Create clusters in PCA plot and use those clusters as data groups

[nuno-agostinho]

These clusters could then be used in survival curves and other analyses.

• [ ] Separate based on user-created line (maybe validated with Fisher's exact test)
• [x] Separate by k-means / PAM / CLARA
• [ ] Assist methods by showing optimal number of clusters (using total within sum of squares, silhouette width and gap statistics)
• [ ] Validate with cluster validity methods
• [x] Plot clusters in PCA using the polygon series
• [x] Create groups based on clusters

See also cluster analysis in R (1 and 2) and cluster validation using cIVALID

[nuno-agostinho]

Plot convex hulls from given k-means:

library(cluster)
X <- ruspini
pca <- prcomp(ruspini)
pca <- pca$x[ , 1:2]

# Cluster
pamx <- pam(pca, 5)
cluster <- pamx$cluster

# Plot
plot(pca)
for (each in unique(cluster)) {
    segment <- pca[cluster == each, ]
    hpts <- chull(segment)
    polygon(segment[hpts, ], border=each + 1)
}