If you're looking to make a nice binned scatter plot with a regression line and you don't need to account for any control variables use seaborn.regplot! If you're looking for a Python analog to Stata's binscatter, read on.

Stata's binscatter is described fully by Michael Stepner here. You can use this Python version in essentially the same way you use Matplotlib functions like plot and scatter. A more extensive description is here.

Getting started

1. Copy and paste: Binscatter's meaningful code consists of consists of just one file. You can copy binscatter/binscatter.py into the directory the rest of your code is in. If you are missing dependencies, you may first need to install them. One way of doing that is with pip install -r requirements.txt.

2. Install binscatter via pip: To make it easier to use binscatter in multiple projects and directories, open a terminal and

   • git clone https://github.com/esantorella/binscatter.git
   • cd binscatter
   • pip install .

Usage

import binscatter
import pandas as pd
import numpy as np
from matplotlib import pyplot as plt

# Create fake data
n_obs = 1000
data = pd.DataFrame({"experience": np.random.poisson(4, n_obs) + 1})
data["Tenure"] = data["experience"] + np.random.normal(0, 1, n_obs)
data["Wage"] = data["experience"] + data["Tenure"] + np.random.normal(0, 1, n_obs)
fig, axes = plt.subplots(2, sharex=True, sharey=True)

# Binned scatter plot of Wage vs Tenure
axes[0].binscatter(data["Tenure"], data["Wage"])

# Binned scatter plot that partials out the effect of experience
axes[1].binscatter(
    data["Tenure"],
    data["Wage"],
    controls=data["experience"],
    recenter_x=True,
    recenter_y=True,
)
axes[1].set_xlabel("Tenure (residualized, recentered)")
axes[1].set_ylabel("Wage (residualized, recentered)")

plt.tight_layout()
plt.show()