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] Parameter coverage iPython notebook #419
Conversation
This pull request introduces 1 alert and fixes 1 when merging a40dca1 into fe87bae - view on LGTM.com new alerts:
fixed alerts:
|
Codecov Report
@@ Coverage Diff @@
## master #419 +/- ##
=======================================
Coverage 49.25% 49.25%
=======================================
Files 35 35
Lines 9577 9577
=======================================
Hits 4717 4717
Misses 4860 4860 Continue to review full report at Codecov.
|
Closed and reopened -- Trying to get travis to resume building this PR |
This pull request introduces 1 alert and fixes 1 when merging 501a1d8 into fe87bae - view on LGTM.com new alerts:
fixed alerts:
|
Summary
This PR adds a parameter coverage checker hosted in an iPython notebook. This notebook produces a detailed report of which chemical groups a ForceField object is unable to parameterize. These chemistries are reported as
These results are printed out in the iPython notebook itself, and the final cells provide functionality to write to file as well.
My hope is that the Open Force Field industry partners can use this tool on a small dataset which encapsulates the diversity of the compounds they might simulate, and report critical areas where our force fields lack coverage. Industry partners probably can not reveal whole molecules from internal data sets, so I have tried to design this notebook such that it is easy to take screenshots of relevant chemistry, without revealing the whole molecule.
This notebook is suitable for analysis of hundreds to thousands of molecules, but probably not much more than that. It takes about 1 core-second per molecule, though I do offer a parallelized checker as well. Significant speedups past 0.1 sec/molecule would require several additional days of engineering, affect internal OFF Toolkit code, and decrease code readability. Thus, while we probably want to make a "fast" version of this some day, it would not be in the form of an interactive notebook.
Checklist
MissingValenceParameterException
to also contain relevantTopologyAtom
s for exception handlingStatus