tests/test01-distance.R

In ClassDiscovery: Classes and Methods for "Class Discovery" with Microarrays or Proteomics

library(ClassDiscovery)
suppressWarnings( RNGversion("3.5.3") )
set.seed(593996)
dd <- matrix(rnorm(100*5, rnorm(100)), nrow=100, ncol=5)
distanceMatrix(dd, 'pearson')
distanceMatrix(dd, 'euclid')
distanceMatrix(dd, 'sqrt')
distanceMatrix(dd, 'weird')
distanceMatrix(dd, 'cosine')
rm(dd) # cleanup

# simulate data from three different groups
d1 <- matrix(rnorm(100*10, rnorm(100, 0.5)), nrow=100, ncol=10, byrow=FALSE)
d2 <- matrix(rnorm(100*10, rnorm(100, 0.5)), nrow=100, ncol=10, byrow=FALSE)
d3 <- matrix(rnorm(100*10, rnorm(100, 0.5)), nrow=100, ncol=10, byrow=FALSE)
dd <- cbind(d1, d2, d3)

# perform hierarchical clustering using correlation
hc <- hclust(distanceMatrix(dd, 'pearson'), method='average')
cols <- rep(c('red', 'green', 'blue'), each=10)
labs <- paste('X', 1:30, sep='')

# plot the dendrogram with color-coded groups
plotColoredClusters(hc, labs=labs, cols=cols)

#cleanup
rm(d1, d2, d3, dd, hc, cols, labs)