Hello,

I am trying to weight observations when estimating growth models in Stata, and I am having trouble. The model isn’t converging and keeps saying “Iteration ##: log pseudolikelihood = -21377506 (backed up).”

As a simple version of the model I’m trying to run, say I want to run the below random slopes model:

mixed Trips Time || Zone: Time, covariance(unstructured)

I have annual data spanning 11 years for ~700 zones. I want to weight by the square root of dollars spent in a zone in a year (variable labeled ZoneWeight). I am not sure how I should enter this weight. I have played around with entering it as different types (fweight, pweight, etc.) before the two vertical lines, but I have issues with convergence. Can anyone please advise the appropriate way to weight by the dollars spent in a zone?

Thank you!