onlyHeterocyclics option for Machine Leaning Estimator Settings #1621
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Codecov Report
@@ Coverage Diff @@
## master #1621 +/- ##
==========================================
+ Coverage 41.88% 41.88% +<.01%
==========================================
Files 176 176
Lines 29354 29368 +14
Branches 6052 6059 +7
==========================================
+ Hits 12294 12301 +7
- Misses 16183 16188 +5
- Partials 877 879 +2
Continue to review full report at Codecov.
|
|
One comment before I review in detail. Since heterocyclics are a subset of cyclics, it seems to me like it would make more sense for
|
|
@mliu49 What you said makes more sense. I will make changes accordingly |
1 similar comment
|
@mliu49 What you said makes more sense. I will make changes accordingly |
yunsiechung
force-pushed
the
MLthermo_heterocyclic
branch
from
70eb9c8 to
8ccb7a0
Compare
mliu49
reviewed
Jun 19, 2019
| species will be estimated. ``onlyHeterocyclics`` means that only heterocyclic species will be estimated. Note that | ||
| if ``onlyHeterocyclics`` setting is set to True, the machine learning estimator is restricted to only heterocyclic | ||
| species regardless of the ``onlyCyclics`` setting. If ``onlyCyclics`` is False and ``onlyHeterocyclics`` is True, | ||
| RMG will log a warning that ``onlyCyclics`` should also be True and because heterocyclics are a subset of cyclics. |
There was a problem hiding this comment.
Slightly awkward wording "and because heterocyclics are a subset of cyclics"
Maybe "and the estimator will be restricted to heterocyclics because they are a subset of cyclics`?
rmgpy/data/thermoTest.py
Outdated
| @@ -407,6 +430,7 @@ def test_thermo_generation_ml_settings(self): | |||
| self.assertIsInstance(thermo, ThermoData) | |||
| self.assertTrue('ML Estimation' in thermo.comment, 'Thermo not from ML estimation, test purpose not fulfilled') | |||
|
|
|||
|
|
|||
There was a problem hiding this comment.
Could you remove this extra line?
|
Could you address my last two comments and rebase? I think this can be merged soon. |
'isHeterocyclic' function will return True if a molecule is heterocyclic or return False if not
3 unit tests for molecule.isHeterocyclic() were added. It checks whether it returns False for cyclohexanol and returns True for furan and pyridine
A new option 'onlyHeterocyclics' is added for Machine Learning Thermo Estimator. If 'onlyHeterocyclics' is True, the thermo of only heterocyclics species are estimated by mlEstimator regardless of 'onlyCyclics' setting. If 'onlyCyclics' is False and 'onlyHeterocyclics' is True, then RMG will log a warning that 'onlyCyclics' should be also True since heterocyclics is a subset of cyclics. I also added lines that will log a warning when 'onlyCyclics' is False and 'minCycleOverlap' is greater than zero since the machine learning estimator is restricted to only cyclic species when 'minCycleOverlap' is greater than zero. If 'onlyCyclics' is False and 'heteroCyclics' is True and 'minCycleOverlap' is greater than zero, it will log a warning and the machine learning estimator will be restricted to only heterocyclic species with the speficied minimum cycle overlap.
4 unit tests for 'onlyHeterocyclics' option in Machine Learning Thermo
Estimator are added.
Case 1: 'onlyCyclics' = True, 'onlyHeterocyclics' = True
test cyclic species gives None
Case 2: 'onlyCyclics' = True, 'onlyHeterocyclics' = True
test heterocyclic species gives ML thermo
Case 3: 'onlyCyclics' = False, 'onlyHeterocyclics' = True
test cyclics species gives None
Case 4: 'onlyCyclics' = False, 'onlyHeterocyclics' = True
test heterocyclic species gives ML thermo
yunsiechung
force-pushed
the
MLthermo_heterocyclic
branch
from
8ccb7a0 to
3e495db
Compare
|
@mliu49 I applied all the changes you have suggested :) |
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation or Problem
Adds one more option for Machine Learning Estimator settings to allow one to use machine learning for only heterocyclic species and group additivity for the rest
Description of Changes
onlyHeterocyclicsoption is added to Machinermgpy.rmg.input.mlEstimator, and a new methodisHeterocyclicis added tormgpy.molecule.molecule.py. Relevant changes are made in other rmg files including the documentation.If both
onlyCyclicsandonlyHeterocyclicsoptions are True, thenonlyCyclicsoption overrides, and all cyclic species thermo will be estimated by machine learning.Testing
3 unit tests are added to
rmgpy.molecule.moleculeTest.pyto check that species can be correctly identified as heterocyclic or not byisHeterocyclicmethod.3 unit tests are added to
rmgpy.data.thermoTest.pyto check that mlEstimator is applied for heterocyclic species for appropriate mlEstimator settings.Reviewer Tips
Check unit tests and run simple mlEstimators examples with heterocyclic settings turned on and off