Add "negative_A" option if required in cti files #68
Conversation
If a reaction has a negative pre-exponential factor, it must have a negative_A option in the cti file. Fixes Niemeyer-Research-Group#67
pymars/soln2cti.py
Outdated
| @@ -432,7 +432,10 @@ def write(solution, output_filename='', path=''): | |||
| raise NotImplementedError(f'Unsupported reaction type: {type(reaction)}') | |||
|
|
|||
| if reaction.duplicate: | |||
| reaction_string += ',\n' + indent[8] + 'options = "duplicate"' | |||
| if getattr(reaction, 'allow_negative_pre_exponential_factor', default=False): | |||
There was a problem hiding this comment.
Remove the default= so this line is
if getattr(reaction, 'allow_negative_pre_exponential_factor', False):
There was a problem hiding this comment.
I already made that change on my local version of the solution file. I'll adjust it here on the pull request as well.
Removed the `default=False` and changed it to `False` since it gives an attribute type error.
Codecov Report
@@ Coverage Diff @@
## master #68 +/- ##
==========================================
- Coverage 66.98% 66.87% -0.11%
==========================================
Files 12 12
Lines 1266 1268 +2
Branches 294 295 +1
==========================================
Hits 848 848
- Misses 316 317 +1
- Partials 102 103 +1
Continue to review full report at Codecov.
|
|
That is implicit in my code, and I think it's a safe assumption: An overall net negative rate coefficient is forbidden by Cantera (and physically insane), so you must always have some duplicate reaction so the sum adds up to a positive number. |
If a reaction has a negative pre-exponential factor, it must have a negative_A option in the cti file.
Fixes #67
Suggested by @rwest
Not yet properly tested