大数据实战（五十一）：电商数仓（三十四）之系统业务数据仓库（七）数仓搭建（五）DWS层之用户行为宽表

1）为什么要建宽表

需求目标，把每个用户单日的行为聚合起来组成一张多列宽表，以便之后关联用户维度信息后进行，不同角度的统计分析

1 创建用户行为宽表

hive (gmall)>
drop table if exists dws_user_action;
create external table dws_user_action 
(   
    user_id          string      comment ‘用户 id‘,
    order_count     bigint      comment ‘下单次数 ‘,
    order_amount    decimal(16,2)  comment ‘下单金额 ‘,
    payment_count   bigint      comment ‘支付次数‘,
    payment_amount  decimal(16,2) comment ‘支付金额 ‘,
    comment_count   bigint      comment ‘评论次数‘
) COMMENT ‘每日用户行为宽表‘
PARTITIONED BY (`dt` string)
stored as parquet
location ‘/warehouse/gmall/dws/dws_user_action/‘;

2 向用户行为宽表导入数据

1）导入数据

hive (gmall)>
with 
tmp_order as
(
    select 
        user_id, 
count(*)  order_count,
        sum(oi.total_amount) order_amount
    from dwd_order_info oi
    where date_format(oi.create_time,‘yyyy-MM-dd‘)=‘2019-02-10‘
    group by user_id
) ,
tmp_payment as
(
    select
        user_id, 
        sum(pi.total_amount) payment_amount, 
        count(*) payment_count 
    from dwd_payment_info pi 
    where date_format(pi.payment_time,‘yyyy-MM-dd‘)=‘2019-02-10‘
    group by user_id
),
tmp_comment as
(
    select
        user_id,
        count(*) comment_count
    from dwd_comment_log c
    where date_format(c.dt,‘yyyy-MM-dd‘)=‘2019-02-10‘
    group by user_id
)

insert overwrite table dws_user_action partition(dt=‘2019-02-10‘)
select
    user_actions.user_id,
    sum(user_actions.order_count),
    sum(user_actions.order_amount),
    sum(user_actions.payment_count),
    sum(user_actions.payment_amount),
    sum(user_actions.comment_count)
from 
(
    select
        user_id,
        order_count,
        order_amount,
        0 payment_count,
        0 payment_amount,
        0 comment_count
    from tmp_order

    union all
    select
        user_id,
        0,
        0,
        payment_count,
        payment_amount,
        0from tmp_payment

    union all
    select
        user_id,
        0,
        0,
        0,
        0,
        comment_count
    from tmp_comment
 ) user_actions
group by user_id;

View Code

2）查询导入结果

hive (gmall)> select * from dws_user_action;

3 用户行为数据宽表导入脚本

1）在/home/atguigu/bin目录下创建脚本dws_db_wide.sh

[atguigu@hadoop102 bin]$ vim dws_db_wide.sh

在脚本中填写如下内容

#!/bin/bash

# 定义变量方便修改

APP=gmall

hive=/opt/module/hive/bin/hive

# 如果是输入的日期按照取输入日期；如果没输入日期取当前时间的前一天

if [ -n "$1" ] ;then

do_date=$1

else

do_date=`date -d "-1 day" +%F`  

fi

sql="

with  

tmp_order as

(

    select

        user_id,

        sum(oi.total_amount) order_amount,

        count(*)  order_count

    from "$APP".dwd_order_info  oi

    where date_format(oi.create_time,‘yyyy-MM-dd‘)=‘$do_date‘

    group by user_id

)  ,

tmp_payment as

(

    select

        user_id,

        sum(pi.total_amount) payment_amount,

        count(*) payment_count

    from "$APP".dwd_payment_info pi

    where date_format(pi.payment_time,‘yyyy-MM-dd‘)=‘$do_date‘

    group by user_id

),

tmp_comment as

(  

    select  

        user_id,

        count(*) comment_count

    from "$APP".dwd_comment_log c

    where date_format(c.dt,‘yyyy-MM-dd‘)=‘$do_date‘

    group by user_id

)

Insert overwrite table "$APP".dws_user_action partition(dt=‘$do_date‘)

select

    user_actions.user_id,

    sum(user_actions.order_count),

    sum(user_actions.order_amount),

    sum(user_actions.payment_count),

    sum(user_actions.payment_amount),

    sum(user_actions.comment_count)

from

(

    select

        user_id,

        order_count,

        order_amount,

        0 payment_count,

        0 payment_amount,

        0 comment_count

    from tmp_order

    union all

    select

        user_id,

        0,

        0,

        payment_count,

        payment_amount,

        0

    from tmp_payment

    union all

    select

        user_id,

        0,

        0,

        0,

        0,

        comment_count

    from tmp_comment

 ) user_actions

group by user_id;

"

$hive -e "$sql"

2）增加脚本执行权限

[atguigu@hadoop102 bin]$ chmod 777 dws_db_wide.sh

3）执行脚本导入数据

[atguigu@hadoop102 bin]$ dws_db_wide.sh 2019-02-11

4）查看导入数据

hive (gmall)> 

select * from dws_user_action where dt=‘2019-02-11‘ limit 2;

原文：https://www.cnblogs.com/qiu-hua/p/13547364.html