plot.hcoef.post: Plot Method for Hierarchical Model Coefficients

In cudaBayesreg: CUDA Parallel Implementation of a Bayesian Multilevel Model for fMRI Data Analysis

Description

plot.hcoef.post plots arrays of hierarchical coefficients.

Usage

## S3 method for class 'hcoef.post'
plot(x,spmname,spm, burnin=trunc(.1*R), nsamp=30, ...)

Arguments

x	betadraw object generated by the MCMC simulation
spmname	name associated with the thresholded voxels, e.g. "activated", "non-activated"
spm	threshold active, or non-active voxel, coefficients
burnin	n. of draws to burnin, def: .1*R
nsamp	number of random voxels to use in the plots (default: 30)
...	standard graphics parameters

Details

See description of similar function in bayesm::plot.bayesm.hcoef.

See Also

cudaMultireg.slice, pmeans.hcoef, regpostsim, post.simul.hist, post.simul.betadraw

Examples

## Not run: 
## load simulation
fsave <- paste(tempdir(),"/simultest1",fileext = ".sav", sep="")
load(file=fsave)
cat("loaded",fsave,"\n")
vreg <- 2
pmeans <- pmeans.hcoef(out$betadraw) 
px <- regpostsim(pmeans, vreg=vreg)
spma <- px$spma # active voxels
spmn <- px$spmn # non-active voxels
plot(out$betadraw,spmname="activated",spm=spma)
plot(out$betadraw,spmname="non-activated",spm=spmn)

## End(Not run)

cudaBayesreg documentation built on May 29, 2017, 6:19 p.m.