I am trying to fit a model where the response is a mixture of
multinomial distributions using the NLMIXED procedure. In order to
compute the likelihood for this model, I need to specify the response
to be the number of times each possible category is observed in
repeated sampling. As a result, my response is a vector which has length
equal to the number of possible outcomes allowed by the multinomial
Is there any way to specify my response variable to be a vector using
PROC NLMIXED, or does it have to be univariate?