Dear NMusers, How many of NMusers had to look at the sources routine during their daily PK/PD modeling? And Why ?
In the past I had to look at the source code to decipher vague errors that I could not understand until I went to the fortran routine and see at which circumstances this error is issued. Other instances where I looked at the source code is when: I needed to see which commons have the correlation matrix/ the R/ RINV, S matrices...to be able to output these with more precision. (Happily, NMVII will allow more control on the outputs precision). Also I wanted to look at some undocumented features to be able to use them: e.g: the PHI function was previously available but undocumented. Also to see more details about the numerical method used to approximate the phi function... In the nonmem help many entries ask the interested user to study the NM code to understand and see how can use these advanced features. With the encryption of the source code, can we expect more extensive and complete documentation and flexible user controls? Bests, Samer Mouksassi
