[英]How to get roc auc for binary classification in sklearn
I have binary classification problem where I want to calculate the roc_auc of the results.我有一个二元分类问题,我想计算结果的 roc_auc。 For this purpose, I did it in two different ways using sklearn.为此,我使用 sklearn 以两种不同的方式完成了这项工作。 My code is as follows.我的代码如下。
Code 1:代码 1:
from sklearn.metrics import make_scorer
from sklearn.metrics import roc_auc_score
myscore = make_scorer(roc_auc_score, needs_proba=True)
from sklearn.model_selection import cross_validate
my_value = cross_validate(clf, X, y, cv=10, scoring = myscore)
print(np.mean(my_value['test_score'].tolist()))
I get the output as 0.60
.我得到的输出为0.60
。
Code 2:代码 2:
y_score = cross_val_predict(clf, X, y, cv=k_fold, method="predict_proba")
from sklearn.metrics import roc_curve, auc
fpr = dict()
tpr = dict()
roc_auc = dict()
for i in range(2):
fpr[i], tpr[i], _ = roc_curve(y, y_score[:,i])
roc_auc[i] = auc(fpr[i], tpr[i])
print(roc_auc)
I get the output as {0: 0.41, 1: 0.59}
.我得到的输出为{0: 0.41, 1: 0.59}
。
I am confused since I get two different scores in the two codes.我很困惑,因为我在两个代码中得到了两个不同的分数。 Please let me know why this difference happens and what is the correct way of doing this.请让我知道为什么会发生这种差异以及这样做的正确方法是什么。
I am happy to provide more details if needed.如果需要,我很乐意提供更多详细信息。
It seems that you used a part of my code from another answer, so I though to also answer this question.看来你从另一个答案中使用了我的一部分代码,所以我也回答了这个问题。
For a binary classification case, you have 2 classes and one is the positive class.对于二元分类案例,您有 2 个类,一个是正类。
For example see here .例如,请参见此处。 pos_label
is the label of the positive class. pos_label
是正类的标签。 When pos_label=None
, if y_true
is in {-1, 1}
or {0, 1}
, pos_label
is set to 1
, otherwise an error will be raised..当pos_label=None
,如果y_true
在{-1, 1}
或{0, 1}
,则pos_label
设置为1
,否则将引发错误..
import matplotlib.pyplot as plt
from sklearn import svm, datasets
from sklearn.metrics import roc_curve, auc
from sklearn.multiclass import OneVsRestClassifier
from sklearn.model_selection import cross_val_predict
from sklearn.linear_model import LogisticRegression
import numpy as np
iris = datasets.load_iris()
X = iris.data
y = iris.target
mask = (y!=2)
y = y[mask]
X = X[mask,:]
print(y)
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1]
positive_class = 1
clf = OneVsRestClassifier(LogisticRegression())
y_score = cross_val_predict(clf, X, y, cv=10 , method='predict_proba')
fpr = dict()
tpr = dict()
roc_auc = dict()
fpr[positive_class], tpr[positive_class], _ = roc_curve(y, y_score[:, positive_class])
roc_auc[positive_class] = auc(fpr[positive_class], tpr[positive_class])
print(roc_auc)
{1: 1.0}
and和
from sklearn.metrics import make_scorer
from sklearn.metrics import roc_auc_score
from sklearn.model_selection import cross_validate
myscore = make_scorer(roc_auc_score, needs_proba=True)
clf = OneVsRestClassifier(LogisticRegression())
my_value = cross_validate(clf, X, y, cv=10, scoring = myscore)
print(np.mean(my_value['test_score'].tolist()))
1.0
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.