Reputation: 3
I am making a table with amount of transactions from other banks. First row will be the total one. First column with absolute numbers, second with % of the total amount and third, last column, will be with name of the senders bank.
eg:
TRN | % | BANK |
---|---|---|
8000 | 100% | ALL |
4000 | 50% | BANK1 |
2000 | 25% | BANK2 |
2000 | 25% | BANK3 |
I have trouble getting the % column as in my script the data are groupped already.
SELECT COUNT(*)
,'100%' as %,
,'ALL' as BANK
FROM table A
UNION ALL
SELECT COUNT(*)
,**IDK**
,BANK_ID BANK
FROM TABLE2 B
GROUP BY A.BANK_ID
ORDER BY 1 DESC
I know that there should be possibility with the over partition, but I can't figure out how.
Thanks for any tips!
Upvotes: 0
Views: 80
Reputation: 7801
The MODEL clause could be used for this:
-- S a m p l e d a t a
WITH
tbl (BANK_ID, TRN) AS
(
Select 'Bank 1', 500 From Dual Union All
Select 'Bank 1', 3500 From Dual Union All
Select 'Bank 2', 1200 From Dual Union All
Select 'Bank 2', 800 From Dual Union All
Select 'Bank 3', 2000 From Dual
)
-- M a i n S Q L
SELECT TRN, PCT, BANK_ID
FROM ( SELECT BANK_ID "BANK_ID", Sum(TRN) "TRN"
FROM tbl
GROUP BY BANK_ID
ORDER BY BANK_ID
)
MODEL Dimension By(BANK_ID )
Measures(TRN, 0 as PCT)
RULES
( TRN['ALL'] = Sum(TRN)[BANK_ID != 'ALL'],
TRN[ANY] = Sum(TRN)[CV()],
PCT['ALL'] = 100,
PCT[ANY] = Sum(TRN)[CV()] * 100 / Sum(TRN)[BANK_ID != 'ALL']
)
ORDER BY BANK_ID
-- R e s u l t
TRN PCT BANK_ID
---------- ---------- -------
8000 100 ALL
4000 50 Bank 1
2000 25 Bank 2
2000 25 Bank 3
Upvotes: 0
Reputation: 16001
Sample data:
create table demo (trn number, bank_id varchar2(10));
insert all
into demo values (1000, 'BANK1')
into demo values (1000, 'BANK1')
into demo values (1000, 'BANK1')
into demo values (1000, 'BANK1')
into demo values (1000, 'BANK2')
into demo values (1000, 'BANK2')
into demo values (1000, 'BANK3')
into demo values (1000, 'BANK3')
select * from dual;
Query:
select sum(trn)
, 100 * ratio_to_report(sum(trn)) over () * 2 as percent
, nvl(bank_id,'ALL') as bank
from demo
group by rollup (bank_id);
SUM(TRN) PERCENT BANK
---------- ---------- ----------
4000 50 BANK1
2000 25 BANK2
2000 25 BANK3
8000 100 ALL
group by rollup()
generates a total row.
ratio_to_report()
returns a value between 0 and 1, so I multiply by 100 to present it as a percentage.
The * 2
is because ratio_to_report()
includes the whole column including the rollup total, so every value is halved.
I've assumed bank_id
cannot be null and so a null value in the report must be the total row. If you need to distinguish between the total row and other null values, you can use grouping(bank_id)
in a case
expression, and it will return 1 for the total row and 0 for the rest. (You could also use this in an order by
clause if you want to display the total row first.)
Upvotes: 1
Reputation: 3
I got to this solution and it works thankfully,
with alll as
(
SELECT
COUNT(*) trn
,'100%' AS prcnt
,'ALL' AS BANK_ID
FROM table
)
, bank as
(
SELECT distinct
count(1) over (partition by BANK_ID) cnt_B
,to_char(round(count(1) over (partition by BANK_ID)/count(*) over (partition by 1),3)*100) || '%' as prcnt
,BANK_ID
FROM table
)
select * from alll
UNION ALL
select * from bank
Upvotes: 0
Reputation: 762
I am not overly familiar with Oracle syntax so the below may need some adjustments, but it should be portable enough. This is in MS SQL. Apologies for not being able to provide you a solution in the direct syntax you need. If someone is more familiar, please feel free to edit.
DECLARE @totalCount BIGINT = (SELECT Count(*) FROM table);
SELECT
@TotalCount AS [Count],
FORMAT(1, 'P') AS [%],
'ALL' AS [BANK]
--removed table here as the only derived value is the count and we already have it in @totalCount
UNION ALL
SELECT
COUNT(*) AS [Count],
FORMAT(CAST(Count(*) AS DECIMAL) / @TotalCount,'P') AS [%],
BANK_ID AS [BANK]
FROM [tableName]
GROUP BY [tableName].BANK_ID
--ORDER BY 1
Upvotes: 0
Reputation: 11
Hi I think this should do the trick.
You can use the over partition with specifying anything in the brackets.
SELECT COUNT(*)
,'100%' as '%',
,'ALL' as BANK
UNION ALL
SELECT COUNT(*)
,cast(count(*)*100/count(*) over () as varchar) +'%'
,BANK_ID BANK
FROM TABLE2 B
GROUP BY B.BANK_ID
ORDER BY 1 DESC
Upvotes: 1