使用OneHotEncoding更改类别变量的问题

Question

我有一列性别，其值是男性（1）和女性（0）。当我使用以下代码对性别列进行一次编码时：

onehotencoder = OneHotEncoder(categorical_features=['gender'])
data = onehotencoder.fit_transform(data).toarray()

我收到以下错误：

IndexError: arrays used as indices must be of integer (or boolean) type

Answer 1

以下描述可以在OneHotEncoder的文档中找到：

categorical_features : ‘all’ or array of indices or mask, default=’all’ Specify what features are treated as categorical. ‘all’: All features are treated as categorical. array of indices: Array of categorical feature indices. mask: Array of length n_features and with dtype=bool.

因此，除了传递列名之外，还应该传递列索引，这样就可以解决您的问题。

顺便说一句，请注意文档中规定的弃用：
Deprecated since version 0.20: The categorical_features keyword was deprecated in version 0.20 and will be removed in 0.22. You can use the ColumnTransformer instead.