标签编码器和OneHotEncoder的值误差

Question

我试图将一个分类字符串列转换为几个虚拟变量二进制列，但我得到了一个值错误。

下面是代码：

import sys, os
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from dateutil import parser
import math
import traceback
import logging
datasetMod = pd.read_csv('data.csv')

X = datasetMod.iloc[:, 3:6].values
y = datasetMod.iloc[:, 1].values
print(X[:, 0])

# Encoding categorical data
from sklearn.preprocessing import LabelEncoder, OneHotEncoder
try:
    labelencoder_X = LabelEncoder()
    X[:, 0] = labelencoder_X.fit_transform(X[:, 0])
    onehotencoder = OneHotEncoder(categorical_features = [0])
    X = onehotencoder.fit_transform(X).toarray()
except Exception as e:
    exc_type, exc_obj, exc_tb = sys.exc_info()
    fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
    print(exc_type, fname, exc_tb.tb_lineno)

以下是错误：

<class 'ValueError'> multipleLinearRegression.py 23

该列的打印语句的结果是：

['Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Weekend' 'Workday' 'Workday' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday' 'Workday'
 'Workday' 'Workday' 'Workday' 'Workday' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend' 'Weekend'
 'Weekend' 'Weekend' 'Weekend' 'Weekend']

字符串本身似乎没有什么问题，中间没有空格，也没有像符号那样的数字。所以我不明白为什么我得到的值类型不能将字符串转换为浮动错误。

任何帮助都将不胜感激。

更新

一个编码器现在工作得还不错，但是最终结果是类型对象，而它应该是float64类型的：

labelencoder_X = LabelEncoder()
X[:, 1] = labelencoder_X.fit_transform(X[:, 1])
X[:, 2] = labelencoder_X.fit_transform(X[:, 2])
X[:, 3] = labelencoder_X.fit_transform(X[:, 3])
onehotencoder = OneHotEncoder(categorical_features = [1,2,3])
onehotencoder.fit(X[:, 1])
onehotencoder.fit(X[:, 2])
onehotencoder.fit(X[:, 3])
onehotencoder.transform(X[:, 1])
onehotencoder.transform(X[:, 2])
onehotencoder.transform(X[:, 3])
X = onehotencoder.toArray()  

更新2

from sklearn.preprocessing import LabelEncoder, OneHotEncoder

labelencoder_X = LabelEncoder()
X[:, 1] = labelencoder_X.fit_transform(X[:, 1])
X[:, 2] = labelencoder_X.fit_transform(X[:, 2])
X[:, 3] = labelencoder_X.fit_transform(X[:, 3])

onehotencoder = OneHotEncoder(categorical_features = [1,2,3])
X[:, 1] = onehotencoder.fit_transform(X[:, 1]).toarray()
X[:, 2] = onehotencoder.fit_transform(X[:, 2]).toarray()
X[:, 3] = onehotencoder.fit_transform(X[:, 3]).toarray()

print(X.dtype) #object

最终代码

由于categorical_features已经指定了索引，所以我可以对整个矩阵X进行fit_transform()。感谢@mkos的耐心！

from sklearn.preprocessing import LabelEncoder, OneHotEncoder
labelencoder_X = LabelEncoder()
X[:, 1] = labelencoder_X.fit_transform(X[:, 1])
X[:, 2] = labelencoder_X.fit_transform(X[:, 2])
X[:, 3] = labelencoder_X.fit_transform(X[:, 3])
onehotencoder = OneHotEncoder(categorical_features = [1,2,3])
X = onehotencoder.fit_transform(X)

Answer 1

这应该能起作用：

onehotencoder = OneHotEncoder(categorical_features = [1,2,3])
X = onehotencoder.fit_transform(X)

你可以用以下方式打印：

print(X.toArray())

将X作为稀疏矩阵并不坏，因为它节省了内存。如果您想要看到它，那么您可以将它转换为带有np.array的普通toArray()。