i.cluster

[DCertain]

I'm experiencing some problems with the i.cluster routine in 4.1.  I'm using a
mask to define a subset of an MSS image for clustering and classification.
However, when I attempt to create my signature file I consistently get 100%
convergence and only 1 signature.  It doesn't matter how many initial clusters
that I request or the minimum seperation value (have set as low as 0.1).  I'm
actually clustering on pixels that clustered seperately on an earlier
clustering run (wow, that's alot of clustering!) so it's not that the pixel
values are that similar.  Any suggestions as to why I'm only getting 1
signature instead of several?  Please reply to dcertain at uokmvsa.backbone.uoknor
.edu .

D. Certain
Oklahoma Biological Survey