小编use*_*751的帖子

我在scikit学习中创建了一个管道,

pipeline = Pipeline([
    ('bow', CountVectorizer()),  
    ('classifier', BernoulliNB()), 
])

并使用交叉验证计算准确性

scores = cross_val_score(pipeline,  # steps to convert raw messages      into models
                     train_set,  # training data
                     label_train,  # training labels
                     cv=5,  # split data randomly into 10 parts: 9 for training, 1 for scoring
                     scoring='accuracy',  # which scoring metric?
                     n_jobs=-1,  # -1 = use all cores = faster
                     )

如何报告混淆矩阵而不是"准确性"？

我有一个由0,1,-1组成的pandas数据帧.

import pandas as pd
df=pd.DataFrame({'indicator':[0, 0, 0, -1,0,0,1,0,-1,1,0,-1,0,1]})

我不想找到每-1和1的索引,这样-1后面跟着一些或没有零和一个+1.对于上面例子中的exapmle,我想得到

[(3,6),(8,9),(11,13)]