The BayesSPM package helps to implement SPM analysis for 2 groups and multiple groups using Bayesian inference, particularly the Bayes factor approach in the Bayesian school of thoughts.
You can install the development version of BayesSPM like so:
devtools::install_github("LindaAmoafo/BayesSPM")This is a basic example which shows you how to solve a common problem:
## Simulate some dataset
Y = mvtnorm::rmvnorm(40,mean=rep(0,50))
grp=rep(c(1,2,3,4),each=10)
time=1:50Run the SPMBayes function
library(BayesSPM)
out<- SPMBayes(Y,group=grp, time, threshold=0.95, Q.threshold=0.05, Hypothesis="alt")
head(summary(out))## No significance at the specified Q-value setting to a Q-value = 0.3154751 that has at least 1 significant timepoint
Generate summary plot
plot(x=out, summary = TRUE) ## Get summary plot
Generate
SPM probability Maps and check significance
plot(x=out, summary = FALSE) ## Get SPM Map plot