How arnabaich96/causalmixgpd works
CausalMixGPD is an R package for Bayesian nonparametric conditional-density modeling, with heavy-tail/GPD extensions for causal inference and clustering. The principal workflow is to specify a model, select a density kernel and tail behavior, fit it, then inspect posterior results or derive causal estimands and cluster summaries. The graph also preserves the package’s kernel family, registry, bundled-data, and result-processing branches. Paths and module roles are based on the repository tree and README; sampled source excerpts were unrelated vendored website JavaScript, so internal call wiring is not verified.
Main entry points: 1-registry.R Backend registry · build-run.R Fit runner · wrappers.R Model wrappers · causal.R Causal models
Components 22
Model specification
- Model wrappers
R/wrappers.R - Causal models
R/causal.R - Clustering models
R/cluster.R - Model compiler
R/compile-spec.R - Backend registry
R/1-registry.R
Inference runtime
- Fit runner
R/build-run.R - Inference engine
R/engine.R
Analysis results
- Posterior extraction
R/extract.R - Posterior summaries
R/summary.R - Fit methods
R/methods.R - Causal estimands
R/causal.R - Cluster summaries
R/cluster.R - Example datasets
R/data-datasets.R
Density backends
- GPD tails
R/0-base-distributions.R - Normal kernel
R/normal.R - Gamma kernel
R/gamma.R - Amoroso kernel
R/amoroso.R - Cauchy kernel
R/cauchy.R - Inverse Gaussian
R/invgauss.R - Laplace kernel
R/laplace.R - Lognormal kernel
R/lognormal.R
Other components
- R user
Connections 23
- R user → Model wrappers: specifies and fits
- R user → Causal models: requests causal fit
- R user → Clustering models: requests clustering fit
- Model wrappers → Model compiler: builds specification
- Model compiler → Backend registry: resolves backend
- Backend registry → GPD tails: selects tail
- Backend registry → Normal kernel: selects kernel
- Backend registry → Gamma kernel: selects kernel
- Backend registry → Amoroso kernel: selects kernel
- Backend registry → Cauchy kernel: selects kernel
- Backend registry → Inverse Gaussian: selects kernel
- Backend registry → Laplace kernel: selects kernel
- Backend registry → Lognormal kernel: selects kernel
- Model wrappers → Fit runner: runs fit
- Fit runner → Inference engine: executes inference
- Causal models → Fit runner: fits model
- Clustering models → Fit runner: fits model
- R user → Posterior extraction: requests draws
- Posterior extraction → Posterior summaries: summarizes draws
- R user → Fit methods: inspects fit
- R user → Causal estimands: requests estimands
- R user → Cluster summaries: requests profiles
- R user → Example datasets: loads examples