Show TOP 4 rows into 4 columns in SQL SELECT Query - mysql

Question

I need to show values from one column in a SQL tabe into 4 columns according its top 4 values.
Example current table:
----------------------------------
ID | Name | Amount
1 | Test | 80
1 | Test2| 70
1 | Test3| 40
1 | Any | 25
1 | Any1 | 15
1 | Any2 | 12
1 | Any3 | 5
2 | TS1 | 70
2 | TS2 | 55
2 | TS3 | 30
2 | TS4 | 19
2 | Any | 11
--------------------------
Example expected SELECT Query result:
----------------------------------
ID | Col1 | Col2 | Col3 | Col4
1 | 80 | 70 | 40 | 25
2 | 70 | 55 | 30 | 19
----------------------------------
The issue here is to group the top 4 amount in 4 columns not considering names, just the numbers.
Is there some way to reach this result in table like that?

Answer 1

Please try this pseudocode. Where data retrieves as per given ordering in sample input or storing position in database.
SELECT t.id
, MAX(CASE WHEN row_num = 1 THEN t.amount END) col1
, MAX(CASE WHEN row_num = 2 THEN t.amount END) col2
, MAX(CASE WHEN row_num = 3 THEN t.amount END) col3
, MAX(CASE WHEN row_num = 4 THEN t.amount END) col4
FROM (SELECT id
, amount
, ROW_NUMBER() OVER (PARTITION BY id) row_num
FROM test) t
WHERE t.row_num <= 4
GROUP BY t.id;
Please check from url https://dbfiddle.uk/?rdbms=mysql_8.0&fiddle=8d3b9a3da33f9177f2bb6957ef08e21b
Also you can use amount in ORDER BY clause in descending order as per expected result if needed. Try this pseudocode then.
SELECT t.id
, MAX(CASE WHEN row_num = 1 THEN t.amount END) col1
, MAX(CASE WHEN row_num = 2 THEN t.amount END) col2
, MAX(CASE WHEN row_num = 3 THEN t.amount END) col3
, MAX(CASE WHEN row_num = 4 THEN t.amount END) col4
FROM (SELECT id
, amount
, ROW_NUMBER() OVER (PARTITION BY id ORDER BY amount DESC) row_num
FROM test) t
WHERE t.row_num <= 4
GROUP BY t.id;

Answer 2

FIRST STEP: Create Ranking (1 to 4) with order by DESC amount as TOP 4 will be highest to lowest
Second Step: USE MAX() on ranking to get all the rankings in one row rather than diagonally
WITH data as (
SELECT
ID,
AMOUNT,
ROW_NUMBER() OVER (PARTITION BY ID ORDER BY AMOUNT DESC) RANKING
FROM [TABLE NAME]
)
SELECT
ID,
MAX(CASE WHEN RANKING = 1 THEN AMOUNT ELSE NULL END) COL_1,
MAX(CASE WHEN RANKING = 2 THEN AMOUNT ELSE NULL END) COL_2,
MAX(CASE WHEN RANKING = 3 THEN AMOUNT ELSE NULL END) COL_3,
MAX(CASE WHEN RANKING = 4 THEN AMOUNT ELSE NULL END) COL_4
FROM data
GROUP BY 1

Answer 3

Alternatively you can use window functionality to achieve this result.
Below is the query written in postgresql
select
id,col1,col2,col3,col4
from
(
SELECT
tbl.id
,tbl.amount AS col1 -- 1st
,LEAD(tbl.amount, 1) OVER (ORDER BY id ASC) as col2 --2nd
,LEAD(tbl.amount, 2) OVER (ORDER BY id ASC) as col3 --3rd
,LEAD(tbl.amount, 3) OVER (ORDER BY id ASC) as col4 --4th
,ROW_NUMBER() OVER (PARTITION BY id order by amount desc) rn
FROM
(
SELECT
id
, amount
, ROW_NUMBER() OVER (PARTITION BY id order by amount desc) rn
FROM <TABLE_NAME>
) tbl
WHERE 1=1 and tbl.rn <= 4 -- 4: variable
)tbl2
where 1=1 and tbl2.rn =1 -- 1: fixed
;
LEAD(column, n) returns column's value at the row n rows aer the current row