sklearn.preprocessing
1、對(duì)數(shù)據(jù)分類編碼
from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()
df_user3['user_id']=le.fit_transform(list(df_user3['user_id']))
le.classes_查看編碼對(duì)應(yīng)的user_id
生成對(duì)應(yīng)表
d = pd.Series(le.classes_)
print(d)
d.to_csv('D:/huashuData/dianhui/dianhui_user_code.csv')
2、對(duì)分類特征進(jìn)行二進(jìn)制（0，1）編碼
圖片.png
from sklearn.preprocessing import MultiLabelBinarizer
mlb=MultiLabelBinarizer()
mo = mo.join(pd.DataFrame(mlb.fit_transform(mo.pop('genres')),
columns=mlb.classes_,
index=mo.index))
代碼后效果.png
