bsamGP: R-Package for Bayesian Spectral Analysis Models using Gaussian Process Priors
by Seongil Jo, Taeryon Choi, Beomjo Park, and Peter Lenk,  October. 2017.

bsamGP implements Bayesian Spectral Analysis Models that include regression with/without shape restrictions [1], quantile regression with/without shape restrictions [2], additive models, generalized linear models and density estimation [3]. For detailed description for the package, please check the manual and the paper [4]. The package is availabe on CRAN (link)To install package, use R command install.packages("bsamGP"). (For experimental version that supports additional M and W shape restriction for BSAR and BSAQ model, you may download here (Windows / Mac).) 
 

We provide a simple example. To carry out examples, install bsamGP into R and run following example R-code.

Bayesian Spectral Analysis Regression (BSAR)
  • Example for mean regression of function with monotone increasing and convex constraint.
Bayesian Spectral Analysis Quantile Regression (BSAQ)
  • Example for median regression of function with monotone convex to concave increasing constraint
Bayesian Spectral Analysis Generalized Linear Regression (GBSAR)
  • Example for probit regression of function with monotone increasing and convex constraint
Bayesian Spectral Analysis Density Estimation (BSAD)
  • Example for semiparametric density estimation with truncated gamma distribution and sigmoid density.

[1] Lenk P, Choi T (2017).“Bayesian analysis of shape-restricted functions using Gaussian process
priors.” Stat Sinica, 27(1), 43–69.
[2] Jo S, Roh T, Choi T (2016). “Bayesian spectral analysis models for quantile regression with
Dirichlet process mixtures.” J. Nonparametr. Stat., 28(1), 177–206.
[3] Lenk P (2003). “Bayesian semiparametric density estimation and model verification using a
logistic-Gaussian process.” J. Comput. Graph. Stat., 12(3), 548–565
[4] Jo, S., Choi, T., Park, B., & Lenk, P. (2017) "bsamGP: An R package for Bayesian Spectral Analysis Models using Gaussian Process Priors", Submitted to Journal of Statistical Software