New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[WIP] Minimal working, automated and tested examples #844
Conversation
|
Ok, I noticed that sometimes I get an openmm error as follows:
I have the feeling has something to do with the platform, I have forced it to 'CPU'. Any ideas why this may be happening or if you have experienced similar issues? |
@mikemhenry Well, tests are failing and no idea why, haha. Maybe it wasn't such a great idea to remove all of this? Though I really cannot tell why the tests are failing that way, it works in my computer and in the HPC system. |
Locally I get these errors as well, for example:
Which looks like it was unable to find the This is how I run the tests locally to simulate being on GHA: |
@mikemhenry Thanks for letting me know how to run these tests simulating GHA. By the way, I think you probably have a dirty |
Good catch, I didn't clean that directory! The danger of dirty/untracked files and globbing :) |
Ok, so at least I made the > cwd = os.getcwd()
E FileNotFoundError: [Errno 2] No such file or directory Doesn't make much sense to me, but maybe something not getting cleaned up properly when the examples test is run with subprocess? |
@mikemhenry If you think about it, it makes sense. Since it is complaining now that it is not finding any neighboring waters (because vacuum). And it takes any word different from "vacuum" for
phase keyword (I guess not putting anything still works).
|
I think we want a vacuum since that will be faster. |
@mikemhenry please do note that the test has been running without vacuum this whole time. If we really need vacuum, I think the test has to be done differently, because I believe the assertion in perses/perses/rjmc/topology_proposal.py Line 586 in de5397e
|
Ah good catch! It must have been running in the solvent phase since vacuum was spelled wrong. We should probably just change the test case to a different mutant residue (i.e. instead of mutating from ALA to ASP, mutate from ALA to THR (or any non-charged amino acid). We won't be able to do charge changing transformations in vacuum because charge changing mutations require transforming water into an ion to account for the charge change. In vacuum, there are no waters present to allow for this. |
@zhang-ivy Yes, that sounds good to me. I'll be making the changes and hopefully that will do it for this PR. |
One of the tests is failing but for a whole new different reason. There is definitely a syntax error here, that @jchodera Do we want that to be an actual |
@ijpulidos it would be nice to remove that stop and just see what the error is, that would help us figure out what we want to do here |
@mikemhenry So yeah, I was afraid this would happen. It is not hitting the exception/error now. It is something that does not happen every time, apparently. I guess we can merge this and raise an issue separately for it. |
We should catch it on our nightly CI eventually -- let's wait until we hear from @jchodera with what his intention was with |
🎉 great work @ijpulidos Having examples that we know will work will make testing things (as well as brining new users to |
@ijpulidos looks good, auto merge is on! |
Whoops! I must have put that in accidentally as part of debugging. The |
Description
Removing old and obsolete examples. Leaving current atom-mapping notebook and cleaning up the protein mutation non-equilibrium switching API example.
Still missing protein-ligand and only ligand examples, for both CLI and API.
Motivation and context
Resolves #813
How has this been tested?
Just local tests so far (manually checking it runs and it gives reasonable output), need to write automated tests for this.
Change log