tests/test-05recolor.R

In BinaryMatrix: Clustering and Visualizing Binary Matrices

library("BinaryMatrix")
data("CML500")

# remove extra identical feature vectors
CML500 <- removeDuplicateFeatures(CML500)
# jacc, pear, manh, euc
vis1 <- DistanceVis(CML500, "jacc", "mds", K=8)
vis2 <- DistanceVis(CML500, "sokal", "mds", K=8)
vis3 <- remapColors(vis1, vis2)

# par(mfrow=c(1,3),  cex=1.5)
plot(vis1@view[[1]], col=vis1@colv, pch=vis1@symv)
plot(vis2@view[[1]], col=vis2@colv, pch=vis2@symv)
plot(vis3@view[[1]], col=vis3@colv, pch=vis3@symv)

A <- recoverCluster(vis2)
B <- recoverCluster(vis3)
table(A, B)
X <- recoverCluster(vis1)
table(A, X)
table(B, X)

library(cluster)
clus <- pam(vis1@distance, k = 12, diss=TRUE, cluster.only=TRUE)
vis4 <- recolor(vis1, clus)
plot(vis4@view[[1]], col=vis4@colv, pch=vis4@symv)