How to split an SQL Table into half and send the other half of the rows to new columns with SQL Query? - sql

Question

Country Percentage
India 12%
USA 20%
Australia 15%
Qatar 10%
Output :
Country1 Percentage1 Country2 Percentage2
India 12% Australia 15%
USA 20% Qatar 10%
For example there is a table Country which has percentages, I need to divide the table in Half and show the remaining half (i.e. the remaining rows) in the new columns. I've also provided the table structure in text.

Answer 1

First, this type of operation should be done at the application layer and not in the database. That said, it can be an interesting exercise to see how to do this in the database.
I would use conditional aggregation or pivot. Note that SQL tables are inherently unordered. Your base table has no apparent ordering, so the values could come out in any order.
select max(case when seqnum % 2 = 0 then country end) as country_1,
max(case when seqnum % 2 = 0 then percentage end) as percentage_1,
max(case when seqnum % 2 = 1 then country end) as country_2,
max(case when seqnum % 2 = 1 then percentage end) as percentage_2
from (select c.*,
(row_number() over (order by (select null)) - 1) as seqnum
from country c
) c
group by seqnum / 2;

Answer 2

Try this
declare #t table
(
Country VARCHAR(20),
percentage INT
)
declare #cnt int
INSERT INTO #T
VALUES('India',12),('USA',20),('Australia',15),('Quatar',12)
select #cnt = count(1)+1 from #t
;with cte
as
(
select
SeqNo = row_number() over(order by Country),
Country,
percentage
from #t
)
select
*
from cte c1
left join cte c2
on c1.seqno = (c2.SeqNo-#cnt/2)
and c2.SeqNo >= (#cnt/2)
where c1.SeqNo <= (#cnt/2)

Answer 3

My variant
SELECT 'A' Country,1 Percentage INTO #Country
UNION ALL SELECT 'B' Country,2 Percentage
UNION ALL SELECT 'C' Country,3 Percentage
UNION ALL SELECT 'D' Country,4 Percentage
UNION ALL SELECT 'E' Country,5 Percentage
;WITH numCTE AS(
SELECT
*,
ROW_NUMBER()OVER(ORDER BY Country) RowNum,
COUNT(*)OVER() CountOfCountry
FROM #Country
),
set1CTE AS(
SELECT Country,Percentage,ROW_NUMBER()OVER(ORDER BY Country) RowNum
FROM numCTE
WHERE RowNum<=CEILING(CountOfCountry/2.)
),
set2CTE AS(
SELECT Country,Percentage,ROW_NUMBER()OVER(ORDER BY Country) RowNum
FROM numCTE
WHERE RowNum>CEILING(CountOfCountry/2.)
)
SELECT
s1.Country,s1.Percentage,
s2.Country,s2.Percentage
FROM set1CTE s1
LEFT JOIN set2CTE s2 ON s1.RowNum=s2.RowNum
DROP TABLE #Country

Answer 4

I just wanted to try something. I have used the function OFFSET. It does the requirement i think for your sample data, but dont know if its bulletproof all the way:
SQL Code
declare #myt table (country nvarchar(50),percentage int)
insert into #myt
values
('India' ,12),
('USA' ,20),
('Australia' ,15),
('Qatar' ,10),
('Denmark',10)
DECLARE #TotalRows int
SET #TotalRows = (select CEILING(count(*) / 2.) from #myt);
WITH dataset1 AS (
SELECT *,ROW_NUMBER() over(order by country ) as rn from (
SELECT Country,percentage from #myt a
ORDER BY country OFFSET 0 rows FETCH FIRST #TotalRows ROWS ONLY
) z
)
,dataset2 AS (
SELECT *,ROW_NUMBER() over(order by country ) as rn from (
SELECT Country,percentage from #myt a
ORDER BY country OFFSET #TotalRows rows FETCH NEXT #TotalRows ROWS ONLY
) z
)
SELECT * FROM dataset1 a LEFT JOIN dataset2 b ON a.rn = b.rn
Result

Answer 5

Assuming you want descending alphabetic country names, but the left column is determined by where India is located in the result:
with CoutryCTE as (
select c.*
, row_number() over (order by country)-1 as rn
from country c
)
, Col as (
select rn % 2 as num from CoutryCTE
where Country = 'India'
)
select max(case when rn % 2 = Col.num then country end) as country_1
, max(case when rn % 2 = Col.num then percentage end) as percentage_1
, max(case when rn % 2 <> Col.num then country end) as country_2
, max(case when rn % 2 <> Col.num then percentage end) as percentage_2
from CoutryCTE
cross join Col
group by rn / 2
;
SQLFiddle Demo
| country_1 | percentage_1 | country_2 | percentage_2 |
|-----------|--------------|-----------|--------------|
| India | 12% | Australia | 15% |
| USA | 20% | Qatar | 10% |
nb: this is extremely similar to an earlier answer by Gordon Linoff