Python的Pandas：例外：數據必須是1維的

Question

這是我從教程中得到的

# Data Preprocessing

# Importing the libraries
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

# Importing the dataset
dataset = pd.read_csv('Data.csv')
X = dataset.iloc[:, :-1].values
y = dataset.iloc[:, 3].values

# Taking care of missing data
from sklearn.preprocessing import Imputer
imputer = Imputer(missing_values = 'NaN', strategy = 'mean', axis = 0)
imputer = imputer.fit(X[:, 1:3])
X[:, 1:3] = imputer.transform(X[:, 1:3])

# Encoding categorical data
# Encoding the Independent Variable
from sklearn.preprocessing import LabelEncoder, OneHotEncoder
labelencoder_X = LabelEncoder()
X[:, 0] = labelencoder_X.fit_transform(X[:, 0])
onehotencoder = OneHotEncoder(categorical_features = [0])
X = onehotencoder.fit_transform(X).toarray()
# Encoding the Dependent Variable
labelencoder_y = LabelEncoder()
y = labelencoder_y.fit_transform(y)

這是帶有編碼虛擬變量的X矩陣

1.000000000000000000e+00    0.000000000000000000e+00    0.000000000000000000e+00    4.400000000000000000e+01    7.200000000000000000e+04
0.000000000000000000e+00    0.000000000000000000e+00    1.000000000000000000e+00    2.700000000000000000e+01    4.800000000000000000e+04
0.000000000000000000e+00    1.000000000000000000e+00    0.000000000000000000e+00    3.000000000000000000e+01    5.400000000000000000e+04
0.000000000000000000e+00    0.000000000000000000e+00    1.000000000000000000e+00    3.800000000000000000e+01    6.100000000000000000e+04
0.000000000000000000e+00    1.000000000000000000e+00    0.000000000000000000e+00    4.000000000000000000e+01    6.377777777777778101e+04
1.000000000000000000e+00    0.000000000000000000e+00    0.000000000000000000e+00    3.500000000000000000e+01    5.800000000000000000e+04
0.000000000000000000e+00    0.000000000000000000e+00    1.000000000000000000e+00    3.877777777777777857e+01    5.200000000000000000e+04
1.000000000000000000e+00    0.000000000000000000e+00    0.000000000000000000e+00    4.800000000000000000e+01    7.900000000000000000e+04
0.000000000000000000e+00    1.000000000000000000e+00    0.000000000000000000e+00    5.000000000000000000e+01    8.300000000000000000e+04
1.000000000000000000e+00    0.000000000000000000e+00    0.000000000000000000e+00    3.700000000000000000e+01    6.700000000000000000e+04

問題是沒有列標簽 。 我試過了

something = pd.get_dummies(X)

但我得到以下例外

Exception: Data must be 1-dimensional

Answer 1

大多數sklearn方法都不關心列名，因為它們主要關注它們實現的ML算法背后的數學。 如果您可以提前確定標簽編碼，則可以在fit_transform()之后將列名添加回OneHotEncoder輸出。

首先，從原始dataset獲取預測變量的列名，不包括第一個（我們為LabelEncoder保留）：

X_cols = dataset.columns[1:-1]
X_cols
# Index(['Age', 'Salary'], dtype='object')

現在獲取編碼標簽的順序。 在這種特殊情況下，看起來LabelEncoder()按字母順序組織其整數映射：

labels = labelencoder_X.fit(X[:, 0]).classes_ 
labels
# ['France' 'Germany' 'Spain']

合並這些列名，然后在轉換為DataFrame時將它們添加到X ：

# X gets re-used, so make sure to define encoded_cols after this line
X[:, 0] = labelencoder_X.fit_transform(X[:, 0])
encoded_cols = np.append(labels, X_cols)
# ...
X = onehotencoder.fit_transform(X).toarray()
encoded_df = pd.DataFrame(X, columns=encoded_cols)

encoded_df
   France  Germany  Spain        Age        Salary
0     1.0      0.0    0.0  44.000000  72000.000000
1     0.0      0.0    1.0  27.000000  48000.000000
2     0.0      1.0    0.0  30.000000  54000.000000
3     0.0      0.0    1.0  38.000000  61000.000000
4     0.0      1.0    0.0  40.000000  63777.777778
5     1.0      0.0    0.0  35.000000  58000.000000
6     0.0      0.0    1.0  38.777778  52000.000000
7     1.0      0.0    0.0  48.000000  79000.000000
8     0.0      1.0    0.0  50.000000  83000.000000
9     1.0      0.0    0.0  37.000000  67000.000000

注意：例如我正在使用此數據集的數據 ，它看起來與OP使用的數據非常相似或相同。 注意輸出如何與OP的X矩陣相同。