Importing CSV with pd.read_json()

[pbenner]

The file format of DATA_FILES.mp_energies is now CSV:

matbench-discovery/models/cgcnn/train_cgcnn.py

Line 65 in 4a90dee

df_in = pd.read_json(data_path).set_index(id_col)

Also, the structure column was removed, which is required for training.

[janosh]

Great catch! I found 3 more places with similar problems. Hopefully didn't miss any.

Might have been smarter to store everything as JSON (at the cost of ~50% larger files).

[pbenner]

Thanks for the quick fix. Now there is two times an assignment to the variable data_files. This line should be deleted:

matbench-discovery/models/cgcnn/train_cgcnn.py

Line 64 in ee5b161

data_path = DATA_FILES.mp_computed_structure_entries

[janosh]

Thanks!