Fix duplicate degeneracy multiplication and add training rules to kinetics families

[Copilot]

Problem

The rmg_kinetics.py script had two issues affecting the accuracy of kinetic rate coefficient calculations:

1. Double multiplication of the A factor by degeneracy: The code was multiplying the rate coefficient by degeneracy twice - once internally in family.get_kinetics() (which accepts a degeneracy parameter) and again explicitly via kinetics.change_rate(deg_rxn.degeneracy). This resulted in incorrect rate coefficients.

2. Missing training rules incorporation: When loading the RMG database, training rules were not being added to kinetics families, which could result in incomplete kinetic data.

Changes

1. Removed duplicate degeneracy multiplication (line 101)

# Before
kinetics_list = family.get_kinetics(reaction=deg_rxn, template_labels=deg_rxn.template, degeneracy=deg_rxn.degeneracy)
for kinetics_detailes in kinetics_list:
    kinetics = kinetics_detailes[0]
    kinetics.change_rate(deg_rxn.degeneracy)  # ❌ Double multiplication!
    if hasattr(kinetics, 'to_arrhenius'):
        ...

# After
kinetics_list = family.get_kinetics(reaction=deg_rxn, template_labels=deg_rxn.template, degeneracy=deg_rxn.degeneracy)
for kinetics_detailes in kinetics_list:
    kinetics = kinetics_detailes[0]
    if hasattr(kinetics, 'to_arrhenius'):  # ✅ Degeneracy already applied
        ...

2. Added training rules to kinetics families

def load_rmg_database() -> RMGDatabase:
    rmgdb = RMGDatabase()
    ...
    rmgdb.load_kinetics(path=os.path.join(DB_PATH, 'kinetics'),
                        reaction_libraries=kinetics_libraries,
                        kinetics_families='default',
                        kinetics_depositories=['training'])
    # Add training rules to each family
    for family in rmgdb.kinetics.families.values():
        family.add_rules_from_training(rmgdb)
    return rmgdb

Impact

• Kinetic rate coefficients will now have the correct A factor values (not multiplied by degeneracy twice)
• Kinetics families will include rules derived from training data, providing more complete and accurate kinetic estimates

Original prompt

In the rmg_kinetics.py file, we need to make changes. We need to remove the line 'kin.change_rate(cand.degeneracy)' because that just multiplies the A twice before output. In the load RMG database section we need to call on 'for family in rmgdb.kinetics.families.values(): family.add_rules_from_training(rmgdb)
I think we need to loop, I am not sure.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[alongd]

FWIW, Gemini says the issue is a circular import inside rmgpy.kinetics.model

I tried relocating from rmgpy.kinetics.surface import StickingCoefficient, StickingCoefficientBEP from the main imports into the is_similar_to() func, but it still complaints. Strange that it does it in our env and doesn't cause issues for RMG itself in runs/tests.