Ergodic
Copyright 2026 Ergodic

This product includes software developed at Ergodic (https://ergodic.ai).

Bundled datasets (ergodic.datasets)
-----------------------------------

The package ships a small set of teaching datasets. Their values are public
domain or permissively licensed, and we credit the sources here.

- birthweight: maternal smoking and infant birthweight. From US natality records
  (public domain, National Center for Health Statistics), as used by Almond,
  Chay and Lee, "The Costs of Low Birth Weight", Quarterly Journal of Economics
  (2005), and Cattaneo, "Efficient Semiparametric Estimation of Multi-valued
  Treatment Effects", Journal of Econometrics (2010).

- pension401k: 401(k) eligibility and net financial assets. From the Survey of
  Income and Program Participation (public domain, US Census Bureau), as used by
  Chernozhukov, Chetverikov, Demirer, Duflo, Hansen, Newey and Robins,
  "Double/Debiased Machine Learning", Econometrics Journal (2018).

- smoking: California Proposition 99 and per-capita cigarette sales. Public-domain
  replication data from Abadie, Diamond and Hainmueller, "Synthetic Control
  Methods for Comparative Case Studies", Journal of the American Statistical
  Association (2010).

- germany: German reunification and GDP per capita. Public-domain replication
  data from Abadie, Diamond and Hainmueller, "Comparative Politics and the
  Synthetic Control Method", American Journal of Political Science (2015),
  redistributed via the MIT-licensed pysyncon package.

- berkeley: UC Berkeley graduate admissions, 1973. The published aggregate table
  from Bickel, Hammel and O'Connell, "Sex Bias in Graduate Admissions: Data from
  Berkeley", Science (1975), stored as counts in the loader.

- asia, earthquake: small Bayesian networks. The structures and conditional
  tables are implemented from their source papers (Lauritzen and Spiegelhalter,
  "Local Computations with Probabilities on Graphical Structures", Journal of the
  Royal Statistical Society B, 1988; Pearl, "Probabilistic Reasoning in
  Intelligent Systems", 1988). The data are sampled from those networks, so they
  are the package's own.
