bug report on table_2.ipynb #5
Description
In the last section (DeepBeat Multi-task (pretrain CDAE + AF + Excellent QA)) table_2.ipynb:
# weighted macro-average across all indivduals
test_metrics = util.collecting_individual_metrics(deepbeat, x_test_excellent, p_test_excellent, rhythm_test_excellent, out_message=False)
test_metrics = pd.DataFrame.from_dict(test_metrics).T.rename(columns={0:'TPR', 1:'TNR', 2:'FPR', 3:'FNR', 4:"total_samples"})
" 4:"total_samples" should be " 5:"total_samples", the current code will not produce the results as shown in the notebook. the results it produces currently:
TPR: 0.95
TNR: nan
FPR: nan
FNR: 0.05
PPV: 0.94
NPV: 1.00
F1: 0.96
Column 4 is F1 Score, Column 5 is total samples