In the KMeans Clustering, there are 3 options : Single Step, Use within-cluster std deviations and Shift distances using sampling rates. I try with a simple example and the result with no of these options gives me the best result considering the CCC value. I am asking if I can proced my statistical analysis without any option OR I have to choose one ?
You donot have to select an option in order to run the kmeans clustering. That is, if none is selected, JMP will proceed with its default settings: scale the distances by an overall std deviation (versus within-cluster std deviations), no adjustment applied to the distances per the sizes of clusters, and fit the clusters automatically (versus manually step through iterations one at a time.).