Software

Open-source software for biomedical data.

A small portfolio of open-source software developed alongside the methodological work described on the Research page. Each package implements a method from a peer-reviewed paper or in-progress manuscript. Issue reports, pull requests, and questions are welcome through GitHub.

Current packages
causalGEM
Python · published

Bivariate causal discovery via the generative exposure model.

Purkayastha & Song (2026). Statistica Sinica, DOI: 10.5705/ss.202025.0236.

GitHub·Paper
swaoBN
R · in development

Survey-weighted, covariate-adjusted ordinal Bayesian network learning.

Purkayastha, Frank, Russell & Hausmann (2026). Manuscript in submission.

GitHub
FedGEE
R · in development

Federated generalized estimating equations for privacy-preserving population-level inference in multicenter longitudinal data.

Chang & Purkayastha (2026). Manuscript in submission.

GitHub
comet
R · published

Collider–mediator testing using information theory.

Purkayastha & Song (2024).

GitHub
Earlier work
SEIRfansy
R · CRAN

Extended SEIR compartmental model with false-negative correction and symptom-based testing for COVID-19 surveillance.

Bhaduri, Kundu, Purkayastha et al. (2022). Statistics in Medicine, 41(13):2317–2337.

CRAN·GitHub·Paper