Data in the cube is showing multiple entries when compared with ODS

Question

Hello BW Gurus,
We have a waste report in production planning on Cube and ODS separately. The same info package loads both targets (which means same infosource) but when we run a report on Cube, the records are showing multiple entries (i.e. Key Figures are not matching when compared to ODS) where as the ODS records are showing correctly as it was in R/3. There are totally 6 key figures out of which 4 pulled from R/3 and 2 are populated in BW.
An Example:
Waste report in PP run for plant 1000 for 12/2005 and process order 123456. The operational scrap should be 2.46% and the component scrap should be 3.00% for material 10000000. The report is showing 7.87% for planned operational waste % and 9.6% for planned component waste %. These values are not correct. The ODS values for order 123456 matched the data in R/3 for component and operational scrap.
There is a Start routine to the ODS and also to the cube. I am not good at ABAP so requesting your Help.
Here is the ODS Code:
tables: /BI0/PPRODORDER.
loop at data_package.
select single COORD_TYPE
PRODVERS
into (/BI0/PPRODORDER-COORD_TYPE,
/BI0/PPRODORDER-PRODVERS)
from /BI0/PPRODORDER
where PRODORDER = data_package-PRODORDER
and OBJVERS = 'A'.
if sy-subrc = 0.
if /BI0/PPRODORDER-COORD_TYPE = 'XXXX'
or /BI0/PPRODORDER-COORD_TYPE = 'YYYY'.
data_package-PRODVERS = space.
else.
data_package-PRODVERS = /BI0/PPRODORDER-PRODVERS.
endif.
endif.
if data_package-calday = space
or data_package-calday = '00000000'.
if data_package-TGTCONSQTY NE 0.
data_package-calday = data_package-ACTRELDATE.
endif.
endif.
modify data_package.
endloop.
Here is Cube Code:
tables: /BI0/PPRODORDER,
/BIC/ODS.
TYPES:
BEGIN OF ys_mat_unit,
material TYPE /bi0/oimaterial,
mat_unit TYPE /bi0/oimat_unit,
numerator TYPE /bi0/oinumerator,
denomintr TYPE /bi0/oidenomintr,
END OF ys_mat_unit.
DATA:
l_s_mat_unit TYPE ys_mat_unit,
e_factor type p decimals 5.
loop at data_package.
select single COORD_TYPE
PRODVERS
into (/BI0/PPRODORDER-COORD_TYPE,
/BI0/PPRODORDER-PRODVERS)
from /BI0/PPRODORDER
where PRODORDER = data_package-PRODORDER
and OBJVERS = 'A'.
if sy-subrc = 0.
if /BI0/PPRODORDER-COORD_TYPE = 'XXX'
or /BI0/PPRODORDER-COORD_TYPE = 'YYY'.
data_package-PRODVERS = space.
else.
data_package-PRODVERS = /BI0/PPRODORDER-PRODVERS.
endif.
endif.
if data_package-calday = space
or data_package-calday = '00000000'.
if data_package-TGTCONSQTY NE 0.
data_package-calday = data_package-ACTRELDATE.
endif.
endif.
data_package-agsu = 'GSU'.
data_package-agsu_qty = 0.
select single gr_qty
base_uom
into (/BIC/ODS-gr_qty,
/BIC/ODS-base_uom)
from /BIC/ODS
where prodorder = data_package-prodorder
and material = data_package-material.
if sy-subrc = 0.
if /BIC/ODS-base_uom = 'GSU'.
data_package-agsu_qty = /BIC/ODS-gr_qty.
else.
SELECT SINGLE * FROM /bi0/pmat_unit
INTO CORRESPONDING FIELDS OF l_s_mat_unit
WHERE material = data_package-material
AND mat_unit = 'GSU'
AND objvers = 'A'.
IF sy-subrc = 0.
IF l_s_mat_unit-denomintr <> 0.
e_factor = l_s_mat_unit-denomintr /
l_s_mat_unit-numerator.
multiply /BIC/ODS-gr_qty by e_factor.
data_package-agsu_qty = /BIC/ODS-gr_qty.
ENDIF.
else.
CALL FUNCTION 'UNIT_CONVERSION_SIMPLE'
EXPORTING
INPUT = /BIC/ODS-gr_qty
NO_TYPE_CHECK = 'X'
ROUND_SIGN = ' '
UNIT_IN = /BIC/ODS-base_uom
UNIT_OUT = 'GSU'
IMPORTING
OUTPUT = DATA_PACKAGE-gsu_qty
EXCEPTIONS
CONVERSION_NOT_FOUND = 1
DIVISION_BY_ZERO = 2
INPUT_INVALID = 3
OUTPUT_INVALID = 4
OVERFLOW = 5
TYPE_INVALID = 6
UNITS_MISSING = 7
UNIT_IN_NOT_FOUND = 8
UNIT_OUT_NOT_FOUND = 9
OTHERS = 10.
endif.
endif.
endif.
modify data_package.
endloop.
some how the AGSU qyt is not populating in the cube and when I dbug the code, I could see a clean record in the internal table but not in the cube.
your suggestion and solutions would be highly appreciated.
thanks,
Swathi.

Answer 1

Hi Swathi
In ODs we have option of overwriting and addition however in Cube we have only adition.Thats why you are getting multiple enteries.
If you are running daily full load on the cube then please delete the earlier requests.
So at one point of time there should be only one full load request in cube. Hope this will solve your problem.
Regards,
Monika