Clustering Analytics w/ Constraint

[marstf]

Hello everyone!

I'm clustering a CLIENT STORE NETWORK for a CONSUMER INDUSTRY with the purpose of routing (the logic is to do a Cluster-First, Route-Second approach; and I want to cluster the clients in 5 segments, one for each weekday).

I've already modelled the Clustering model in Alteryx, however I want to include a constraint: every cluster must have a total of 20% (could be 20% +/- 2%). The logic is to have ~20% volume in each of the 5 segments, to ensure a balanced volume for each weekday.

(1) Is it possible to do it with K-Centroid Cluster Analysis, and how?

(2) If not, what is a possible alternative?

Thank you!

Marcelo