intra_H_migration reaction marked as duplicate in the chemkin file

[xiaoruiDong]

I am using the latest version of RMG-Py to generate a model and afterwards when I try to create the 'Chemistry Set' in CHEMKIN using the file 'chem_annotated.inp', an error popped out, saying several reactions are marked duplicate but no duplicate be found. These reactions are all from intra_H_migration.

PS: version of RMG The current git HEAD for RMG-Py is: e32da33dd628c9386ffeffdf2e439a316c60387c Wed Feb 14 17:04:45 2018 -0500

The current git HEAD for RMG-database is: 3d68758a71da7e0f9cbeab08ffd94ed584f8b59b

chem_annotated.txt cycloP_gas.txt

[alongd]

Could you attach the 'chem_annotated.inp'?

[xiaoruiDong]

@alongd files are attached. The related reactioin # is 1731 7995 8476 8681 10011 (from cycloP_gas.out)

[mliu49]

I have a theory. This is not due to duplicate marking while saving the chemkin file, but rather the duplicate marking that is done as a result of the multiple TS changes. Following the multiple TS PR, additional duplicate marking is done in find_degenerate_reactions and checkForExistingReaction. However, those steps occur before the reaction is added to the model. As a result, two reactions (from two transition states, which is most commonly seen in intra_H_migration), will be marked as duplicate, but only one might make it into the core. Then, when saving the chemkin file, a non-duplicated reaction will have a duplicate flag.

Potential solutions:

• Reset duplicate flags when adding a reaction to the model
• Reset duplicate flags right before chemkin.markDuplicateReactions
• Check for single reactions with duplicate flags when marking duplicates
• Refactor multiple TS handling to not use the duplicate attribute

The first two are essentially equivalent, only differing in when it's done, and shouldn't add that much computational time. The third doesn't really accomplish much more than the first two, but might be more robust at the cost of computation time. The last might be the most difficult to implement, and may or may not be more efficient.

@goldmanm, what do you think?

Edit: I realized that it's odd that the duplicate reactions aren't added to the core together. However, I think this may be possible if pruning is on. @xiaoruiDong, were you using pruning?

[xiaoruiDong]

@mliu49 I double checked that I didn't use pruning.

[github-actions[bot]]

This issue is being automatically marked as stale because it has not received any interaction in the last 90 days. Please leave a comment if this is still a relevant issue, otherwise it will automatically be closed in 30 days.