Is the following possible? Came off a stats forum so hope this is clear
enough.
=========
Plot a 90% confidence intervals of a Kmeans-cluster analysis into a
scatterplot.
I've got a scatterplot of PCA results (first and second factor). On the same
dataset (which are measurements) I'm running the Kmeans cluster analysis.
I'm naming the point scatters of the PCA-plot after the analyzed clusters.
Into this mixed plot I would like to add the 90% confidence intervals (which
should come up as a ellipse) of the Kmeans cluster centroids.
==========
Tks
Locked
