我试图转换多个包含大量分类值数据的列，但在使用OneHotEncoder时遇到了错误

我的数据框架

1) 将列分离到X_census和Y_census中（X_census包含分类值）：

X_census  = df[['workclass',               'education',               'marital-status',               'occupation',               'relationship',               'race',               'sex',               'native-country']]Y_census = df['income']

2) 使用LabelEncoder处理X_census中的分类值：

from sklearn.preprocessing import LabelEncoder

le = LabelEncoder()X_1 = X_census.apply(le.fit_transform)X_2 = X_1.to_numpy()

3) 现在对X_2使用OneHotEncoder将分类值转换为数值：

from sklearn.preprocessing import OneHotEncoderfrom sklearn.compose import ColumnTransformeroh = OneHotEncoder()onehotencoder_census = ColumnTransformer(transformers=[('OneHot', oh, X_2[:])],remainder='passthrough')X_census = onehotencoder_census.fit_transform(X_census) # 错误出现在这里！

错误信息

回答：

你可以使用pandas.get_dummies

df = pd.DataFrame({“marital_status”:[‘S’,’M’,’D’,’S’,’M’,’D’,’S’,’M’,’D’],”sex”:[“male”,”female”,”male”,”female”,”male”,”female”,”male”,”female”,”male”],”education”:[‘grad’,’post-grad’,’grad’,’post-grad’,’grad’,’post-grad’,’grad’,’post-grad’,’grad’],”income”:[125,135,120,110,90,150,180,130,110]})

pd.get_dummies(df)