I am attaching a project I made that takes advantage of our unique transit statement. To make sure fraction is between 0 and 1, I used the inverse logit function(ilogit).
I recreated similar conditions than in the paper. The advantage is that if you simulate data , you can fit the model to these data and optimize the number of transit compartments nd it will work with both single and multiple doses.
Let me know if all is clear. Look at the pml user guide for the syntax.
I updated the project just in case the one I sent just before would not work. Click on initial estimates and you will see the impact of both ktr and ntr.