StataFlow

A Python econometrics toolkit designed to reproduce Stata 17 estimation results with field-level validation.

from stataflow.compat.stata import reghdfe

result = reghdfe(
    df,
    y="lwage",
    x=["exper", "edu"],
    absorb="firm_id year_id",
    vce="cluster",
    cluster="firm_id",
)
result.display()

Why StataFlow

StataFlow is for researchers who want Python workflows without giving up the empirical conventions they rely on in Stata. The project is not a generic statistics library: public capabilities are validated against Stata 17 with synthetic cases, public real-data cases, and Stata/Python dual-run evidence where Stata is available.

The current public line is 1.1.0 Stable, with an additional v1.2.0+ correctness-hardening release-candidate sync prepared on July 9, 2026.

Features

• 14 Stata-style commands in Python: regress, xtreg_fe, areg, reghdfe, ivregress_2sls, ivreghdfe, logit, probit, poisson, ppmlhdfe, did_imputation, eventstudyinteract, csdid, and rdrobust.
• Two API layers: a Stata-compatible command layer (stataflow.compat.stata) and a Python-native estimator layer (stataflow.estimators).
• Stata-style output: result.display() prints compact regression tables with coefficients, standard errors, test statistics, p-values, and fit statistics.
• High-dimensional fixed effects: MAP absorption for large FE designs, multi-FE workflows, singleton handling, individual slopes, and cluster-aware VCE paths.
• Instrumental variables: 2SLS, GMM2S, LIML, Fuller/k-class, first-stage diagnostics, weak-instrument tests, and overidentification tests.
• Binary, count, and PPML models: Logit, Probit, Poisson, and PPML-HDFE with robust and clustered covariance estimators.
• Causal inference: BJS DID imputation, Sun-Abraham event-study interactions, Callaway-Sant'Anna DID, and sharp/fuzzy regression discontinuity.
• Stata-compatible syntax subsets: factor variables, analytic weights, multiple fixed effects, common VCE choices, and hard rejection of unsupported parameters.
• Validation-first development: public commands are backed by field-level Stata 17 comparison evidence.

Installation

pip install StataFlow

Python 3.10, 3.11, or 3.12 is required. Core dependencies are NumPy, pandas, SciPy, scikit-learn, and PyYAML.

Quick Start

Stata-Compatible API

from stataflow.compat.stata import regress, reghdfe, logit, ivregress_2sls, ppmlhdfe

# OLS with robust standard errors
result = regress(df, y="wage", x=["edu", "exper"], vce="robust")
result.display()

# High-dimensional fixed effects
result = reghdfe(
    df,
    y="wage",
    x=["edu", "exper"],
    absorb="firm_id year_id",
    vce="cluster",
    cluster="industry",
)

# Logit
result = logit(df, y="inlf", x=["nwifeinc", "educ", "exper"])
result.display()

# 2SLS with robust VCE
result = ivregress_2sls(
    df,
    y="lwage",
    x_exog=["educ"],
    x_endog=["exper"],
    instruments=["age", "kidslt6"],
    vce="robust",
)

# PPML with high-dimensional fixed effects
result = ppmlhdfe(
    df,
    y="trade",
    x=["lndist", "contig", "fta"],
    absorb=["exporter", "importer", "year"],
    vce="cluster",
    cluster="exporter",
)

Native Python API

from stataflow import OLS, AbsorbingOLS, Logit

model = OLS(data=df, y="wage", x=["edu", "exper"])
result = model.fit(vce="robust")
result.display()

Programmatic Results

result.display(show_ci=True)

for coef in result.coefficients:
    print(f"{coef.name}: b={coef.beta:.6f}, se={coef.std_err:.6f}, t={coef.t_stat:.2f}")

print(f"R2 = {result.fit.r2:.4f}, N = {result.sample.nobs}")

Supported Models

See Open-Source Status and Known Issues for exact support boundaries.

Validation

Recent local release-candidate checks (July 9, 2026):

• Public unit/integration suite: 405 passed
• Internal modular audit suite: 95 passed
• Golden dual-run collection guard: 839 tests collected
• Example smoke scripts: all four public demos passed
• Wheel build: stataflow-1.1.0-py3-none-any.whl built successfully
• Open-source export dry-run: 150 files selected, 0 orphan removals

Golden Stata dual-run tests require a local Stata 17 installation and are not part of the public CI gate.

Documentation

• User Guide (中文)
• Cookbook (中文)
• Examples
• Validation Evidence
• Changelog

Running Tests

# Unit and integration tests
pytest tests/ -v --ignore=tests/golden/ --ignore=tests/audit_v1_3

# Golden dual-run tests (require local Stata 17)
pytest tests/golden/ -v

License

This project is licensed under the MIT License. See LICENSE for details.