search

tensfeldt / openNCA

openNCA computation engine is an R package that provides for generation of pharmacokinetic parameter estimates using non-compartmental (NCA) pharmacokinetic analysis methods.
MIT License
27 stars 5 forks source link

manual fix NAMESPACE #301

Open roninsightrx opened 6 months ago

roninsightrx commented 6 months ago

The current master branch of tensfeldt/openNCA cannot be installed in recent R versions (testing with 4.4.0) due to a mismatch between man files and code. Best would be to rerun devtools::document() for this package with a recent version of roxygen, but wanted to keep this fix minimal.

TomTensfeldt commented 6 months ago

The current master branch of tensfeldt/openNCA cannot be installed in recent R versions (testing with 4.4.0) due to a mismatch between man files and code. Best would be to rerun devtools::document() for this package with a recent version of roxygen, but wanted to keep this fix minimal.

Note that I’ll have to review. This version of openNCA and the entire commit history is subject to update and change to bring the commit history in line with what was used in serial releases previously. An update to the main/master branch is in progress to align this and update the code rather significantly. Let’s discuss if your needs are more urgent, perhaps a near term separate branch for a while would work.

roninsightrx commented 6 months ago

Thanks Tom, no problem. I did indeed realize as well that the current master is not able to run the test cases. It does seem to run fine with the specific "qualified" archived package (c3d3f48) mentioned in the install scripts, so I'll work off of that version for now. That hash doesn't seem to map to a specific commit hash though? Could you point me to which commit in the commit history it matches to?

TomTensfeldt commented 6 months ago

Thanks Tom, no problem. I did indeed realize as well that the current master is not able to run the test cases. It does seem to run fine with the specific "qualified" archived package (c3d3f48) mentioned in the install scripts, so I'll work off of that version for now. That hash doesn't seem to map to a specific commit hash though? Could you point me to which commit in the commit history it matches to?

Ron, I’ll have to check if that’s even possible. Unfortunately, we had a big diversion from the GitHub repo and I’m in the process of completely rehashing the code to a more modern, maintainable state. I hope to, probably awkwardly, retain the commit history for, I think, 4 main “internal” releases and then go from there. The new release will be modernized significantly, removing a long of duplication in the code and probably relaxing the current focus on maintaining a pure BASE R approach. I’m really pleased that there’s any interest at all so want to work with you or support you in whatever way I can.

I’m been distracted on building out the “app” that supports this code for quite a while since that is being changed dramatically to refocus it as a portable (webR) based web framework runnable everywhere rather than an electron app with its overhead. If you have any interest in that, let me know. We WILL be releasing that in a separate repo here in Github but are working currently in a self hosted repo until it’s ready for broader exposure testing. If you want to discuss any of this in more detail, let me know. I’m open to suggestions. I’ll give you my direct email if you’d like to do so. I’m eager to hear about what you’re using this for and how it could evolve to help your use cases as well.

roninsightrx commented 6 months ago

I see, thanks for clarifying. Would indeed be good if we can discuss in more detail offline, I can share the use case in more detail. My email is ron[at]insight-rx.com.