Unsupervised Classification runs on wrong number of classes?

ScottBarron — Tue, 06 Aug 2013 17:01:04 GMT

Hi,
So my problem is that I am running an unsupervised classification, and for some reason it is not running on the correct number of classes. If I try to run a 12 class classification, it returns only 10 classes. I have no idea why this is happening, as it runs correctly for a different set of data.

#Unsupervised Classification
    Classified= Divided
    Output_signature_file= "Scrap.gsg"
    outUnsupervised = IsoClusterUnsupervisedClassification(Filtered, 12, 20, 10, Output_signature_file)
    outUnsupervised.save(Classified)
    arcpy.AddMessage("Unsupervised classification complete.")

Advice?

Re: Unsupervised Classification runs on wrong number of classes?

curtvprice — Wed, 07 Aug 2013 03:43:56 GMT

Hi,
So my problem is that I am running an unsupervised classification, and for some reason it is not running on the correct number of classes. If I try to run a 12 class classification, it returns only 10 classes. I have no idea why this is happening, as it runs correctly for a different set of data.

I would look closely at your dataset. It's possible it your data distribution does not allow it to separate into 12 discernible groups. The help article How Iso Cluster Works discusses this:

The specified Number of classes value is the maximum number of clusters that can result from the clustering process. However, the number of clusters in the output signature file may not be the same as the number specified for the number of classes.

Have you tried running the output signatures through the Dendrogram tool? It may shed some light.

topic Unsupervised Classification runs on wrong number of classes? in Geoprocessing Questions

Unsupervised Classification runs on wrong number of classes?

Re: Unsupervised Classification runs on wrong number of classes?