metric.py

from sklearn.metrics import f1_score, precision_recall_fscore_support, classification_report

def evaluation(real_labels, pred_labels):
    f1_micro = f1_score(real_labels, pred_labels, average='micro')
    f1_macro = f1_score(real_labels, pred_labels, average='macro')
    f1_weighted = f1_score(real_labels, pred_labels, average='weighted')
    #f1_binary = f1_score(real_labels, pred_labels, average='binary')
    #f1_samples = f1_score(real_labels, pred_labels, average='samples')

    micro_p, micro_r, micro_f1, _ = precision_recall_fscore_support(real_labels, pred_labels, average='micro')
    macro_p, macro_r, macro_f1, _ = precision_recall_fscore_support(real_labels, pred_labels, average='macro')
    

    report = classification_report(real_labels, pred_labels)

    print('f1 micro: ',f1_micro)
    print('f1 macro: ',f1_macro)
    print('f1 weighted: ',f1_weighted)
    #print('f1 binary: ',f1_binary)
    #print('f1 samples: ',f1_samples)
    print('micro p, micro r, micro f1:', micro_p, micro_r, micro_f1)
    print('macro p, macro r, macro f1:', macro_p, macro_r, macro_f1)
    print(report)

if __name__ == '__main__':
    real_labels = [1,1,1,2,3,3,3,3]
    pred_labels = [1,1,2,2,1,3,1,2]
    evaluation(real_labels, pred_labels)