rmsb: Bayesian Regression Modeling Strategies for Logistic and Ordinal Models

Frank E. Harrell's 'rmsb' package provides a Bayesian companion to the 'rms' package for regression modeling. It implements Bayesian logistic binary and ordinal regression models, including options for clustering, censoring, and partial proportional odds. The package's 'blrm()' fitting function produces objects compatible with established 'rms' utilities for estimation and visualization.

Use Cases

• Fit Bayesian logistic regression models based on the implemented 'blrm()' function
• Perform post-fit estimation and hypothesis testing based on compatibility with 'rms' functions like 'contrast()' and 'summary()'
• Generate predictive nomograms and visualizations based on integration with 'rms' graphics and 'Predict()'
• Model ordinal outcomes with departures from proportional odds based on the partial proportional odds model

Strengths

• Implements a specific, well-cited statistical model: the partial proportional odds model of Peterson and Harrell (1990)
• Designed for interoperability, as fit objects can be processed by established functions from the 'rms' package

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download
• Column-level documentation is absent; field semantics must be inferred after download

Provenance

Source

Frank E. Harrell

Collection Method

Software package for statistical modeling

Time Range

null

Freshness

Last update date is unknown; freshness unverified

Geography

null