scikit learn 中的特殊性

I need specificity对于我的分类，其定义为：TN/(TN+FP)

我正在编写一个自定义记分器函数：

from sklearn.metrics import make_scorer
def specificity_loss_func(ground_truth, predictions):
    print predictions
    tp, tn, fn, fp = 0.0,0.0,0.0,0.0
    for l,m in enumerate(ground_truth):        
        if m==predictions[l] and m==1:
            tp+=1
        if m==predictions[l] and m==0:
            tn+=1
        if m!=predictions[l] and m==1:
            fn+=1
        if m!=predictions[l] and m==0:
            fp+=1
    `return tn/(tn+fp)

score = make_scorer(specificity_loss_func, greater_is_better=True)

Then,

from sklearn.dummy import DummyClassifier
clf_dummy = DummyClassifier(strategy='most_frequent', random_state=0)
ground_truth = [0,0,1,0,1,1,1,0,0,1,0,0,1]
p  = [0,0,0,1,0,1,1,1,1,0,0,1,0]
clf_dummy = clf_dummy.fit(ground_truth, p)
score(clf_dummy, ground_truth, p)

当我运行这些命令时，我得到p打印为：

[0 0 0 0 0 0 0 0 0 0 0 0 0]
1.0

为什么是我的p当我输入时更改为一系列零p = [0,0,0,1,0,1,1,1,1,0,0,1,0]

你可以获得specificity来自confusion matrix https://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html。对于二元分类问题，它会是这样的：

from sklearn.metrics import confusion_matrix
y_true = [0, 0, 0, 1, 1, 1, 1, 1]
y_pred = [0, 1, 0, 1, 0, 1, 0, 1]

tn, fp, fn, tp = confusion_matrix(y_true, y_pred).ravel()
specificity = tn / (tn+fp)