The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.
31/01/2022
This package contains algorithms for Scalable Spike-and-Slab (S^3), a scalable Gibbs sampling implementation for high-dimensional Bayesian regression with the continuous spike-and-slab prior.
It is based on the article “Scalable Spike-and-Slab”, by Niloy
Biswas, Lester Mackey and Xiao-Li Meng. The folder inst
contains scripts to reproduce the results of the article.
The package can be installed from R via:
# install.packages("devtools")
::install_github("niloyb/ScaleSpikeSlab/R_package")
devtools
# Install dependencies Rcpp, RcppEigen
install.packages(c("Rcpp", "RcppEigen"))
# Install additional packages to help with parallel computation and plotting
install.packages(c("doParallel", "doRNG", "foreach", "dplyr", "tidyr",
"ggplot2", "latex2exp", "reshape2", "ggpubr"))
set.seed(1)
library(ScaleSpikeSlab)
# Riboflavin linear regression dataset of Buhlmann et al. (2014)
data(riboflavin)
<- riboflavin$x
X <- t(X)
Xt <- riboflavin$y
y
# Choose hyperparamters
<- spike_slab_params(n=nrow(X),p=ncol(X)) params
library(doParallel)
registerDoParallel(cores = detectCores()-1)
library(foreach)
<- 50
no_chains <-
sss_chain_z_output foreach(i = c(1:no_chains), .combine=rbind)%dopar%{
<- spike_slab_linear(chain_length=5e3,burnin=1e3,X=X,Xt=Xt,y=y,
sss_chain tau0=params$tau0,tau1=params$tau1,q=params$q,
verbose=FALSE,store=FALSE)
return(as.vector(sss_chain$z_ergodic_avg))
}
library(dplyr)
library(ggplot2)
library(latex2exp)
<-
riboflavin_df data.frame(post_prob_mean=apply(sss_chain_z_output,2,mean),
post_prob_sd=apply(sss_chain_z_output,2,sd),
cov_index=c(1:ncol(X)), no_chains=no_chains) %>%
arrange(desc(post_prob_mean)) %>%
mutate(xaxis =1:n())
ggplot(riboflavin_df, aes(x=xaxis, y=post_prob_mean)) +
geom_point(size=2) +
geom_errorbar(aes(ymax=(post_prob_mean+3*post_prob_sd/sqrt(no_chains)),
ymin=(post_prob_mean-3*post_prob_sd/sqrt(no_chains))),
position=position_dodge(.9)) +
xlab('Riboflavin Covariates') +
ylab(TeX('Marginal posterior probabilities')) +
scale_x_continuous(trans='log10') + theme_classic(base_size = 12)
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.