All,
I’m attempting to calculate the average silhouette for a hierarchical cluster analysis using ward’s method. I’m toying with R, since there appears to be a way to calculate it there. But I’m running against the wall of my R ignorance in that attempt. I thought about just doing the calculations for average silhouette in SPSS. I understand the calculation in theory. I need to find the dissimilarity between all points within a cluster, and all points between clusters. Using Euclidean distance I’m not clear how it is calculated when you have 8 variables in the cluster analysis. Maybe there’s a python or r extension that can do this?
Thanks
Matt