ReactionMechanismGenerator / T3

The Tandem Tool (T3) for automated chemical kinetic model development
https://reactionmechanismgenerator.github.io/T3/
MIT License
11 stars 4 forks source link

T3 should also store the final libs in the output folder #30

Open alongd opened 4 years ago

alongd commented 4 years ago

Useful if T3 is run on a server, and the user would like to copy the thermo/kinetics library to their repo w/o digging too much on the server

mefuller commented 3 years ago

If I understand this right, it's just adding a copy operation when execution terminates (pseudocode follows, the BASH stuff would be implemented in python):

T3.project_directory= $PROJDIR
{read folder names to get $LAST_ITER}
cp $PROJDIR/$LAST_ITER/RMG/cantera/*.cti $PROJDIR
cp $PROJDIR/$LAST_ITER/RMG/chemkin/chem.inp $PROJDIR
cp $PROJDIR/$LAST_ITER/RMG/chemkin/chem_annotated $PROJDIR
cp $PROJDIR/$LAST_ITER/RMG/chemkin/tran.dat $PROJDIR
cp $PROJDIR/$LAST_ITER/RMG/chemkin/species_dictionary.txt $PROJDIR

Is that what you had in mind?

alongd commented 3 years ago

Your suggestion addresses an important issue of storing the final model (Chemkin, Cantera and RMS) in a designated output folder. It's currently being stored under PROJDIR/iteration_x/RMG/<format>/, and we'd like to copied into something like PROJDIR/output/model/ (I don't think we opened a feature request for that).

In addition, this particular feature request asks that we store the RMG libraries T3 creates with the computations it did throughout all iterations of a specific T3 run. The current situation is that T3 looks for the RMG-database repo location on the machine it runs on, opens a new library there if needed and stores thermo and rate coefficients there for the RMG run of the next iteration could use it.

I think we'd like that it'll also open another new library file under PROJDIR/output/libraries, and store the data calculated during this T3 run there as well. Independently, it should also, at the end of the run, copy the RMG library it appended to (which might be more comprehensive than the computations it made in this run, since previous T3 runs might have also wrote into it) over to PROJDIR/output/libraries/project.

I prefer the implementation to be done in Python, we already have a dictionary with paths to library destinations that we should use (and in this case, extend)