机器学习基于缺失值的填充法则_18340082396

代码数据网址：https://work.datafountain.cn/forum?id=68&type=2&source=1

### 导入数据集
import numpy as np    # Python中进行数值计算的库
import pandas as pd    # Python中进行数据处理的库
import warnings
warnings.filterwarnings('ignore') #  忽略弹出的warnings
 
 
data=pd.read_csv("loan.csv",encoding="utf-8")
data.head(5)
print(data.shape)
 
me v:18340082396

查看缺失值的情况

### 查看数据集缺失情况
missingDf = data.isnull().sum().sort_values(ascending=False).reset_index()
missingDf.columns = ['feature', 'miss_num']    
missingDf['miss_percentage'] = missingDf['miss_num'] / data.shape[0]    # 缺失值比例 data.shape= (400000, 145)
missingDf.head(10)    # 缺失值最多的前十列特征