题解 | 筛选某店铺最有价值用户中消费最多前5名

import pandas as pd
from numpy import True_
sales = pd.read_csv('sales.csv')
pd.set_option('display.max_columns', None)# 显示所有的列，而不是以……显示
pd.set_option('display.max_rows', None)# 显示所有的行，而不是以……显示
pd.set_option('display.width', None) # 不自动换行显示
def rank(s,q):
    top_quantile=s[q].quantile(0.75)
    down_quantile=s[q].quantile(0.25)
    med=s[q].median()
    quartile=[]
    if q=='recency':
        for i in s[q]:
            if i<= down_quantile:
                quartile.append(4)
            elif i>down_quantile and i<=med:
                quartile.append(3)
            elif i>med and i<=top_quantile:
                quartile.append(2)
            else:
                quartile.append(1)
    else:
        for i in s[q]:
            if i<= down_quantile:
                quartile.append(1)
            elif i>down_quantile and i<=med:
                quartile.append(2)
            elif i>med and i<=top_quantile:
                quartile.append(3)
            else:
                quartile.append(4)
    return quartile
sales['R_Quartile']=rank(sales,'recency')
sales['F_Quartile']=rank(sales,'frequency')
sales['M_Quartile']=rank(sales,'monetary')

sales['RFMClass'] = sales['R_Quartile'].apply(str) + sales['F_Quartile'].apply(str)+ sales['M_Quartile'].apply(str)
sales.drop(['R_Quartile', 'F_Quartile', 'M_Quartile'], axis = 1, inplace=True)
print(sales.head())
print()
# print('\n') 
result = sales[sales['RFMClass'] == '444'].sort_values(by = 'monetary', ascending= False).reset_index(drop=True)
print(result.head())