: Coefficient (Std. Error)

Combining different packages

You can also combine models from different packages in one table.

# Here we just reestimaate the linear probability model using pyfixest instead of statsmodels
est0=pf.feols("foreign_i ~ weight + length + price", data=df)

# Make the table
mt.ETable([est0, est1, est2], model_stats=["N","r2","pseudo_r2"], 
          model_heads=["OLS (PyFixest)","OLS (Statsmodels)","Probit"],
          caption="Regression Results combining different packages")

Regression Results combining different packages
	Foreign (indicator)
	OLS (PyFixest)	OLS (Statsmodels)	Probit
	(1)	(2)	(3)
coef
Weight (lbs.)	-0.001*** (0.000)	-0.001*** (0.000)	-0.004*** (0.002)
Length (in.)	0.007 (0.005)	0.007 (0.005)	0.033 (0.045)
Price	0.000*** (0.000)	0.000*** (0.000)	0.001*** (0.000)
Intercept	0.656 (0.583)	0.656 (0.583)	1.327 (4.992)
stats
Observations	74	74	74
R2	0.552	0.552	-
Pseudo R2	-	-	0.606
Significance levels: * p < 0.1, ** p < 0.05, *** p < 0.01. Format of coefficient cell: Coefficient (Std. Error)