synthdid-py
===========

This package is a Python port of the reference R implementation of synthetic
difference-in-differences:

    synthdid: Synthetic Difference-in-Difference Estimation
    Dmitry Arkhangelsky, Susan Athey, David A. Hirshberg, Guido W. Imbens
    and Stefan Wager
    https://github.com/synth-inference/synthdid
    Copyright 2019, Stanford University
    Licensed under GPL (>= 2) | BSD 3-Clause; this port takes the BSD option.

The estimator, the Frank-Wolfe solvers, the three variance estimators and the
diagnostic plots follow that implementation closely, so that results agree.

Method reference
----------------
    Dmitry Arkhangelsky, Susan Athey, David A. Hirshberg, Guido W. Imbens and
    Stefan Wager (2021). "Synthetic Difference-in-Differences."
    American Economic Review 111(12), 4088-4118.

Staggered adoption
------------------
The staggered-adoption extension in `synthdid.staggered` is not part of the R
package. Its cohort decomposition and treated-unit-period weighting follow the
Stata package `sdid`:

    Damian Clarke, Daniel Pailanir, Susan Athey and Guido Imbens (2023).
    "Synthetic Difference-in-Differences Estimation."
    https://github.com/Daniel-Pailanir/sdid

Bundled data
------------
`california_prop99.csv` is redistributed from the R package, which derives it
from the replication materials of:

    Alberto Abadie, Alexis Diamond and Jens Hainmueller (2010). "Synthetic
    Control Methods for Comparative Case Studies: Estimating the Effect of
    California's Tobacco Control Program." Journal of the American Statistical
    Association 105(490), 493-505.

`CPS.csv` and `PENN.csv` are likewise redistributed from the R package.
