How can I create MODData with my DataFrame?

[mhlee216]

How can I create MODData with my DataFrame from my CSV file?
Or how can I append a DataFrame to featurized MODData (featurize()) ?

[ppdebreuck]

Hi @mhlee216,

You should first extract the structures (list of pymatgen structures or compositions) and targets (2D array) from the CSV file.
Example: you read the csv as a dataframe df with columns "structures" and "e_form".

data = MODData(materials = df["structures"],
               targets = df[["e_form"]].values,
               target_names = ["formation_energy"]
              )

If you already have features you can pass them to the df_featurized keyword. Otherwise use data.featurize().

Finally, you can append other features to the featurized MODData by using join or concat: data.df_featurized = data.df_featurized.join(df_with_my_features). (Features are simply stored as a dataframe at data.df_featurized.)

Let me know if this works.

[mhlee216]

This works for me. Thank you so much for this solution!