stratification = housing.groupby('income_cat',observed = True,group_keys = True) stratified_sampling = stratification.apply(lambda y:y.sample(frac=0.1, random_state=1234), include_groups=False) vs stratification = housing.groupby('income_cat',observed = True,group_keys = False) stratified_sampling = stratification.apply(lambda y:y.sample(frac=0.1, random_state=1234), include_groups=True) When group_keys = False, include_groups = True, it was included as a general column, not as a multi-index. Before include_groups=True disappears. Did it print the same result? Continue reading...
