CODEWITHSUNDEEP
sqlpostgresql
rockstardev
rockstardev

Reputation: 13527

Merging SQL data?

This seems so simple, but I can't seem to figure out how to do it. I have two data sets:

SET1
DATE       | TOTAL1 | TOTAL2 | TOTAL3
1 Jun 2013 | 0      | 0      | 5
2 Jun 2013 | 0      | 0      | 12
3 Jun 2013 | 0      | 0      | 34
4 Jun 2013 | 0      | 0      | 50

SET2
DATE       | TOTAL1 | TOTAL2 | TOTAL3
1 Jun 2013 | 1      | 2      | 0
2 Jun 2013 | 4      | 12     | 0
3 Jun 2013 | 5      | 12     | 0
4 Jun 2013 | 6      | 10     | 0

I want to create a third dataset the merges these two sets into the following:

SET3
DATE       | TOTAL1 | TOTAL2 | TOTAL3
1 Jun 2013 | 1      | 2      | 5
2 Jun 2013 | 4      | 12     | 12
3 Jun 2013 | 5      | 12     | 34
4 Jun 2013 | 6      | 10     | 50

Joining the tables does not work. I need to join them in a way that will add the totals if the dates match up. Any idea how to do this?

Upvotes: 0

Views: 58

Answers (5)

Iceman
Iceman

Reputation: 21

You can use something similar:

INSERT INTO set3 (date,Total1,Total2,Total3)
    SELECT s1.date
       ,case when s1.date=s2.date then s1.Total1+s2.Total1 end as Total1
       ,case when s1.date=s2.date then s1.Total2+s2.Total2 end as Total2
       ,case when s1.date=s2.date then s1.Total3+s2.Total3 end as Total3
    FROM set1 s1, set2 s2
    WHERE s1.date=s2.date

It works only when you want to add columns with same date. It will not handle rows where the date is different...

Upvotes: 0

Olaf H
Olaf H

Reputation: 496

Use UNION ALL to create 1 table of both sets. Then GROUP BY date and SUM up all totals.

create table set1 (
 d date,
 total1 number,
 total2 number,
 total3 number
);

create table set2 (
 d date,
 total1 number,
 total2 number,
 total3 number
);

insert into set1 (d, total1, total2, total3) values (to_date('01.06.2013', 'dd.mm.yyyy'), 0,0,5);
insert into set1 (d, total1, total2, total3) values (to_date('02.06.2013', 'dd.mm.yyyy'), 0,0,12);
insert into set1 (d, total1, total2, total3) values (to_date('03.06.2013', 'dd.mm.yyyy'), 0,0,34);
insert into set1 (d, total1, total2, total3) values (to_date('04.06.2013', 'dd.mm.yyyy'), 0,0,50);

insert into set2 (d, total1, total2, total3) values (to_date('01.06.2013', 'dd.mm.yyyy'), 1,2,0);
insert into set2 (d, total1, total2, total3) values (to_date('02.06.2013', 'dd.mm.yyyy'), 4,12,0);
insert into set2 (d, total1, total2, total3) values (to_date('03.06.2013', 'dd.mm.yyyy'), 5,12,0);
insert into set2 (d, total1, total2, total3) values (to_date('04.06.2013', 'dd.mm.yyyy'), 6,10,0);

commit;

select d, sum(total1) as total1, sum(total2) as total2, sum(total3) as total3 from (
  select d, total1, total2, total3 from set1
  union all
  select d, total1, total2, total3 from set2
) group by d
order by d;

Upvotes: 0

George
George

Reputation: 1054

SELECT
     DATE,
     SUM(TOTAL1) AS TOTAL1,
     SUM(TOTAL2) AS TOTAL2,
     SUM(TOTAL3) AS TOTAL3
FROM
(
     SELECT
          DATE,
          TOTAL1,
          TOTAL2,
          TOTAL3
     FROM
         SET1

     UNION ALL

     SELECT
          DATE,
          TOTAL1,
          TOTAL2,
          TOTAL3
     FROM
         SET2
) SubQueryAlias
GROUP BY
     DATE

Upvotes: 5

Lamak
Lamak

Reputation: 70638

I'm guessing that you want a FULL JOIN:

SELECT  COALESCE(T1.DATE,T2.DATE) AS DATE,
        COALESCE(T1.TOTAL1,0)+COALESCE(T2.TOTAL1,0) AS TOTAL1,
        COALESCE(T1.TOTAL2,0)+COALESCE(T2.TOTAL2,0) AS TOTAL2,
        COALESCE(T1.TOTAL3,0)+COALESCE(T2.TOTAL3,0) AS TOTAL3
FROM Table1 T1
FULL JOIN Table2 T2
    ON T1.DATE = T2.DATE

Upvotes: 2

Daniel A. White
Daniel A. White

Reputation: 190944

You still want to do a join but name the columns explicitly, something like this.

SELECT Date, T1.Total1 + T2.Total1 AS TOTAL1, ...
FROM T1 JOIN T2 ON T1.Date = T2.Date

Upvotes: 0

Related Questions