how to divide the obtained sum of data - sql

Question

I need to get the sum of the records matching the key and the result of it is my TOTALQUANTITY
then i need to split it for packages, max package size is 5.
so when i get a result(TOTALQUANTITY) equals to 13 i should get something like this:
package |ofpackages |totalquantity |quantityofpackage
1 |3 |13 |5
2 |3 |13 |5
3 |3 |13 |3
my attempt:
SELECT
count(*) as TOTALQUANTITY,
get_token(data1) data1,
get_token(data2) data2,
get_token(data3) data3,
floor((row_number() over (partition BY get_token(data2) order by get_token(data2) ) - 1 ) / 5) package
FROM
--my working code
WHERE
--my working conditions
GROUP BY
get_token(data1),
get_token(data2),
get_token(data3)
ORDER BY
get_token(data2)
Totalquantity gives the correct value, but the package unfortunately doesn't :(
How it should be? how can i get the remaining values?

Answer 1

Add the use MOD() to determine the remainder, if zero output the division, otherwise add the package amount to the totalquantity then divide and round.
Using a derived table makes this a little easier I believe:
SELECT
TOTALQUANTITY,
get_token(data1) data1,
get_token(data2) data2,
get_token(data3) data3,
package,
case when MOD(TOTALQUANTITY,package) = 0
then TOTALQUANTITY + package
else round((TOTALQUANTITY + package)/package,0)
end as quantityofpackage
FROM (
SELECT
count(*) as TOTALQUANTITY,
get_token(data1) data1,
get_token(data2) data2,
get_token(data3) data3,
floor((row_number() over (partition BY get_token(data2) order by get_token(data2) ) - 1 ) / 5) package
FROM
--my working code
WHERE
--my working conditions
GROUP BY
get_token(data1),
get_token(data2),
get_token(data3)
) derived
ORDER BY
data2
Note you don't specify which dbms you are using so I have assumed it supports MOD() if not it will have an equivalent e.g. % in MS SQL Server

Answer 2

I'm not sure from the example code how that's supposed to work. Suppose there's a table which contains packages and quantities.
drop table if exists #package_quantities;
go
create table #package_quantities(
package int not null,
quantity int not null);
go
insert #package_quantities(package, quantity) values
(10, 10),
(10, 1),
(10, 2),
(20, 5),
(20, 9),
(30, 12),
(40, 20);
To generate new rows the query uses a tally tvf called fnNumbers. The CTE pack_cte sums quantitites for each package. The query then splits the packages into bundles with a maximum quantity of 5. Each subpackage has a unique sequence number 1, 2, 3...
fnNumbers tvf
create function [dbo].[fnNumbers](
#zero_or_one bit,
#n bigint)
returns table with schemabinding as return
with n(n) as (select null from (values (1),(2),(3),(4)) n(n))
select 0 n where #zero_or_one = 0
union all
select top(#n) row_number() over(order by (select null)) n
from n na, n nb, n nc, n nd, n ne, n nf, n ng, n nh,
n ni, n nj, n nk, n nl, n nm, n np, n nq, n nr;
Query
with pack_cte(package, sum_quantity) as (
select package, sum(quantity)
from #package_quantities
group by package)
select p.package, p.sum_quantity, p_calc.num_packages, fn.n subpackage,
case when p_calc.num_packages<>fn.n or (p_calc.num_packages=fn.n and p.sum_quantity%5=0)
then 5 else p.sum_quantity%5 end quanityofpackage
from pack_cte p
cross apply (select ceiling(p.sum_quantity/5.0) num_packages) p_calc
cross apply fnNumbers(1, p_calc.num_packages) fn;
Output
package sum_quantity num_packages subpackage quanityofpackage
10 13 3 1 5
10 13 3 2 5
10 13 3 3 3
20 14 3 1 5
20 14 3 2 5
20 14 3 3 4
30 12 3 1 5
30 12 3 2 5
30 12 3 3 2
40 20 4 1 5
40 20 4 2 5
40 20 4 3 5
40 20 4 4 5