I'm trying to do some cluster analysis of local neighbourhood by proximity, but I need to use a maximum size for the stopping rule for the cluster generation rather than a fixed number. There around around 7000 neighbourhood in my dataset and for each I have their X/Y co-ordinate and and the population (which ranges from 2 to 30). I need to cluster so that there are a minimum of 50 people in each cluster but no more than 100.
Is this something that anyone has come across before and can it be done in Stata (I'm using v16).
Many thanks for any help
Rob
Is this something that anyone has come across before and can it be done in Stata (I'm using v16).
Many thanks for any help
Rob
