我正在运行一个查询,以获取两个时段(7月和8月)中特定客户的总数。然后将两者相减得出差值。但是,即使该客户端的Jul和Aug值都存在,我在减去时也会得到空值。
即使我知道值/数字存在,select语句的这一部分仍将返回空值。 “ AUG.amount-JUL.amount作为Final_Result”
我尝试分别运行Jul和Aug表,它们确实返回了AH的总和。
SELECT AUG.party_name, AUG.amount - JUL.amount as Final_Result
FROM (select cl.party_name, sum(cl.amount)/1000 as amount
from CLIENT_PROFIT CL
where to_char(trade_date,'YYYYMMDD') between '20190801' and '20190830'
AND cl.party_name = 'AH'
group by cl.party_name
) AUG
LEFT JOIN
(select cl.party_name, sum(cl.amount)/1000 as amount
from CLIENT_PROFIT CL
where to_char(trade_date,'YYYYMMDD') between '20190701' and '20190731'
AND cl.party_name = 'AH'
group by cl.party_name) JUL
ON AUG.party_name = JUL.party_name
group by AUG.party_name
order by Final_Result DESC
预期结果应显示AH的8月和7月之间的差异
答案 0 :(得分:2)
您可以使用以下方法直接获取它: 编辑查询以按月使用trunc。
select cl.party_name,
(sum(case when trunc(trade_date,'month') = date '2019-08-01' then cl.amount end)
- sum(case when trunc(trade_date,'month') = date '2019-07-01' then cl.amount end) ) /1000 as amount
from CLIENT_PROFIT CL
where trunc(trade_date) between date '2019-07-01' and date '2019-08-31'
AND cl.party_name = 'AH'
group by cl.party_name;
此外,您忘记在查询中包括2019年8月31日。
干杯!
答案 1 :(得分:2)
您可以使用Conditional Aggregation
而不使用group by
表达式,因为您只有一个参与方名称:
select 'AH' as party_name,
sum(case when to_char(trade_date,'YYYYMM')='201908' then cl.amount/1000 end) -
sum(case when to_char(trade_date,'YYYYMM')='201907' then cl.amount/1000 end)
as Final_Result
from CLIENT_PROFIT CL
where cl.party_name = 'AH'