Skip to content
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

Fixes and improvements for SPAM and volmap, 3d grids #838

Merged
merged 7 commits into from
Aug 22, 2020

Conversation

drroe
Copy link
Contributor

@drroe drroe commented Aug 21, 2020

This PR has multiple fixes and improvements related to SPAM, volmap and reading/writing 3D data.

  • The online help and manual documentation are greatly improved for both.
  • The volmap command now creates a data set for the peaks data. This means that among other things you can run spam and volmap in the same cpptraj run. This could include running volmap several times with the same peak data and different settings.
  • Some checks have been introduced into spam to trap "bad" peaks (poor occupancy, extremely large energies, etc) that previously could lead to the code producing nonsense or failing outright.
  • OpenDX data can now be read in single (the previous behavior) or double precision if desired.
  • OpenDX output now respects if the format of a data set has been changed (via e.g. dataset precision) from the default of %g.
  • Enable divide and multiply for 3D datasets.

Miscellaneous fixes:

  • Improve Tinker format recognition.

drroe added 7 commits August 21, 2020 13:29
…for OpenDX files, option to read grids in as double precision, enable DIV and MULT for grids, warn if extra args remain after 'purewater' specified for SPAM, use DataSet/DataFile framework for peaks in SPAM, add DataIO for read and write of peaks data, fix Tinker file recognition, make sure SPAM is not hidden, skip peaks where bin calc could overflow, avoid using zero bandwidth for KDE hist when only 1 value available.
…p to be compiled double precision if desired. Fix some uninitialized vars.
@AmberJenkins
Copy link
Collaborator

The PGI build in Jenkins failed.

@drroe
Copy link
Contributor Author

drroe commented Aug 21, 2020

Weird - Jenkins says PGI failed again, but it clearly passed.

@drroe
Copy link
Contributor Author

drroe commented Aug 21, 2020

Travis-CI mac framework is having issues:

Error: Failure while executing; `tar xof /Users/travis/Library/Caches/Homebrew/downloads/9e4ec3ec2b678643fc676aad73df9fd9e838a4d764ecba365714e1e7f5852631--nss-3.56.high_sierra.bottle.tar.gz -C /var/folders/nz/vv4_9tw56nv9k3tkvyszvwg80000gn/T/d20200821-83344-mhy1fh` exited with 1. Here's the output:

tar: Error opening archive: Failed to open '/Users/travis/Library/Caches/Homebrew/downloads/9e4ec3ec2b678643fc676aad73df9fd9e838a4d764ecba365714e1e7f5852631--nss-3.56.high_sierra.bottle.tar.gz'

I'll try to restart in a bit I guess...

@drroe
Copy link
Contributor Author

drroe commented Aug 21, 2020

😠 The Mac stuff is still failing

@drroe
Copy link
Contributor Author

drroe commented Aug 22, 2020

OK, the real error is a few lines above:

==> Downloading https://homebrew.bintray.com/bottles/nss-3.56.high_sierra.bottle

curl: (22) The requested URL returned error: 404 Not Found

Error: Failed to download resource "nss"

Download failed: https://homebrew.bintray.com/bottles/nss-3.56.high_sierra.bottle.tar.gz

Warning: Bottle installation failed: building from source.

Indeed, trying that URL results in a The requested path was not found. So right now I'm stuck.

@drroe
Copy link
Contributor Author

drroe commented Aug 22, 2020

Looks like Mac builds on Travis are finally working again!

@drroe drroe merged commit 1cf93d3 into Amber-MD:master Aug 22, 2020
@drroe drroe deleted the fixspam2 branch August 22, 2020 18:39
@swails
Copy link
Contributor

swails commented Aug 22, 2020

Weird - Jenkins says PGI failed again, but it clearly passed.

It actually didn't pass. The problem is that if that stage itself failed, then the entire pipeline would fail (which we don't want). If you click on the green PGI stage in the Blue Ocean UI, you will see the individual steps, and the build/test step is red:

PGI Failure

The PR post is placed in the catch part of a try/catch control block in the Jenkinsfile, so it's only accessible when the steps before it fail.

The whole point behind having the pipeline post to a PR when the PGI build fails is so we can be notified if it didn't pass without making the whole Jenkins pipeline fail (which would block it from building the libcpptraj docker image and testing pytraj against it).

As an aside, the latest master build doesn't work with pytraj it seems. @hainm, are you aware of this?

@hainm
Copy link
Contributor

hainm commented Aug 23, 2020

As an aside, the latest master build doesn't work with pytraj it seems. @hainm, are you aware of this?

The error is here: FAILED test_analysis/test_jcoupling.py::test_jcoupling - ValueError
And I think @drroe already fixed that in Amber-MD/pytraj#1547

@drroe
Copy link
Contributor Author

drroe commented Aug 24, 2020

It actually didn't pass.

@swails you're 100% right! Thanks - I think I've found the issue, should be fixed by #839 (I hope).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants