以淘宝APP数据为原始数据集，通过行业常见行业指标对淘宝用户行为进行分析，从而构建用户画像。
# 导入数据集，查看数据量import pandas as pddf = pd.read_csv('taobao.csv')df.head()
#导入所需库import numpy as npimport pandas as pdimport warningswarnings.filterwarnings('ignore')from datetime import datetime%matplotlib inlinefrom matplotlib import pyplot as pltplt.rcParams['font.sans-serif'] = 'SimHei' # 设置中文显示plt.rcParams['axes.unicode_minus'] = False
# 数据抽样，并预览数据结构df = df.sample(frac=0.3,random_state=None).reset_index()df.head()
df.info()
# 查看是否存在缺失值df.isnull().sum()
# 删除缺失字段df.drop('user_geohash',axis=1,inplace=True)df
# 将time日期拆分为日期和时间df['date'] = df['time'].str[0:10]df['date'] = pd.to_datetime(df['date'],format='%Y-%m-%d')df['time'] = df['time'].str[11:]df['time'] = df['time'].astype(int)df.head()
# 构造面板数据，将时段分为凌晨，上午，中午，下午，晚上df['hour'] = pd.cut(df['time'],bins=[0,5,10,13,18,24],labels=['凌晨','上午','中午','下午','晚上'])df.head()
# 生成用户标签列表users = df['user_id'].unique()labels = pd.DataFrame(users,columns=['user_id'])labels
构建用户行为标签
# 对用户和时段进行分组，统计浏览次数time_browse = df[df['behavior_type']==1].groupby(['user_id','hour']).item_id.count().reset_index()time_browse.rename(columns={'item_id':'hour_counts'},inplace=True)time_browse.head()
# 统计每个用户浏览次数最多的时间段time_browse_max = time_browse.groupby('user_id').hour_counts.max().reset_index()time_browse_max.rename(columns={'hour_counts':'read_counts_max'},inplace=True)# 两个结果做连接time_browse = pd.merge(time_browse,time_browse_max,how='left',on='user_id')time_browse.head()
# 选取每个用户浏览次数最多的时间段time_browse_hour = time_browse.loc[time_browse['hour_counts']==time_browse['read_counts_max'],'hour']# 存在并列多个时用逗号连接，得到用户活跃时间段标签time_browse_hour = time_browse_hour.groupby(time_browse['user_id']).aggregate(lambda x:','.join(x)).reset_index()time_browse_hour.head()
# 将用户浏览活跃时间加入用户标签中labels = pd.merge(labels,time_browse_hour,how='left',on='user_id')labels.rename(columns={'hour':'time_browse'},inplace=True)labels.head()
#用户购买活跃时间段标签time_buy = df[df['behavior_type']==4].groupby(['user_id','hour']).item_id.count().reset_index().rename(columns={'item_id':'hour_counts'})time_buy_max = time_buy.groupby('user_id').hour_counts.max().reset_index().rename(columns={'hour_counts':'buy_counts_max'})time_buy = pd.merge(time_buy,time_buy_max,how='left',on='user_id')time_buy_hour = time_buy.loc[time_buy['hour_counts']==time_buy['buy_counts_max'],'hour']time_buy_hour = time_buy_hour.groupby(time_buy['user_id']).aggregate(lambda x:','.join(x)).reset_index()time_buy_hour.head()
# 将用户购买活跃时间段加入到用户标签表中labels = pd.merge(labels,time_buy_hour,how='left',on='user_id').rename(columns={'hour':'time_buy'})labels.head()
用户浏览最多的商品类别# 对用户与类别进行分组，统计浏览次数df_browse = df.loc[df['behavior_type']==1,['user_id','item_id','item_category']]df_cate_most_browse = df_browse.groupby(['user_id','item_category']).item_id.count().reset_index()df_cate_most_browse.rename(columns={'item_id':'item_category_counts'},inplace=True)# 每个用户浏览最多的商品类别df_cate_most_browse_max = df_cate_most_browse.groupby('user_id').item_category_counts.max().reset_index()df_cate_most_browse_max.rename(columns={'item_category_counts':'item_category_counts_max'},inplace=True)df_cate_most_browse = pd.merge(df_cate_most_browse,df_cate_most_browse_max,how='left',on='user_id')df_cate_most_browse['item_category'] = df_cate_most_browse['item_category'].astype(str)# 选取每个用户浏览次数最多的类别，存在并列时，用逗号连接df_cate_browse = df_cate_most_browse.loc[df_cate_most_browse['item_category_counts']==df_cate_most_browse['item_category_counts_max'],'item_category'].groupby(df_cate_most_browse['user_id']).aggregate(lambda x:','.join(x)).reset_index()# 将用户浏览量最多的类目加入到用户标签中labels = pd.merge(labels,df_cate_browse,how='left',on='user_id')labels.rename(columns={'item_category':'cate_most_browse'},inplace=True)labels.head()
