Hemispheric validation, right vs. left
parky
Fri Mar 24 10:20:50 2017
library(mbstructure)
data(MBconnectome)
suppressMessages(library(mclust))
out <- generate.graph(newrdat, vdf.right)
g <- out$g
vdf <- out$vdf
dmax <- 50
Xhat <- doEmbed(g, dmax, plotElbow = FALSE)
Kmax <- 19
mc <- Mclust(Xhat, 2:Kmax)
vdf$cluster <- factor(mc$class)
Xhat4 <- Xhat[vdf$type=="KC",]
out100 <- semiparYQ(Xhat4, K=100, verbose=FALSE)
right.vs.left(Xhat, vdf, out100)
