https://github.com/fabian-s/spikeSlabGAM

R-package: Bayesian variable selection, model choice, and regularized estimation for (spatial) generalized additive mixed regression models via stochastic search variable selection with spike-and-slab priors.
https://github.com/fabian-s/spikeSlabGAM

Last synced: 5 days ago
JSON representation

R-package: Bayesian variable selection, model choice, and regularized estimation for (spatial) generalized additive mixed regression models via stochastic search variable selection with spike-and-slab priors.

• Host: GitHub
• URL: https://github.com/fabian-s/spikeSlabGAM
• Owner: fabian-s
• License: other
• Created: 2015-07-21T13:59:03.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2024-10-22T13:42:03.000Z (over 1 year ago)
• Last Synced: 2025-12-09T09:09:46.816Z (about 2 months ago)
• Language: R
• Size: 2.03 MB
• Stars: 14
• Watchers: 3
• Forks: 3
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • Changelog: NEWS
  • License: LICENSE

Awesome Lists containing this project

• feature-selection - Code - project.org/web/packages/spikeSlabGAM/), [Paper](https://www.jstatsoft.org/article/view/v043i14)) (Uncategorized / Uncategorized)

README

          
`spikeSlabGAM`

===================

  [![R-CMD-check](https://github.com/fabian-s/spikeSlabGAM/workflows/R-CMD-check/badge.svg)](https://github.com/fabian-s/spikeSlabGAM/actions)

  

An R-package for Bayesian variable selection, model choice, and regularized estimation for (spatial) generalized additive mixed regression models via stochastic search variable selection with spike-and-slab priors.

- Fits **additive models** for **Gaussian, Binary/Binomial and Poisson responses** 

- (Correlated) **random effects** 

- Automagically performs **variable selection** and **function selection** (i.e., do I need this model term at all, is it linear or is it non-linear?), also for interactions between multiple covariates. 

- Yields **marginal posterior inclusion probabilities** for each term as well as **posterior model probabilities** and model-averaged effect estimates.

- Convenient **formula-based model specification**

- Fully Bayesian via MCMC, multiple parallelized chains for diagnostics and faster mixing, sampler implemented in `C`. 

--------------------------------------------------------------------------------

### References:

Short & applied intro (also the vignette that comes with the package, with some minor modifications...):

> Fabian Scheipl. (2011) `spikeSlabGAM`: Bayesian Variable Selection, Model Choice and Regularization for Generalized Additive Mixed Models in R. *Journal of Statistical Software*, 43(14). [[pdf]](http://www.jstatsoft.org/v43/i14)

More theory, simulation studies and real-world case studies:

> Fabian Scheipl, Ludwig Fahrmeir, Thomas Kneib (2012). Spike-and-Slab Priors for Function Selection in Structured Additive Regression Models. *Journal of the American Statistical Association*, 107(500), 1518-1532. [[pdf on arXiv]](http://arxiv.org/abs/1105.5250)