Python roc_auc_score() 和 auc() 的不同结果
声明：本页面是StackOverFlow热门问题的中英对照翻译，遵循CC BY-SA 4.0协议，如果您需要使用它，必须同样遵循CC BY-SA许可，注明原文地址和作者信息，同时你必须将它归于原作者（不是我）：StackOverFlow
原文地址: http://stackoverflow.com/questions/31159157/
Warning: these are provided under cc-by-sa 4.0 license. You are free to use/share it, But you must attribute it to the original authors (not me):
StackOverFlow
Different result with roc_auc_score() and auc()
提问by gowithefloww
I have trouble understanding the difference (if there is one) between roc_auc_score()and auc()in scikit-learn.
我很难理解scikit-learnroc_auc_score()和auc()scikit-learn之间的区别（如果有的话）。
Im tying to predict a binary output with imbalanced classes (around 1.5% for Y=1).
我想预测具有不平衡类的二进制输出（Y = 1 时约为 1.5%）。
Classifier
分类器
model_logit = LogisticRegression(class_weight='auto')
model_logit.fit(X_train_ridge, Y_train)
Roc curve
洛克曲线
false_positive_rate, true_positive_rate, thresholds = roc_curve(Y_test, clf.predict_proba(xtest)[:,1])
AUC's
AUC的
auc(false_positive_rate, true_positive_rate)
Out[490]: 0.82338034042531527
and
和
roc_auc_score(Y_test, clf.predict(xtest))
Out[493]: 0.75944737191205602
Somebody can explain this difference ? I thought both were just calculating the area under the ROC curve. Might be because of the imbalanced dataset but I could not figure out why.
有人可以解释这种差异吗？我认为两者都只是在计算 ROC 曲线下的面积。可能是因为数据集不平衡，但我不知道为什么。
Thanks!
谢谢！
采纳答案by oopcode
AUC is not always area under the curve of a ROC curve. Area Under the Curve is an (abstract) area under somecurve, so it is a more general thing than AUROC. With imbalanced classes, it may be better to find AUC for a precision-recall curve.
AUC 并不总是 ROC 曲线下的面积。曲线下面积为下（抽象）地区的一些曲线，所以它比AUROC更一般的事情。对于不平衡的类，最好为精确召回曲线找到 AUC。
See sklearn source for roc_auc_score:
请参阅 sklearn 源代码roc_auc_score：
def roc_auc_score(y_true, y_score, average="macro", sample_weight=None):
# <...> docstring <...>
def _binary_roc_auc_score(y_true, y_score, sample_weight=None):
# <...> bla-bla <...>
fpr, tpr, tresholds = roc_curve(y_true, y_score,
sample_weight=sample_weight)
return auc(fpr, tpr, reorder=True)
return _average_binary_score(
_binary_roc_auc_score, y_true, y_score, average,
sample_weight=sample_weight)
As you can see, this first gets a roc curve, and then calls auc()to get the area.
如你所见，这首先得到一个 roc 曲线，然后调用auc()得到面积。
I guess your problem is the predict_proba()call. For a normal predict()the outputs are always the same:
我想你的问题是predict_proba()电话。对于正常predict()的输出总是相同的：
import numpy as np
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import roc_curve, auc, roc_auc_score
est = LogisticRegression(class_weight='auto')
X = np.random.rand(10, 2)
y = np.random.randint(2, size=10)
est.fit(X, y)
false_positive_rate, true_positive_rate, thresholds = roc_curve(y, est.predict(X))
print auc(false_positive_rate, true_positive_rate)
# 0.857142857143
print roc_auc_score(y, est.predict(X))
# 0.857142857143
If you change the above for this, you'll sometimes get different outputs:
如果为此更改上述内容，有时会得到不同的输出：
false_positive_rate, true_positive_rate, thresholds = roc_curve(y, est.predict_proba(X)[:,1])
# may differ
print auc(false_positive_rate, true_positive_rate)
print roc_auc_score(y, est.predict(X))
回答by Andreus
predictreturns only one class or the other. Then you compute a ROC with the results of predicton a classifier, there are only three thresholds (trial all one class, trivial all the other class, and in between). Your ROC curve looks like this:
predict只返回一个类或另一个类。然后你用predict分类器上的结果计算 ROC，只有三个阈值（尝试所有一个类，所有其他类都是微不足道的，以及介于两者之间）。您的 ROC 曲线如下所示：
..............................
|
|
|
......|
|
|
|
|
|
|
|
|
|
|
|
Meanwhile, predict_proba()returns an entire range of probabilities, so now you can put more than three thresholds on your data.
同时，predict_proba()返回整个概率范围，因此现在您可以对数据设置三个以上的阈值。
.......................
|
|
|
...|
|
|
.....|
|
|
....|
.|
|
|
|
|
Hence different areas.
因此不同的领域。
回答by Dayvid Oliveira
When you use the y_pred (class labels), you already decided on the threshold. When you use y_prob (positive class probability) you are open to the threshold, and the ROC Curve should help you decide the threshold.
当您使用 y_pred（类标签）时，您已经决定了阈值。当您使用 y_prob（正类概率）时，您对阈值持开放态度，并且 ROC 曲线应该可以帮助您确定阈值。
For the first case you are using the probabilities:
对于第一种情况，您使用的是概率：
y_probs = clf.predict_proba(xtest)[:,1]
fp_rate, tp_rate, thresholds = roc_curve(y_true, y_probs)
auc(fp_rate, tp_rate)
When you do that, you're considering the AUC 'before' taking a decision on the threshold you'll be using.
当你这样做时，你是在“在”决定你将使用的阈值之前考虑 AUC。
In the second case, you are using the prediction (not the probabilities), in that case, use 'predict' instead of 'predict_proba' for both and you should get the same result.
在第二种情况下，您使用的是预测（而不是概率），在这种情况下，对两者都使用 'predict' 而不是 'predict_proba'，您应该得到相同的结果。
y_pred = clf.predict(xtest)
fp_rate, tp_rate, thresholds = roc_curve(y_true, y_pred)
print auc(fp_rate, tp_rate)
# 0.857142857143
print roc_auc_score(y, y_pred)
# 0.857142857143
