I have recently started diving into PML. At the moment I am trying to write PML code in which the Clearance (1-cmt) is scaled by the indidivual’s body weight with body weight being a parameter in the model code. However, I don’t know how to setup a PML variable that is mapped to the body weight in my dataset.
Do I have to use fcovariate(wt) for example to make the body weight “visible” in the PML code? But then I am not sure if Phoenix really handles it as some generic variable or as a covariate which might lead to diverse issues?!
i think fcovariate(wt) will be enought to make wt ‘visible’ to the PML code without automatically causing other associatiions. I am wondering though why you don’t want to try wt as a covariate on that parameter from the built in options? it sounds like you could be making things more complicated than they need to be…
agree with Simon, please take a look at our recent PML School webinar where we showed how to use the built-in functions in Phoenix to setup covariates:
I am trying to bring a NONMEM model to Phoenix and in this NONMEM model the clearance was scaled with body weight. I guess I will try fcovariate(wt) but I will also test the covariate setup and compare the model performances with the original NONMEM model.
When I use fcovariate(wt), can I be sure that Phoenix will use the correct body weight for each individual? (Don’t know exactly how the covariates are handled).