获取一个表的所有行，并从另一个表中匹配行，其中SQL中可能不存在条目

Question

so what I have is a database where records have an ID and a YEAR associated with them. Each YEAR will only have one record for each ID , but every ID doesn't necessarily exist in a given YEAR . So, what I'm trying to do is get a list of every record from my table for the 2018 YEAR , and get a certain value from the three previous years. My issue is that if a record from 2018 didn't exist in 2017, I won't get it at all. I want to just fill in NULL fields with a 0 or something, but still keep that ID .

What I have so far (which obviously doesn't work) is:

SELECT a.ID, b.VAL, c.VAL, d.VAL
FROM table as a, table as b, table as c, table as d
WHERE a.YEAR = 2018 AND b.YEAR = 2017 AND c.YEAR = 2016 AND d.YEAR = 2015 AND a.ID = b.ID AND b.ID = c.ID AND c.ID = d.ID;

I tried doing a join like:

SELECT a.ID, b.VAL FROM table as a LEFT JOIN table as b on a.ID = b.ID WHERE...

but that still showed the same issue. I should also mention I'm doing this in MS ACCESS.

Answer 1

Hopefully I understand the data structure you're describing. The variable table below is my interpretation. If this is correct then the code below it should give you what you're looking for.

The key difference is the 'AND' in the 'ON' clause vs in the 'WHERE'. If you put the criteria in the 'WHERE' then you're essentially turning your LEFT OUTER JOIN into an INNER JOIN by asserting that the criteria must be true after the join is completed. Whereas in the 'ON' you're telling SQL that it must be true in order to do the join itself.

declare @tbl table ([year] int, [id] int, [someval] varchar(50))
insert into @tbl values (2018, 1, 'one')
insert into @tbl values (2018, 2, 'two')
insert into @tbl values (2017, 1, 'three')
insert into @tbl values (2016, 1, 'four')
insert into @tbl values (2015, 1, 'five')
insert into @tbl values (2015, 2, 'six')
insert into @tbl values (2015, 3, 'seven')

;with t2018 as(select * from @tbl where [year]=2018)
,t2017 as(select * from @tbl where [year]=2017)
,t2016 as(select * from @tbl where [year]=2016)
,t2015 as(select * from @tbl where [year]=2015)
select isnull(t2018.id,isnull(t2017.id,isnull(t2016.id,t2015.id))) as id,
     t2018.someval as [2018SomeVal],
     t2017.someval as [2017SomeVal], 
     t2016.someval as [2016SomeVal], 
     t2015.someval as [2015SomeVal]
from t2018
    full outer join t2017 on t2017.id = t2018.id
    full outer join t2016 on t2016.id = t2018.id
    full outer join t2015 on t2015.id = t2018.id

Answer 2

I would just use conditional aggregation:

SELECT a.ID,
       SUM(IIF(a.YEAR = 2018, a.VAL, 0)) as val_2018,
       SUM(IIF(a.YEAR = 2017, a.VAL, 0)) as val_2017,
       SUM(IIF(a.YEAR = 2016, a.VAL, 0)) as val_2016,
       SUM(IIF(a.YEAR = 2015, a.VAL, 0)) as val_2015
FROM table as a
WHERE a.YEAR IN (2015, 2016, 2017, 2018)
GROUP BY a.id;

If you only want rows where 2018 is present, then include:

HAVING MAX(a.YEAR) = 2018