AxisError：计算 roc_auc_score 时，轴 1 超出维度 1 数组的范围

Question

rf = RandomForestClassifier(max_features=5, n_estimators=500)
rf.fit(d2_train_dataset, Y_train)

nb = GaussianNB()
nb.fit(d2_train_dataset, Y_train)

rf_probs = rf.predict_proba(d2_train_dataset)
nb_probs = nb.predict_proba(d2_train_dataset)

rf_probs = rf_probs[:, 1]
nb_probs = nb_probs[:, 1]

rf_auc = roc_auc_score(Y_test, rf_probs,multi_class="ovo")
nb_auc = roc_auc_score(Y_test, nb_probs,multi_class="ovo")

那么这是此单元格中的错误“AxisError：axis 1 is out of bounds for array of dimension 1”

rf_auc = roc_auc_score(Y_test, rf_probs,multi_class="ovo")
nb_auc = roc_auc_score(Y_test, nb_probs,multi_class="ovo")

Answer 1

遇到了同样的问题，但这在我的情况下有效，请检查我的答案一次

roc_auc_score(Y_test,rf_probs,multi_cass = "ovo",average= "weighted")

否则，请尝试不执行这两个步骤

rf_probs = rf_probs[:, 1]
nb_probs = nb_probs[:, 1]