I am committed to developing open-source software that implements the methods I work on, contributing to shared research infrastructure and promoting transparent, reproducible science.
R library: bml
bml is an R library for fitting Bayesian multiple-membership multilevel models with parameterizable weight functions via JAGS. It allows researchers to model how multiple lower-level units jointly influence higher-level outcomes, enabling a principled analysis of micro-to-macro relationships. The package supports diverse outcome types (e.g., linear, logistic, survival), flexible weighting schemes, and complex dependence structures, making it well suited for settings where group outcomes depend on overlapping lower-level actors, such as coalition governments, teams, spatial neighbors, or networked units.
R library: ineqx
ineqx is an R library for descriptive and explanatory variance decomposition, enabling researchers to analyze how inequality in an outcome (e.g., income) divides into within- and between-group components, both cross-sectionally and over time. Beyond classic descriptive approaches, ineqx implements an explanatory framework that assesses how a treatment or binary predictor affects inequality across the full distribution, disentangling changes into treatment effects on within- and between-group inequality, compositional shifts, and pre-treatment differences.