Predict Hierarchical Cluster Error

Hi everybody, I am trying to make a cluster prediction using the R package named clue, but when trying to make the prediction I cannot do that when I am using the Hierarchical cluster option "hclust".

The error says 

Execute failed: Error in R code: "Error: no applicable method for 'predict' applied to an object of class "hclust""

Is interesting because I can do that with the Kmeans. By reading the documentation of the clue package it says that it can predict "an object of class "cl_class_ids" with the predicted class ids, or of class "cl_membership" with the matrix of predicted membership values.", but I does not seem the case for the hclust algo.  See the image.

I am ataching the workflow

Any help wll be appreciated



Well I am answering my own question and according to this response "Some of the clustering algorithms (like those centroid based - kmeans, kmedians etc.) can "label" new instance based on the model created. Unfortunately hierarchical clustering is not one of them - it does not partition the input space, it just "connects" some of the objects given during clustering, so you cannot assign the new point to this model." see:

But atill you can make such prediction using the Weka nodes but not the Knime nodes. 

Best Regards


Hi mau,

Thanks for letting us know why it doesn't work in R. Still, it's interesting that you could do it in Weka. Do you know if they do something different?