Bayesian Model Averaging for linear models with a wide choice of (customizable) priors. Built-in priors include coefficient priors (fixed, flexible and hyper-g priors), 5 kinds of model priors, moreover model sampling by enumeration or various MCMC approaches. Post-processing functions allow for inferring posterior inclusion and model probabilities, various moments, coefficient and predictive densities. Plotting functions available for posterior model size, MCMC convergence, predictive and coefficient densities, best models representation, BMA comparison.
| Version: | 0.3.3 |
| Depends: | R (≥ 2.5), methods |
| Published: | 2013-11-22 |
| Author: | Martin Feldkircher and Stefan Zeugner |
| Maintainer: | Stefan Zeugner <bms at zeugner.eu> |
| License: | Artistic-2.0 |
| URL: | http://bms.zeugner.eu |
| NeedsCompilation: | no |
| Citation: | BMS citation info |
| In views: | Bayesian |
| CRAN checks: | BMS results |
| Reference manual: | BMS.pdf |
| Vignettes: |
Bayesian Model Averaging with BMS |
| Package source: | BMS_0.3.3.tar.gz |
| Windows binaries: | r-devel: BMS_0.3.3.zip, r-release: BMS_0.3.3.zip, r-oldrel: BMS_0.3.3.zip |
| OS X Snow Leopard binaries: | r-release: BMS_0.3.3.tgz, r-oldrel: BMS_0.3.3.tgz |
| OS X Mavericks binaries: | r-release: BMS_0.3.3.tgz |
| Old sources: | BMS archive |
| Reverse depends: | TBSSurvival |
| Reverse imports: | conting |