This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
I've run cluster analyses a number of times, but have never been asked this before....so I did some digging and could not find anything in the Alteryx toolset or on community that would help me answer it.
I understand how to run a cluster analysis and then append clusters with the tools shown below...I can see the average distance within cluster in the typical output. But how to get the distance from each of the k-centroids to each of the members of each cluster? For context I had grouped some store locations, and then someone asked me which stores are in the center of each cluster and which ones on the fringe.
Thank you and I think that would work if I had done the cluster analysis based on latitude/longitude or spatial position. However this is based on various factors including demographics, store performance, competitive factors etc. Is there a way to get the relative 'coordinates' for each observation based on these factors? Thanks!
@ggruccio Thank you for explaining that. The K - Centroids Cluster Analysis tool actually outputs cluster centroids. There are Two fields in the O Output: grp and out. There are a few headers, but if you filter on the Clus_Centers value for the grp field, then you should get a pipe delineated list of cluster centers for each parameter. There is also Clus_Info which gives some stats about the cluster itself. With that and the output of the Append Cluster tool, you should be able to get a lot of insight.
The Find Nearest Neighbors tool may provide what you're looking for. You'll need to grab the cluster assignments to link them with your store locations but you can find the "distance" between them. Also, including the Interactive Lesson in case you want more info!