peal.data.tabular_datasets.ADULT_FEATURES

peal.data.tabular_datasets.ADULT_FEATURES = ['hours-per-week', 'educational-num', 'occupation', 'workclass', 'race', 'age', 'marital-status', 'gender']

The published tabular Adult cell was produced by a one-off curation script (tools/curate_data_adult.py) whose one-hot column names are what configs/tabular_experiments/adaptors/adult_cfkd_dice.yaml selects, e.g. occupation_Professional and gender_Male. AdultDataset used to category-code the table instead, producing integer occupation and gender columns that no committed config can consume, so the cell could not run from a clean clone. The curation now lives here, which is the only arrangement where downloading and running both work with no manual step.