def __init__(self,
idx_list,
transformer_list, n_jobs=None,
transformer_weights=None, verbose=False,
):
FeatureUnion.__init__(transformer_list=transformer_list, n_jobs=n_jobs,
transformer_weights=transformer_weights, verbose=verbose)
self.idx_list = idx_list
def __init__(self, transformer_list, idx_list, n_jobs=1, transformer_weights=None):
'''
:param transformer_list: 做transform操作的方法的列表
:param idx_list: 做transform操作的列（注意是列数）的列表
:param n_jobs:
:param transformer_weights:
'''
self.idx_list = idx_list # 新添加的参数用来存储列名或列序列号（注意从0开始）
FeatureUnion.__init__(self, transformer_list=transformer_list,
n_jobs=n_jobs, transformer_weights=transformer_weights)
def __init__(self,
transformer_list,
idx_list,
n_jobs=1,
transformer_weights=None):
self.idx_list = idx_list
FeatureUnion.__init__(self,
transformer_list=map(
lambda trans: (trans[0], trans[1]),
transformer_list),
n_jobs=n_jobs,
transformer_weights=transformer_weights)
def __init__(self,**kwargs):
self.title_features = Pipeline([('selector',FeatureSelector('title')),
('tfidf',title_vectorizer),
('scale',Normalizer())])
self.top_subject_features = Pipeline([('selector',FeatureSelector('top_subject')),
('encoder',LabelBinarizer(sparse_output=True))])
self.date_features = Pipeline([('selector',FeatureSelector('date')),
('encoder',DateEncoder())])
self.text_features = Pipeline([('selector',FeatureSelector('text')),
('tfidf',text_vectorizer),
('scale',Normalizer())])
FeatureUnion.__init__(self,
transformer_list=[('title',self.title_features),
('top_subject',self.top_subject_features),
('date',self.date_features),
('text',self.text_features)],
**kwargs)
def __init__(self, transformer_list, idx_list, n_jobs=1, transformer_weights=None):
self.idx_list = idx_list
FeatureUnion.__init__(self, transformer_list=map(lambda trans:(trans[0], trans[1]), transformer_list), n_jobs=n_jobs, transformer_weights=transformer_weights)
def __init__(self, **kw):
FeatureUnion.__init__(self, **kw)
