京东用户行为数据分析——以Python为主要工具

京东数据分析——以Python为主要工具

---

• 数据探索分析
• 用户购买行为建模

数据基本情况

1.数据来源
2.数据概况
3.本文内容

1.探索分析

import pandas as pd 
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
plt.style.use('ggplot')
plt.rcParams['font.sans-serif']=['SimHei']
plt.rcParams['axes.unicode_minus']=False

df = pd.read_csv('JDdata.csv')
df.head()

customer_id	product_id	action_date	action_id	type	age_range	gender	customer_register_date	customer_level	city_level	brand	shop_id	category	product_market_date	vender_id	fans_number	vip_number	shop_register_date	shop_category	shop_score
0	1068082	5924	2018/4/1	3853455	PageView	6.0	W	2012/8/13	1	4.0	Li-Ning	8002	Badminton	2013/6/14	1457	213610	357580	2013/5/23	Outdoor Sports
1	1068082	10723	2018/4/1	3853455	PageView	6.0	W	2012/8/13	1	4.0	Li-Ning	8002	Badminton	2011/7/19	1457	213610	357580	2013/5/23	Outdoor Sports
2	735392	4559	2018/2/6	9434940	PageView	6.0	M	2009/9/15	7	5.0	Other	6944	Cookie	2015/11/16	355	63969	85597	2014/10/31	Food
3	1516193	374293	2018/3/17	2105553	PageView	6.0	M	2016/5/25	1	4.0	illuma	3830	Milk Power	2017/12/25	25	11576	20927	2016/12/16	Mother and Infant
5	1160556	285165	2018/2/4	8914730	PageView	6.0	M	2012/3/11	7	4.0	Other	744	Coat	2016/5/16	8627	180250	52606	2016/5/16	Clothes

#数据清洗与数据类型变换，将时间变为标准时间格式
df=df.dropna(how='any',axis=0)
df['action_date']=pd.to_datetime(df.action_date)
df['customer_register_date']=pd.to_datetime(df.customer_register_date)
df['shop_register_date']=pd.to_datetime(df.shop_register_date)
df['product_market_date']=pd.to_datetime(df.product_market_date)

#查询用户购买行为情况
order_comment=df.query("type=='Order'| type=='Comment' " )
order_comment.sort_values(by='action_date',ascending=True)
order_comment['month']=order_comment.action_date.values.astype('datetime64[M]')

#根据用户购物的月份生成数据透视表，汇总每月购物的用户
order_month=order_comment.groupby(by='month')
order_count_by_month=order_month.customer_id.count()
plt.figure(1,figsize=(10,4)) #生成图表,生成两张子图
plt.subplot(1,2,1)
order_count_by_month.plot() #子图1 每月购物的用户数
plt.subplot(1,2,2)
order_count_by_month.