我正在尝试为决策树和多项式朴素贝叶斯分类器准备数据。
这是我的数据看起来的样子（pandas数据框）
Label Feat1 Feat2 Feat3 Feat40 1 3 2 11 0 1 1 22 2 2 1 13 3 3 2 3我已经将数据分成了dataLabel和dataFeatures。使用dataLabel.ravel()准备了dataLabel
我需要对特征进行离散化处理，以便分类器将它们视为分类而非数值数据。
我尝试使用OneHotEncoder来实现这一点
enc = OneHotEncoder()enc.fit(dataFeatures)chk = enc.transform(dataFeatures)from sklearn.naive_bayes import MultinomialNBmnb = MultinomialNB()from sklearn import metricsfrom sklearn.cross_validation import cross_val_scorescores = cross_val_score(mnb, Y, chk, cv=10, scoring='accuracy')我得到了这个错误 – bad input shape (64, 16)
这是标签和输入的形状
dataLabel.shape = 72chk.shape = 72,16
为什么分类器不接受onehotencoded特征？
编辑 – 完整的堆栈跟踪代码
/root/anaconda2/lib/python2.7/site-packages/sklearn/utils /validation.py:386: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and willraise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample. DeprecationWarning)Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/root/anaconda2/lib/python2.7/site-packages/sklearn /cross_validation.py", line 1433, in cross_val_scorefor train, test in cv) File "/root/anaconda2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.py", line 800, in __call__while self.dispatch_one_batch(iterator): File "/root/anaconda2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.py", line 658, in dispatch_one_batchself._dispatch(tasks) File "/root/anaconda2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.py", line 566, in _dispatchjob = ImmediateComputeBatch(batch) File "/root/anaconda2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.py", line 180, in __init__self.results = batch() File "/root/anaconda2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.py", line 72, in __call__return [func(*args, **kwargs) for func, args, kwargs in self.items] File "/root/anaconda2/lib/python2.7/site-packages/sklearn/cross_validation.py", line 1531, in _fit_and_scoreestimator.fit(X_train, y_train, **fit_params) File "/root/anaconda2/lib/python2.7/site-packages/sklearn/naive_bayes.py", line 527, in fitX, y = check_X_y(X, y, 'csr') File "/root/anaconda2/lib/python2.7/site-packages/sklearn/utils/validation.py", line 515, in check_X_yy = column_or_1d(y, warn=True) File "/root/anaconda2/lib/python2.7/site-packages/sklearn/utils/validation.py", line 551, in column_or_1draise ValueError("bad input shape {0}".format(shape))ValueError: bad input shape (64, 16)
回答：
首先，你需要交换chk和Y，考虑cross_val_score文档。其次，你没有指定Y是什么，所以我希望它是一个一维数组。最后，最好是将所有转换器结合在一个分类器中使用Pipeline，而不是单独使用。像这样：
from sklearn import metricsfrom sklearn.cross_validation import cross_val_scorefrom sklearn.naive_bayes import MultinomialNBfrom sklearn.pipeline import Pipelineclf = Pipeline([ ('transformer', OneHotEncoder()), ('estimator', MultinomialNB()),])scores = cross_val_score(clf, dataFeatures.values, Y, cv=10, scoring='accuracy')