ValueError：尺寸不匹配

Question

我使用SciPy 和scikit-learn 来训练并应用多项式朴素贝叶斯分类器进行二进制文本分类。准确地说，我使用模块sklearn.feature_extraction.text.CountVectorizer 来创建稀疏矩阵，该矩阵包含来自文本的单词特征计数，模块sklearn.naive_bayes.MultinomialNB 作为分类器实现，用于在训练数据上训练分类器并将其应用于测试数据。

CountVectorizer 的输入是表示为 unicode 字符串的文本文档列表。训练数据远大于测试数据。我的代码看起来像这样（简化）：

vectorizer = CountVectorizer(**kwargs)

# sparse matrix with training data
X_train = vectorizer.fit_transform(list_of_documents_for_training)

# vector holding target values (=classes, either -1 or 1) for training documents
# this vector has the same number of elements as the list of documents
y_train = numpy.array([1, 1, 1, -1, -1, 1, -1, -1, 1, 1, -1, -1, -1, ...])

# sparse matrix with test data
X_test = vectorizer.fit_transform(list_of_documents_for_testing)

# Training stage of NB classifier
classifier = MultinomialNB()
classifier.fit(X=X_train, y=y_train)

# Prediction of log probabilities on test data
X_log_proba = classifier.predict_log_proba(X_test)

问题：只要调用MultinomialNB.predict_log_proba()，我就会得到ValueError: dimension mismatch。根据下面的IPython stacktrace，错误发生在SciPy中：

/path/to/my/code.pyc
--> 177         X_log_proba = classifier.predict_log_proba(X_test)

/.../sklearn/naive_bayes.pyc in predict_log_proba(self, X)
    76             in the model, where classes are ordered arithmetically.
    77         """
--> 78         jll = self._joint_log_likelihood(X)
    79         # normalize by P(x) = P(f_1, ..., f_n)
    80         log_prob_x = logsumexp(jll, axis=1)

/.../sklearn/naive_bayes.pyc in _joint_log_likelihood(self, X)
    345         """Calculate the posterior log probability of the samples X"""
    346         X = atleast2d_or_csr(X)
--> 347         return (safe_sparse_dot(X, self.feature_log_prob_.T)
    348                + self.class_log_prior_)
    349 

/.../sklearn/utils/extmath.pyc in safe_sparse_dot(a, b, dense_output)
    71     from scipy import sparse
    72     if sparse.issparse(a) or sparse.issparse(b):
--> 73         ret = a * b
    74         if dense_output and hasattr(ret, "toarray"):
    75             ret = ret.toarray()

/.../scipy/sparse/base.pyc in __mul__(self, other)
    276 
    277             if other.shape[0] != self.shape[1]:
--> 278                 raise ValueError('dimension mismatch')
    279 
    280             result = self._mul_multivector(np.asarray(other))

我不知道为什么会出现这个错误。有人可以向我解释一下并为这个问题提供解决方案吗？提前非常感谢！

Answer 1

在我看来，就像您只需要对测试数据集使用 vectorizer.transform 一样，因为训练数据集修复了词汇表（毕竟您无法知道包括训练集在内的完整词汇表）。需要明确的是，那是 vectorizer.transform 而不是 vectorizer.fit_transform。

Answer 2

另一个解决方案是使用vector.vocabulary

# after trainning the data
vector = CountVectorizer()
vector.fit(self.x_data)
training_data = vector.transform(self.x_data)
bayes = MultinomialNB()
bayes.fit(training_data, y_data)

# use vector.vocabulary for predict
vector = CountVectorizer(vocabulary=vector.vocabulary_) #vocabulary is a parameter, it should be vocabulary_ as it is an attribute.
text_vector = vector.transform(text)
trained_model.predict_prob(text_vector)