score() now accepts arbitrary **molecular_inputs and these will be passed directly to the scoring functions. So keyword arguments should match otherwise an error will be raised.
Replay buffer now requires a molecule_key as it may store more than just SMILES
Optional mol_ids can be passed, if provided, these will be used for duplication detection
If SMILES are provided, canonicalisation with RDKit is now optional
pytest_configs.py to test new code runs (predominantly just this for now)
To-do
Mol_ids need to be updated if they are duplicate, currently it's just registered as unique=False but the mol_id value should be updated to the first recorded. This is also required for the replay buffer to work correctly.
Update GUI monitor such that it doesn't throw errors if SMILES aren't present
Draft changes addressing #63
Summary of changes so far
score()now accepts arbitrary**molecular_inputsand these will be passed directly to the scoring functions. So keyword arguments should match otherwise an error will be raised.molecule_keyas it may store more than just SMILESpytest_configs.pyto test new code runs (predominantly just this for now)To-do
unique=Falsebut themol_idvalue should be updated to the first recorded. This is also required for the replay buffer to work correctly.