how to have a fix number of data within clusters when use hierarchical cluster

Question

Hello all,

I am using the hierarchical cluster to group similar modules now. My script is below:

dt = current data table();
obj = Hierarchical Cluster(
Y( :A, :B),
Method( "Ward" ),
Standardize Data( 1 ),
by(:SampleDescription),
Number of Clusters( 6 ),
color clusters(1),
Columns Scaled Individually(0)
);
obj << Declutter(1,1);
obj << Go;

However, I really want to set up a fixed number of models in each group. Does anyone know could jmp achieve it?

Thanks a lot,

Wen

Chris_Kirchberg · Answer

Do you mean 5 members per cluster? So that each cluster has exactly 5 members?

If so, I am not familiar with any clustering method that lets you define/limit the number of members in each cluster. Only the number of clusters can be defined/limited in JMP.

unknown · Answer

Thank you for replying me. I was wondering if I could set up a number, for example 5, so JMP could generate clusters that have 5 models in each. I guess there is no way that JMP can do it. But thank you for reply any way.

Thanks

gzmorgan0 · Answer

Wen,

Your question is not clear to me.  You have a column called :SampleDescription so you will get a cluster analysis for each group defined by :SampleDescription. I do not understand the statement "fix number of models in each group."