python - sklearn Imputer() 返回的特征不适合拟合函数

Question

我有一个包含缺失值 NaN 的特征矩阵，因此我需要先初始化这些缺失值。但是，最后一行提示并抛出以下错误行:预期序列或类数组，得到 Imputer(axis=0, copy=True, missing_values='NaN', strategy='mean', verbose=0)。我查了一下，好像是因为train_fea_imputed不是np.array格式，而是sklearn.preprocessing.imputation.Imputer格式。我该如何解决这个问题？
顺便说一句，如果我使用 train_fea_imputed = imp.fit_transform(train_fea)，代码工作正常，但 train_fea_imputed 返回一个比 train_fea 小一维的数组

    import pandas as pd
    import numpy as np
    from sklearn.ensemble import RandomForestClassifier
    from sklearn.preprocessing import Imputer

    imp = Imputer(missing_values='NaN', strategy='mean', axis=0)
    train_fea_imputed = imp.fit(train_fea)

    # train_fea_imputed = imp.fit_transform(train_fea)
    rf = RandomForestClassifier(n_estimators=5000,n_jobs=1, min_samples_leaf = 3)
    rf.fit(train_fea_imputed, train_label)

更新:我改成了

imp = Imputer(missing_values='NaN', strategy='mean', axis=1)

现在尺寸问题没有出现了。我认为插补函数存在一些固有问题。我完成项目后会回来。

Answer 1

使用 scikit-learn，初始化模型、训练模型和获得预测是独立的步骤。在你的情况下你有:

train_fea = np.array([[1,1,0],[0,0,1],[1,np.nan,0]])
train_fea
array([[  1.,   1.,   0.],
       [  0.,   0.,   1.],
       [  1.,  nan,   0.]])

#initialise the model
imp = Imputer(missing_values='NaN', strategy='mean', axis=0)

#train the model
imp.fit(train_fea)

#get the predictions
train_fea_imputed = imp.transform(train_fea)
train_fea_imputed
array([[ 1. ,  1. ,  0. ],
       [ 0. ,  0. ,  1. ],
       [ 1. ,  0.5,  0. ]])