Allow ARCSpecies to be initialized with a conformer list file #104
Conversation
alongd
added
Type: Feature
Type: UX
Codecov Report
@@ Coverage Diff @@
## master #104 +/- ##
==========================================
- Coverage 40.62% 40.55% -0.08%
==========================================
Files 22 22
Lines 5226 5265 +39
Branches 1357 1373 +16
==========================================
+ Hits 2123 2135 +12
- Misses 2760 2780 +20
- Partials 343 350 +7
Continue to review full report at Codecov.
|
There was a problem hiding this comment.
The code looks good. During testing, I find some minor issues:
[A]: The conformers_path argument works well with the ARCSpecies. However, if a user defines the species using an RMG object, such as:
spc = Species(molecule=[Molecule().fromSMILES("C=C[O]")], conformers_path = 'path')
Then a TypeError will occur:
__init__() got an unexpected keyword argument `conformers_path`
Possible solutions:
(1) add 'conformers_path' as a keyword argument to init in rmgpy/species.so
OR (2) add an exception in ARC to notify the users that they have to use ARCSpecies if they want to use conformers_path.
[B]: Currently an AttributeError will occur (related to PR 106) if a user specifies conformers_path as an argument without specifying a smile/adj list. For instance:
spc = ARCSpecies(label='methylamine', conformers_path = 'path')
Will give:
'NoneType' object has no attribute 'enumerate_bonds'
But the following will work
spc = ARCSpecies(label='methylamine', conformers_path = 'path', **smiles = 'CN'**)
Suggestion: we can temporarily get around with this issue by specifying smiles/adj list along with the conformers_path. We should address the issue raised in PR 106, and also figure out why from_xyz does not work as expected.
Many thanks for your contribution. This feature is really useful. Please let me know if you have any questions. Great work 👍
|
Thanks! rmg_spc = Species(molecule=[Molecule().fromSMILES("C=C[O]")])
arc_spc = ARCSpecies(rmg_species=rmg_spc, conformers_path='path/to/conformers.txt')I wouldn't want to add Will #106 solve point B? |
Instead of duplicate code for saving the 'before' and 'after' conformer files
Also tests the Scheduler.save_conformers_file() method Also made the required changes to schedulerTests
Don't generate coformers if their xyzs are available Don't calculate conformers energies if energies already given
To differentiate between cases where the energies are given in a `conformers_after_optimization.txt` file with all energies set to 0. Also set Species.opt_level default value to None
Users can now initialize an ARCSpecies object with either of the
conformers_before_optimization.txtorconformers_after_optimization.txtfiles. These files are automatically generated by ARC and saved in the Species 'geometry' folder. The latter also reports the conformer energies.