Open billdenney opened 3 years ago
Thanks for reporting this. PsN doesn't support this currently.
Hi Rikard, We are testing PsN5 integration with NM750 at the moment. Is a patched version available that is fully compatible with all NM750 features?
Thanks @sundav05 for the interest. The answer is no and it is not in the current plan. We are too few developers to be able to prioritize this at the moment. Most tools of PsN are still expected to continue working with NONMEM 7.5 though.
That being said a lot of development is currently happening in Pharmpy (https://github.com/pharmpy/pharmpy), which is used more and more by PsN and is also compatible with R via https://github.om/pharmpy/pharmr. The particular feature of this issue is not supported by Pharmpy yet though.
If you are interested in support for particular features of NONMEM 7.5 that you know doesn't work it would be useful if you could file separate issues for these. Or even better you could join the development. Let us know if there is some particular feature you would like to work on and we will be more than happy to help you get started.
Understand completely. We will test specific features of interest to us and get back to you if we run into issues. In the meantime, we noticed that PsN is force adding the [NODES}=1 argument to nonmem. This is counterproductive since we use wrappers for launching nonmem that automatically sets the number of NODES to the same as the number of cores requested by users. If NODES value is NOT set, nonmem defaults to a single node anyway and serves NO addl. purpose. Does PsN50 use this setting explicitly for anything internal as PsN490 does not do this.
When using model abbreviations available in NONMEM 7.5 like in pheno_nm750.mod.txt, I get an error with execute:
This appears to be an issue with the parser where it is grouping the
CL=
into the numeric value rather than outside.