sklearn.preprocessing.Imputer
填补缺失值：sklearn.preprocessing.Imputer(missing_values=’NaN’, strategy=’mean’, axis=0, verbose=0, copy=True)主要参数说明：missing_values：缺失值，可以为整数或NaN(缺失值numpy.nan用字符串‘NaN’表示)，默认为NaNstrategy：替换策略，字符串，默认用均值‘mean’替换
①若为mean时，用特征列的均值替换②若为median时，用特征列的中位数替换③若为most_frequent时，用特征列的众数替换axis：指定轴数，默认axis=0代表列，axis=1代表行
copy：设置为True代表不在原数据集上修改，设置为False时，就地修改，存在如下情况时，即使设置为False时，也不会就地修改
①X不是浮点值数组②X是稀疏且missing_values=0③axis=0且X为CRS矩阵④axis=1且X为CSC矩阵statistics_属性：axis设置为0时，每个特征的填充值数组，axis=1时，报没有该属性错误
以X为数组为例：
[*]In [1]: import numpy as np
[*]
...: from sklearn.preprocessing import Imputer
[*]
...: train_X = np.array([[1, 2], 3], [7, 6]])
[*]
...: imp = Imputer(missing_values=np.nan , strategy='mean', axis=0)
[*]
...: imp.fit(train_X)
[*]
...:
[*]
Out[1]: Imputer(axis=0, copy=True, missing_values=nan, strategy='mean', verbose=
[*]
0)
[*]
[*]
In [2]: imp.statistics_
[*]
Out[2]: array([ 4. ,3.66666667])
[*]
[*]
In [3]: test_X = np.array([2], [6, np.nan], [7, 6]])
[*]
...: imp.transform(test_X)
[*]
...:
[*]
Out[3]:
[*]
array([[ 4. ,2. ],
[*]
[ 6. ,3.66666667],
[*]
[ 7. ,6. ]])
[*]
[*]
In [4]: imp.fit_transform(test_X)
[*]
Out[4]:
[*]
array([[ 6.5,2. ],
[*]
[ 6. ,4. ],
[*]
[ 7. ,6. ]])
[*]
[*]
In [5]: imp.statistics_
[*]
Out[5]: array([ 6.5,4. ])
以X为稀疏矩阵为例：
[*]
In [6]: import scipy.sparse as sp
[*]
...: from sklearn.preprocessing import Imputer
[*]
...: X = sp.csc_matrix([[1, 2], [0, 3], [7, 6]])
[*]
...: imp = Imputer(missing_values = 0 , strategy = 'mean',axis=0)
[*]
...: imp.fit(X)
[*]
...: test_X = sp.csc_matrix([[0, 2], [6, 0], [7, 6]])
[*]
...: imp.transform(test_X)
[*]
...:
[*]
Out[6]:
[*]
array([[ 4. ,2. ],
[*]
[ 6. ,3.66666667],
[*]
[ 7. ,6. ]])
[*]
[*]
In [7]: imp.statistics_
[*]
Out[7]: array([ 4. ,3.66666667])
页:
[1]