import pandas as pd
import numpy as np
def one_hot(df, c):
dummies = pd.get_dummies(df[c]) #关键点
# 每一列重命名
dummies.columns = [c + "_" + c1 for c1 in dummies.columns]
df = df.drop(c,axis = 1)
return df.join(dummies)
def one_hot_encoding(X):
cols = X.columns.copy() #复制列明
for c in cols:
if X[c].dtype == np.dtype: #判断每一列的类型
X = one_hot(X, c)
return X
file_name = "train_feature.xlsx"
df = pd.read_excel(file_name, index_col=0)
one_hot_encoding(df).to_excel("one_hot_" + file_name)
pandas实现one_hot编码
猜你喜欢
转载自blog.csdn.net/tailonh/article/details/110728333
今日推荐
周排行