REDUCE 运算符中的小数位精度

Question

In my below code, I declared sum and lt_req-dmbtr as dmbtr , but in the output it is not showing the decimal values and output is also less than 2 or 3 signs than the expected

DATA(total) = REDUCE dmbtr( INIT sum = 0  FOR wa IN lt_req WHERE ( gsber = <fs_ba_det>-gsber AND gjahr = 2020 ) NEXT sum = sum + conv dmbtr( wa-dmbtr )   ).

Sometimes it is thworing the error

CX_SY_ARITHMETIC_OVERFLOW

Answer 1

The sum variable type is qualificative here, not the wa-dmbtr you convert from.

Excerpt from REDUCE help:

The declarations after INIT create local variables x1, x2...

After the full evaluation, the content of the first variable x1 or the memory area pointed to by the first field symbol is assigned to the temporary result of the expression of type type in accordance with the assignment rules .

And in your snippet your sum is zero, so it is Integer and it limits all other types.

The correct form of REDUCE to achieve your requirement:

TYPES:
  BEGIN OF ty_product,
    product TYPE char10,
    price   TYPE dmbtr,
    group   TYPE char10,
  END OF ty_product,
  tt_products TYPE STANDARD TABLE OF ty_product WITH EMPTY KEY.

DATA(t_prod) =
  VALUE tt_products(
    ( product = 'C0001' price = '1.005'     group = 'commodities' )
    ( product = 'C0002' price = '245.48575' group = 'crude' )
    ( product = 'C0003' price = '500.05'    group = 'crude' )
    ( product = 'C0004' price = '32'        group = 'commodities' )
  ).

DATA(total) = REDUCE dmbtr( INIT sum = CONV dmbtr( 0 ) FOR wa IN t_prod WHERE ( group = 'crude' ) NEXT sum = sum + wa-price ).

WRITE total.

That will give 745.53575 as a result.