import pandas as pd

import numpy as np

import matplotlib

import matplotlib.pyplot as plt

import seaborn as sns

matplotlib.rcParams[‘font.family’]=‘SimHei’

data=pd.read_csv(‘OnlineRetail.csv’,encoding=‘ISO-8859-1’)

data.info()

data.apply(lambda x:sum(x.isnull())/len(x),axis=0)

data.drop([‘Description’],axis=1,inplace=True)

data

data[‘CustomerID’]=data[‘CustomerID’].fillna(‘U’)#填充缺失值

data[‘amount’]=data[‘Quantity’]*data[‘UnitPrice’]

data[‘date’]=data[‘InvoiceDate’].str.split(’ ‘).str[0]

data[‘time’]=data[‘InvoiceDate’].str.split(’ ‘).str[1]

data.drop([‘InvoiceDate’],inplace=True,axis=1)

data[‘year’]=data[‘date’].str.split(’/’).str[2]

data[‘month’]=data[‘date’].str.split(’/’).str[0]

data[‘day’]=data[‘date’].str.split(’/’).str[1]

data[‘date’]=pd.to_datetime(data[‘date’])

data=data.drop_duplicates()

data.describe()

data1=data[data[‘UnitPrice’]<=0]

data1.shape[0]/data.shape[0]

data1[‘UnitPrice’].value_counts()

Tuihuo=data[data[‘Quantity’]<0]

Tuihuo

tt=pd.pivot_table(Tuihuo,index=‘year’,columns=‘month’,values=‘amount’,aggfunc=np.sum)

data2=data[(data[‘Quantity’]>0)&(data[‘UnitPrice’]>0)]

data2

pp=pd.pivot_table(data2,index=‘year’,columns=‘month’,values=‘amount’,aggfunc=np.sum)

np.abs(tt/pp).loc[‘2011’].mean()#计算2011年平均退货率

R_value=data2.groupby(‘CustomerID’)[‘date’].max()#求出每个ID最近购买的日期

R_value=(data2[‘date’].max()-R_value)#得到每个用户距离最近购买的时候的天数

R_value=R_value.dt.days#取出天数

R_value#值越小越好

import seaborn as sns

sns.set(style=‘darkgrid’)

plt.hist(R_value,bins=30)#直方图

```python

在这里插入代码片