How to define a function that calculates molecular descriptor values from SMILE strings in a dateframe?

Question

I am trying to define a function that I can use on multiple dataframes with SMILE string as values. The function should calculate the molecular descriptor values, so I can add them to the dataframe and use it in a machine learning model.

cached(cache={})
def get_molecular_descriptor_calculator():
    """Cache call to retreive molecular descriptors function"""
    from rdkit.ML.Descriptors.MoleculeDescriptors import MolecularDescriptorCalculator
    
    descriptor_vals = ['ExactMolWt', 'FpDensityMorgan1', 'FpDensityMorgan2', 'FpDensityMorgan3', 
               'FractionCSP3', 'HallKierAlpha', 'HeavyAtomCount', 'HeavyAtomMolWt', 'Ipc', 'Kappa1', 
               'Kappa2', 'Kappa3', 'LabuteASA', 'MaxAbsEStateIndex', 'MaxAbsPartialCharge', 'MaxEStateIndex', 
               'MaxPartialCharge', 'MinAbsEStateIndex', 'MinAbsPartialCharge', 'MinEStateIndex', 'MinPartialCharge',
               'MolLogP', 'MolMR', 'MolWt']
    
    return MolecularDescriptorCalculator(descriptor_vals), descriptor_vals


def vudf_compute_molecular_properties(smile_string_df: T.PandasDataFrame[str]) -> T.PandasSeries[dict]: 
    """Calculate molecular descriptor from SMILE values in a df"""
        
    #Load the libraries:
    from rdkit import Chem
    from rdkit.ML.Descriptors.MoleculeDescriptors import MolecularDescriptorCalculator
    
    smile_string_df.columns = ['SMILES']
    descriptor_calculator, descriptors = get_molecular_descriptor_calculator()
    
    def smiles_to_descriptors(smile):
        """Helper function to apply to batches of rows"""
        mol = Chem.MolFromSmiles(smile)
        mol_property_vals = descriptor_calculator.CalcDescriptors(mol)
        return dict(zip(descriptors, mol_property_vals))
    
    #smile_string_df['descriptors'] = smile_string_df.SMILES.apply(smi_to_descriptors)
    return smile_string_df.SMILES.apply(smiles_to_descriptors)

I got an error that says : NameError Traceback (most recent call last) Cell In[33], line 15 6 descriptor_vals = ['ExactMolWt', 'FpDensityMorgan1', 'FpDensityMorgan2', 'FpDensityMorgan3', 7 'FractionCSP3', 'HallKierAlpha', 'HeavyAtomCount', 'HeavyAtomMolWt', 'Ipc', 'Kappa1', 8 'Kappa2', 'Kappa3', 'LabuteASA', 'MaxAbsEStateIndex', 'MaxAbsPartialCharge', 'MaxEStateIndex', 9 'MaxPartialCharge', 'MinAbsEStateIndex', 'MinAbsPartialCharge', 'MinEStateIndex', 'MinPartialCharge', 10 'MolLogP', 'MolMR', 'MolWt'] 12 return MolecularDescriptorCalculator(descriptor_vals), descriptor_vals ---> 15 def vudf_compute_molecular_properties(smile_string_df: T.PandasDataFrame[str]) -> T.PandasSeries[dict]: 16 """Calculate molecular descriptor from SMILE values in a df""" 18 #Load the libraries:

NameError: name 'T' is not defined