dataframe 列拆分

# -*- coding:utf-8 -*-
import pandas as pd
import numpy as np

df
    Type        Brands_rank
0   Computer    Mac-Dell-Lenovo
1   Phone       Mac-XiaoMi-HuaWei
2   Pad         Mac-HuaWei

• 需求：将Brands_rank列中的不同品牌名，单独列出来
• 观察数据：不同的品牌名，由横杆(-)隔开，第3行只有1个横杆(-)
• 为实现需求，可以将Brands_rank列拆为3列，或者将3行Brands_rank数据拆为8行。

基于Brands_rank列构造多列

df_column['Brands-First'] = df.Brands_rank.astype('str').str[0:3]

print(df_column)
    Type        Brands-First
0   Computer    Mac
1   Phone       Mac
2   Pad         Mac

df_column['Brands-Second'] = df.Brands_rank.apply(lambda x: x.split('-')[1])  
print(df_column)
    Type      Brands-First  Brands-Second
0   Computer    Mac           Dell
1   Phone       Mac           XiaoMi
2   Pad         Mac           HuaWei

df_column['Brands-Third'] = df.Brands_rank.apply(lambda x: x.split('-')[2] if x.count('-') >= 2 else np.nan)  

print(df_column)
    Type      Brands-First  Brands-Second   Brands-Third
0   Computer     Mac            Dell          Lenovo
1   Phone        Mac           XiaoMi         HuaWei
2   Pad          Mac           HuaWei         NaN

原文链接