Question

I'm following the technique described here to produce a biplot of kmeans cluster analysis.

I would like to know what measures the plot uses (as Dim1 and Dim2)

I've searched, and read the R documentation, but can't find an answer

Was it helpful?

Solution

The dune data has 30 variables. You can calculate distances between each datapoint in 30D space but you can't visualize it. cmdscale takes the distance data calculated on the 30 variables and tries to represent the data in, for example, 2D. So you can think of this approach as dimensionality reduction, similar in spirit to Factor analysis, PCA, etc.

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top