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

#读取数据
a=pd.read_csv(r'sales.csv')
#准备好四分位数
r1=a['recency'].quantile(0.25)
r2=a['recency'].quantile(0.5)
r3=a['recency'].quantile(0.75)

f1=a['frequency'].quantile(0.25)
f2=a['frequency'].quantile(0.5)
f3=a['frequency'].quantile(0.75)

m1=a['monetary'].quantile(0.25)
m2=a['monetary'].quantile(0.5)
m3=a['monetary'].quantile(0.75)

#给用户赋分
a.loc[a['recency']<=r1,'R_Quartile']=4
a.loc[(a['recency']>r1) & (a['recency']<=r2),'R_Quartile']=3
a.loc[(a['recency']>r2) & (a['recency']<=r3),'R_Quartile']=2
a.loc[a['recency']>r3,'R_Quartile']=1

a.loc[a['frequency']<=f1,'F_Quartile']=1
a.loc[(a['frequency']<=f2) & (a['frequency']>f1),'F_Quartile']=2
a.loc[(a['frequency']<=f3) & (a['frequency']>f2),'F_Quartile']=3
a.loc[a['frequency']>f3,'F_Quartile']=4

a.loc[a['monetary']<m1,'M_Quartile']=1
a.loc[(a['monetary']<=f2) & (a['monetary']>f1),'M_Quartile']=2
a.loc[(a['monetary']<=f3) & (a['monetary']>f2),'M_Quartile']=3
a.loc[a['monetary']>f3,'M_Quartile']=4

#由于class类需要合并 转为str而不用数字，否则加到一起了
a['R_Quartile']=a['R_Quartile'].astype(int).astype(str)#先转int类，去掉保留的一位小数，再转str
a['F_Quartile']=a['F_Quartile'].astype(int).astype(str)
a['M_Quartile']=a['M_Quartile'].astype(int).astype(str)

#创建RFMClass列并整理数据
a['RFMClass']=a['R_Quartile']+a['F_Quartile']+a['M_Quartile']
a.drop(columns=['R_Quartile','F_Quartile','M_Quartile'],inplace=True)
b=a.sort_values(by='RFMClass',ascending=False).reset_index(drop=True)

print(a.head(),'\n',b.head())