October 2017
Beginner to intermediate
572 pages
26h 1m
English
Another visualization method is to present an MDS object as a graph. A sample code is listed here:
> library(igraph) > swiss.sample = swiss[1:10,] > g = graph.full(nrow(swiss.sample)) > V(g)$label = rownames(swiss.sample) > layout = layout.mds(g, dist = as.matrix(dist(swiss.sample))) > plot(g, layout = layout, vertex.size = 3)

You can also compare differences between the generated results from MDS and PCA. You can compare their differences by drawing the projected dimensions on the same scatter plot. If you use a Euclidean distance on MDS, the projected dimensions are exactly the same as ...
Read now
Unlock full access