Clustering in KNIME

KNIME Analytics Platform
1

Hi, currently I’m doing clustering in KNIME. However, the results that I got from several nodes are different by each clusters. My objective is to limit each cluster to certain numbers so that the deviation between clusters would not be that much.

For example,
in k-medoids nodes, this is the results of each cluster given the input is distance from distance matrix using euclidian.

image

I want to limit each cluster results to range between 300-500.

Is there any nodes or workflow that could allow me perform this action in limiting clustering results? or even any code in Python that could be integrated in KNIME to achieve this results?

Thanks in advance!

2

Hi I have found this link

ClusterEqualSize.Rmd

It uses R, let me know if you need help with the workflow.

Cheers

Mau

1 Like
3

Hi @mauuuuu5

Thanks for responding!
I am always curious how to implement R/Python node in KNIME. Could you please help me with the shared link on how the workflow would look like on set of data to achieve equal size cluster?

Thanks in advance! :slight_smile:

4

Hi @mauuuuu5

Is there any workflow that I could learn from to implement the link you shared? :slight_smile:

5

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.