DataFrame分组聚合函数

import numpy as np
import pandas as pd

df = pd.read_csv("D:\海运数据\数据需求满足\DB导出\MANIFEST_CONSUME_RECORD_202204201421.csv")
df

92425 rows × 9 columns

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 92425 entries, 0 to 92424
Data columns (total 9 columns):
 #   Column        Non-Null Count  Dtype  
---  ------        --------------  -----  
 0   CUSTOMER_ID   92425 non-null  int64  
 1   COMPANY_NAME  92425 non-null  object 
 2   USER_ID       92425 non-null  int64  
 3   CONSUME_TYPE  92425 non-null  object 
 4   CREATE_TIME   92425 non-null  object 
 5   BILL_NO_NUM   92425 non-null  int64  
 6   CONSUME_NUM   92425 non-null  float64
 7   SHIP_AGENT    91583 non-null  object 
 8   FIRST         92425 non-null  int64  
dtypes: float64(1), int64(4), object(4)
memory usage: 6.3+ MB

def split_ship(x):
    if type(x) is str:
        return str(x[2:])

apply函数

df['SHIP_AGENT'] = df['SHIP_AGENT'].apply(split_ship)

df

92425 rows × 9 columns

groupby

df.groupby(['COMPANY_NAME']).sum()

2797 rows × 5 columns

df.groupby(['COMPANY_NAME']).agg(['max','min','mean','count','std'])

2797 rows × 25 columns

com_gro = df.groupby(by='COMPANY_NAME')
com_gro.agg({'CONSUME_NUM': np.sum,'BILL_NO_NUM':[np.sum,np.mean]})

2797 rows × 3 columns