It looks like method in pumas_be has been removed. Could you explain a bit about the change?
@donaldlee3 can you elaborate what method in pumas_be are you referring to?
As part of the new release, a lot of new features focused on replicated crossover designs, especially for highly variable drugs and narrow therapeutic index drugs. One of the changes as part of the release was to use a linear model for non-replicated crossover designs and linear mixed effects for replicated crossover designs (i.e., the default behavior in the past release). We didn’t find any regulatory guidance that suggested the use of linear models for replicated crossover designs as a convergence of guidance suggested to analyze these through linear mixed models.
Are you trying to analyze a replicated crossover design using a linear model as opposed to a linear mixed model?
Thank you for the explanation @Nosferican.
I used to use method = :lmm, and noticed it wasn’t supported anymore. I was just curious to see how it was being implemented with the update.