Sum value in aggregator

Hi Ipsita, When you use the ABORT function the error message we define will be shown in log. Is that fine, or you need it in different way(like into a table or file)? Thanks,Deeshan.

Hi Ipsita, You can also try the below approach as suggested by my colleague(Krishna Charan). Pipeline1 SQ ----> AGG1--->AGG2--->EXP(ABORT func)-->Filter-->Dummytarget Pipeline2 SQ----->EXP------>Actual Target In target load order at mapping level, keep pipeline1 as first to execute. Thanks,Deeshan.

Similar Messages

Error in printing sum value of DME file in Automatic Payment Run

Hi All,
We are in the process of upgrading from R/3-4.0B to ECC6.0.
In Automatic Payment run in transaction F110, when the spool is generated, the sum values is displaying in correctly. It is displaying and printing as actual value/100 for example if total of all the line items is 1000000, the sum value displaying as 10000 and in words also prints as ten thousands.
Kindly let us know how to fix the issue, if any one has come across with similar issue.
Thanx & Regards
Ashok

Hi VVR,
I got the following errors:
Err     Message text
003    Item is blocked for payemnt
007    Error in creating the payment doc, read job log.....
But if we have a look at the job log it is showing for all the 50,000 items... how to find where the problem is???
thanks.

SUM VALUES

Hi
I need to sum values for different CD_INDICATOR, same CD_CUSTOMER , same NM_CYCLE , same CD_STRUT, same PROFILE
CD_STRUT                    NUMBER(6)
NM_CYCLE                    NUMBER(6)
CD_CUSTOMER                 NUMBER(10)
CD_PROFILE                  NUMBER(2)
CD_INDICATOR                NUMBER(6)
VL_INDICATOR                NUMBER(15,2)
LAST_DATE                   DATE
SELECT CD_STRUT ,   NM_CYCLE ,
    CD_CUSTOMER , CD_PROFILE , CD_INDICATOR ,
   VL_INDICADOR VL_INDICATOR , DT_ULTIMA_ATUALIZACAO LAST_DATE
   FROM MYTABLE PPC
   WHERE PPC.CD_TYPE_STRUCT = 5   ---IS CONSTANT = 5
    AND PPC.CD_STRUT = 4
    AND PPC.NM_CYCLE = 200810
    AND PPC.CD_CUSTOMER = 574970
    AND PPC.CD_PROFILE = 1
    AND PPC.CD_INDICATOR IN (4,2);     CD_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR LAST_DATE
       4   200810      574970          1            2              14.00 6/30/2008 8
       4   200810      574970          1            4              1.00 6/30/2008 8     I must to sum VL_INDICATOR , with rule
     When CD_INDICATOR = 4 AND VL_INDICATOR =1.00 I must return sum
     of VL_INDICATOR 2 ( In case above =14) with 1, but if when
     CD_INDICATOR = 4 AND VL_INDICATOR != 1 I must return only value of of VL_INDICATOR 2 ( In case above = 14)
     I must return new CD_INDICATOR (5) as with below
CD_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR LAST_DATE
       4   200810      574970          1          5              15.00 6/30/2008 8     if the values of CD_INDICATOR (2,4)
CD_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR LAST_DATE
       4   200810      574970          1            2              12.00 6/30/2008 8
       4   200810      574970          1            4              4.00 6/30/2008 8RETURN ME
CD_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR LAST_DATE
       4   200810      574970          1          5              12.00 6/30/2008 8

Hi
I must to sum values OF CD_INDICATOR 309 and 4 ,
When value of CD_INDICATOR 4 is equal 1 (One) then I sum 1 else I sum zero
But if record of CD_INDICATOR 4 not exists I must to sum 0
In example below the sum equal 10 because value of CD_INDICATOR 4 is equal 2
CD_STRUT CD_TYPE_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR
       2             5   200810     3320080          1          309             10.00
       2             5   200810     3320080          1            4              2.00 In example below the sum equal 15
CD_STRUT CD_TYPE_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR
       2             5   200810     3320080          1          309             14.00
       2             5   200810     3320080          1            4              1.00 In example below the sum is 10, because no exist cd_indicator 4
CD_STRUT CD_TYPE_STRUT NM_CYCLE CD_CUSTOMER CD_PROFILE CD_INDICATOR      VL_INDICATOR
       2             5   200810     3320080          1          309             10.00
   SELECT c.CD_STRUT,
          cCD_TYPE_STRUT,
          cNM_CYCLE ,
          cCD_CUSTOMER,
          cCD_PROFILE ,
          C.CD_INDICATOR,
          c.VL_INDICATOR
     from T_CUSOTMER a
     INNER JOIN MYTABLE c ON (    a.cd_type_strut = c.cd_type_strut
        AND a.cd_estru = c.cd_estrut
        AND a.CUSTOMER = c.cd_CUSOTMER
        AND c.nm_cyicle = 200810
        AND c.cd_profile = 1
        AND c.cd_indicator = 309
       and c.cd_customer = 3320080)I Want to a OUTER LEFT JOIN with MYTABLE for to get CD_INDICATOR =4 , but I did know how can I do it
Message was edited by:
muttleychess

Sum Values on 2 Where Clauses

Hi, sorry to trouble, i havent got any data within a database or anything, just thinking about some work that i might have to do, so trying to get a head start. Was wondering how i would sum values in a column but with 2 where clauses. For example if i wanted to sum all values until 6pm in 1 column and all values after 6pm for 1 day how i would write this out.
I know how to write them separately, i.e.
select name, sum(values) as after6
from table
where time > '18:00'
group by name;
select name, sum(values) as upto6
from table
where time < '18:00'
group by name;
how can i combine these to have a columns after6 and upto6 next to each other.
if somebody could please advise.
thanks in advance

Using case you can split the column then sum the column .Try the below
select name,
SUM(CASE WHEN time > '18:00' THEN values ELSE 0 END) as after6,
SUM(CASE WHEN time < '18:00' THEN values ELSE 0 END) as upto6
from table
groupbyname;

Sum values in last row for date format "hh24:mi"

Hi,
I have a big problem creating a sum row with date data.
I created a sql view where the result looks like that:
SELECT category_name,
user_name,
b_mmyyyy,
b_1, ..., b_31
FROM view_category_user_booking
category_name, user_name, b_1, ..., b_31
category1, user1, 122008, 01:00, ..., 03:30
category2, user1, 122008, 02:00, ..., 01:00
category3, user1, 122008, 00:00, ..., 00:15My goal is to integrate a sum row at the end for the columns b_1 ... b_31
category_name, user_name, b_1, ..., b_31
category1, user1, 122008, 01:00, ..., 03:30
category2, user1, 122008, 02:00, ..., 01:00
category3, user1, 122008, 00:00, ..., 00:15
Sum, user1, 122008, 03:00, ..., 04:45I tried it like that:
select decode(grouping(category_name),1,'Summe',category_name),
sum(to_number(replace(b_1,':',','))) as b_1,
sum(to_number(replace(b_31,':',','))) as b_31
from category_user_booking
where user_name = 'user1'
and b_mmyyyy = '122008'
group by rollup(category_name)But the result isn't really successful
category_name, b_1, b_31
category1, 1, 3,3
category2, 2, 1
category3, 0, 0,15
Sum, 3, 4,45Is there somebody with an idea?
Thanks ahead,
Tobias

Hi Avinash,
the problem is not the grouping it is the date format.
Instead of a value like 5,4 > I want 05:40 or 1,75 > I want 02:15 in the sum value.
In my table all time values are saved in the format hh24:mi.
I don't know how to get that via a sql statement.
Output I want:
category_name, user_name, b_1, ..., b_31
category1, user1, 122008, 01:00, ..., 03:30
category2, user1, 122008, 02:00, ..., 01:00
category3, user1, 122008, 00:00, ..., 00:15
Sum, user1, 122008, 03:00, ..., 04:45Regards,
Tobias

SUM value by multiple conditions in SSAS

Hi All,
I have a condition like as below, but the SUM value not working in CASE STATEMENT. Anyone can help me...!
SUM(CASE [Calendar].[Calendar Date]
WHEN [Calendar].[Calendar Date].[Year] = '2005' AND [Calendar].[Calendar Date].[Month] = 1
THEN [Calendar].[Calendar Date].CURRENTMEMBER.CHILDREN
END
, [Measures].[Work Day Flag])
Thanks in Advance...!
Baskaran

That is quite right Simon, although, technically, it's not a sum, you are simply getting the internet sales for january 2011, and you are overriding whatever selection is on the calendar date hierarchy. However, if that is what you want the calc to return,
perfect.
I would like to point out that WHEN [Calendar].[Calendar Date].[Year] = '2005' AND [Calendar].[Calendar Date].[Month] = 1
probably isn't what you want. If you want to know the Year of the current date, you might like to try something like the following
WHEN ancestor([Calendar].[Calendar Date],[Calendar].[Calendar Date].[Year]).member_caption
= '2005' AND ancestor([Calendar].[Calendar Date],[Calendar].[Calendar
Date].[Month]).member_caption = 'January'
Hope that helps,
Richard

Read "Do-Sum" Value

I want to read field-catlog "do-sum" value i.e final grandtotal which diaplay in ALV output . How to read "do-sum" value ?

Hi,
You have to use "get_subtotals" method to get the Totals and subtotals in your ALV THis set of code you will be writing in your "TOP-OF-PAGE" event, Please refer the sample code to do that.
DATA: lo_grid TYPE REF TO cl_gui_alv_grid.
* get the global reference
CALL FUNCTION 'GET_GLOBALS_FROM_SLVC_FULLSCR'
    IMPORTING
      e_grid = lo_grid.
* get the subtotal
DATA: it_00 TYPE REF TO data,
it_01 TYPE REF TO data,
it_02 TYPE REF TO data,it_03 TYPE REF TO data,it_04 TYPE REF TO data.
CALL METHOD lo_grid->get_subtotals
    IMPORTING
      ep_collect00 = it_00
      ep_collect01 = it_01
      ep_collect02 = it_02
      ep_collect03 = it_03
      ep_collect04 = it_04.
FIELD-SYMBOLS: <ft_tab> TYPE ANY TABLE,<ft_tab1> TYPE ANY, <ft_tab2> TYPE ANY TABLE,<ft_tab3> TYPE ANY,
<ft_tab4> TYPE ANY TABLE.
ASSIGN it_00->*    TO <ft_tab>.
ASSIGN it_01->*    TO <ft_tab1>.
ASSIGN it_02->*    TO <ft_tab2>.
ASSIGN it_03->*    TO <ft_tab3>.
ASSIGN it_04->*    TO <ft_tab4>.
Now Your Final Total is in Fieldcatalog <FT_TAB>, and your respective subtotals are in other fieldcatalog.
Thanks
Saurabh
Edited by: Saurabh Siwach on Feb 8, 2010 8:35 AM
Edited by: Saurabh Siwach on Feb 8, 2010 8:39 AM

Break formatting and number format of SUM value

Where can I put number formatting functio, which will format breaking sum values...like:
to_char(#SUM_VALUE#,'fmt 999G999G999G990')....maybe this thread is somewhere posted, but I cannot find it.
Next, how to exclude showing reports total at the end of report?
Is there any working demo to see some hints...
THX!
Message was edited by:
Funky

Hello Damir,
"Maybe a hint for feature releases to have an option to disable total report sum on column …"There is a relative simple workaround, which will allow you exclude or highlight (style) this total report sum line, as you can see in here - http://htmldb.oracle.com/pls/otn/f?p=22710:10
In the Report Attribute page, I'm using the Break Formatting section first field "Display this text when printing report sums" - http://i10.tinypic.com/4ie8ck5.png .
If you want to exclude this line, you should use the same color as your cell background. For the example page, the code is looking like this:
<div id="FinalSum"><script type="text/javascript">var lastRow=html_CascadeUpTill('FinalSum','TR'); lastRow.style.color='#f7f7e7';</script></div>If, on the other hand, you want to highlight the line, or use any other CSS attribute style available, the code would look something like this:
<div id="FinalSum1"><script type="text/javascript">var lastRow=html_CascadeUpTill('FinalSum1','TR'); lastRow.style.backgroundColor='#CCCC99';</script>Total Report Sum</div> Regards,
Arie.

Calculate the max and sum value by weird way !!!

this code calculate the sum value without sum clause :-
DECLARE
x NUMBER;
y NUMBER := 0;
CURSOR cur_sal
IS
SELECT sal
FROM emp;
BEGIN
OPEN cur_sal;
LOOP
FETCH cur_sal
INTO x;
EXIT WHEN cur_sal%NOTFOUND;
y := y + x;
END LOOP;
CLOSE cur_sal;
DBMS_OUTPUT.put_line ('sum sal without using sum function = ' || y);
END;
and this code calculate the max value without max clause :-
DECLARE
x NUMBER;
y NUMBER := 0;
CURSOR cur_sal
IS
SELECT sal
FROM emp;
BEGIN
OPEN cur_sal;
LOOP
FETCH cur_sal
INTO x;
EXIT WHEN cur_sal%NOTFOUND;
IF (x > y)
THEN
y := x;
END IF;
END LOOP;
CLOSE cur_sal;
DBMS_OUTPUT.put_line ('max sal without using max function = ' || y);
END;
the Question is : how suppose that's happen ?
what's the secret in those codes ?
it's just new idea for me , but I don't understand !
is there any name for this way in ( Oracle® Database PL/SQL User's Guide and Reference ) ?
if not , so please anyone give me any help to understand those codes !

Secret? What secret? It slowly and inefficiently fetch each row one at a time and consecutively add them together you get a sum. What the first, horribly inefficient, cursor loop is doing is roughly the equivalent of determining how many rivets are in a bucket by dumping them on the floor and picking them up one at a time rather than weighing one, weighing all, and then doing a simple division. The second loop does the same, horribly inefficient thing, by picking up each rivet and seeing if it is bigger than the previous one.
Cursor loops are obsolete and this code is about as inefficient as you can get without using an Excel macro.

How can I sum values over a date range?

I want to sum values (revenue) for a range of dates (i.e. for each month). I'm trying to use the MONTH function nested in SUMIF and keep getting an error message. I want it to be SUMIF( MONTH (column of dates), cell with month, column with revenue) and it tells me "all ranges must be the same size". It works if I create a separate column where I extract the month from each date and then compare that to a column with the number of the month, but that is really annoying.

You can use the sumif() function to compute the sum of revenues after the "End Date", the sum of revenues before some date and you can use the sum function to summ all revenue. Then you can combine these into a single formula to get the answer by:
sum between two dates is:
sum("of all revenue") - sumif("before date") - sumif("after date"
Cell B4 of the smaller table =SUM(Revenue)-SUMIF(Date, "<"&B1,Revenue)-SUMIF(Date, ">"&B2,Revenue)
I highlighed cells to show before and after dates to highlight the sumif("after") and sumif("before")

How to sum value from a day to next day

First, I provided the CREATE and INSERT sql statements for testing:
CREATE TABLE TBL_CALC
    "YEAR" VARCHAR2(4),
    "MONTH" VARCHAR2(2),
    "DAY"   VARCHAR2(2),
    "HOUR" VARCHAR2(2),
    "VALUE" VARCHAR2(20)
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','1','20','32')
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','1','21','33');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','1','22','53');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','1','23','28');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','2','0','37');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','2','1','35');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','2','2','39');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','2','3','57');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','2','13','42');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','2','14','12');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','31','21','32');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','31','22','52');
insert into tbl_calc (year,month,day,hour,value) values ('2000','1','31','23','29');
insert into tbl_calc (year,month,day,hour,value) values ('2000','2','1','0','27');
insert into tbl_calc (year,month,day,hour,value) values ('2000','2','1','1','45');
insert into tbl_calc (year,month,day,hour,value) values ('2000','2','1','2','69');
insert into tbl_calc (year,month,day,hour,value) values ('2000','2','1','3','68');I have this sql to get the period of time of a day:
         select year,month,day,hr,val,
                 sum(case when hr between 20 and 23 or hr between 0 and 3 then val else 0 end) over (partition by year,month,day) as new_value
                      from(
                             select year,month,day,to_number(hour) hr,to_number(value) val from tbl_calc
         order by year,month,day,hrHere's the output:
YEAR MONTH DAY HR VAL NEW_VALUE
2000 1     1   20 32       146
2000 1     1   21 33       146
2000 1     1   22 53       146
2000 1     1   23 28       146
2000 1     2    0 37       168
2000 1     2    1 35       168
2000 1     2    2 39       168
2000 1     2    3 57       168
2000 1     2   13 42       168
2000 1     2   14 12       168
2000 1     31 21 32       113
2000 1     31 22 52       113
2000 1     31 23 29       113
2000 2     1    0 27       209
2000 2     1    1 45       209
2000 2     1    2 69       209
2000 2     1    3 68       209
17 rows selected The above sql get the sum of time period which in 20-23 and 0-3 in every day.
The values are recorded by each hour. (sometimes there are missed hours)
My problem is how to get the sum of time period over each day?
For example, I need to get the sum of value of a day night time from 20-23 and the next day morning 0-3, to become a sum value in this period.
Let say, the sum from 1st Jan night time to 2nd Jan morning should be combined into 1 value (146+168)=314.
In addition, there is a tricky point is that from every month last day (31th Jan night time to 1st Feb morning time) are also need to be combined into 1 value.
And it appears the new problem is that how about if the last day is in a non-regular day (not 30,31,28,29), just say maybe March consist only 15 period records only and the last day is 19th March.
The expected output:
YEAR MONTH DAY HR VAL NEW_VALUE
2000 1     1   20 32       314
2000 1     1   21 33       314
2000 1     1   22 53       314
2000 1     1   23 28       314
2000 1     2    0 37       314
2000 1     2    1 35       314
2000 1     2    2 39       314
2000 1     2    3 57       314
2000 1     2   13 42       314
2000 1     2   14 12       314
2000 1     31 21 32       322
2000 1     31 22 52       322
2000 1     31 23 29       322
2000 2     1    0 27       322
2000 2     1    1 45       322
2000 2     1    2 69       322
2000 2     1    3 68       322
17 rows selected I think its quite complicated case, please ask me for more if you need to clarify for something.
Thanks so much!! :)
Edited by: 920575 on 2012/7/31 上午 1:09

Are you looking for something like this (partially tested)...
SQL> ed
Wrote file afiedt.buf
1 select year
2        ,month
3        ,day
4        ,hr
5        ,val
6        ,sum(case when hr between 20 and 23 or hr between 0 and 3 then val else 0 end) over (partition by year,month,day) as new_value
7        ,sum(case when hr between 20 and 23 or hr between 0 and 3 then val else 0 end) over (partition by trunc(dt-(4/24),'DD')) as day_value
8 from(
9       select year,month,day,to_number(hour) hr,to_number(value) val
10             ,to_date(year||lpad(month,2,'0')||lpad(day,2,'0')||lpad(hour,2,'0')||'0000','YYYYMMDDHH24MISS') as dt
11       from tbl_calc
12      )
13* order by year,month,day,hr
SQL> /
YEAR MO DA         HR        VAL NEW_VALUE DAY_VALUE
2000 1 1          20         32        146        314
2000 1 1          21         33        146        314
2000 1 1          22         53        146        314
2000 1 1          23         28        146        314
2000 1 2           0         37        168        314
2000 1 2           1         35        168        314
2000 1 2           2         39        168        314
2000 1 2           3         57        168        314
2000 1 2          13         42        168          0
2000 1 2          14         12        168          0
2000 1 31         21         32        113        322
2000 1 31         22         52        113        322
2000 1 31         23         29        113        322
2000 2 1           0         27        209        322
2000 2 1           1         45        209        322
2000 2 1           2         69        209        322
2000 2 1           3         68        209        322
17 rows selected.This turns the values into a proper DATE datatype and then offsets that back by 4 hours to bring the 0-3 hours into the previous day, and then partitions the sum based on that day.

Last Value Exception Aggregation

Hi.
I have a KF with excpetion aggregation with Last Value based on CALWEEK. I have to calculate the 'for the month' and the YTD value for this in a query. When I execute the query, the system fetches the last value for the month for 'for the month'. However, it gets the same last value for the YTD as well. How do I change this to get the sum of last value of all the months for the YTD Value. For example, if I am looking in sep 2008, the for the month value should be the last week of Sep 2008 and the YTD should be the sum of all the last values for months Jan -08 to Sep-08.
Cheers
Anand

Hello Vikram
One possible solution I can think of is to use an extra ODS in order to hold your data in an aggregated form that suits you. Unfortunately the exceptional aggregation behaviour does not work outside reporting thus in the transfer rules from your data provider to this new ODS you should add a formula checking whether the new record to be transfered to the ODS has a date newer than the ones already transfered in the ODS. Else it should skip the current record and proceed to the next one.
Another possible solution would be to use the APD and transfer the results of the query you are already playing in an ODS (this way you take advantage of the exceptional aggregation behavior) and by transfering your records to a second ODS you can achieve the aggregation you need.
Assign points if any of the above helped

Validations for sum value

Hey,
I have a problem with validations:
I have in a product repository a multi-valued qualified lookup into an 'ingredients' table.
In the 'ingredients' table I have several ingredients and as qualifier the percentage of its share in the product.
So now I want to make a validation in the 'product' main table checking if the sum of all ingredients' percentages is 100%.
I haven't found a sum formula though, yet.
Can anyone help, please?
Thanks,
Chris

Hi Christiane,
I do not know a solution for that using SP03, but SP04 will have aggregation functions. You can apply these functions e.g. to the non-qualifier fields in order to sum them up. The bad news is, that you will have to wait until August....
Kind regards
Andreas

How to calculate the total sum value of a particular field that repeats

Hi All,
I have the following Req...File----Idoc Scenario
In the Inbound xml file i will get the Sales Order details with suppose 10 line items( 10 Orders)
Each line item represents one one SO. So totally i wil have 10 Sales Orders in this file.
<?xml version="1.0" encoding="UTF-8"?>
<ns0:MT_Sales_Order xmlns:ns0="http://sap/Sales_Order">
   <Header>
      <COMP_CODE></COMP_CODE>
      <DOC_TYPE></DOC_TYPE>
       <SUPPL_VEND></SUPPL_VEND>
   </Header>
   <Item>
      <ITEM></ITEM>
      <MATERIAL></MATERIAL>
      <PLANT></PLANT>
      <QUANTITY></QUANTITY>
      <Amount></Amount>
</Item>
In the above structure Item Segment will repeats as many no. of Sales Orders comes in a file.
In a file if there are 10 Orders means the Item segment wil repeats 10 times.
I have the Amount field in the Item Segment, each and every time that needs to be added to next Amount value that presents in the next Line Item.
Finally i will have the Another separate field caled Grand Total, and i have to get the total summation of the 10 values of the Amount field at last.
Can we achieve this using UDF or is there any way to do this
REgards

Hi,
Do like this, actually in your case sum is taking place before the condition check for discount type
do a little change in mapping
DiscntType--removeContext--
                                          EqulsS-------IfWithoutElse----Amount---removecontext--then---SUM-
Connstatnt(Value)
--->GrandAmount
Krishna, Check the same question asked by Rajesh in thread Calculate totals of segments that occur multiple times
Thanks!

How to carry over values differently (aggregated or not), or is it possible

Dear all:
In our Account Dimension, we have Beginning Inventory Qty, Ending Inventory Qty, Purchased Amt, Sold Amt, etc.
When we run reports with calculated time members, such as 2008.Total, 2008.H2, 2008.Q4, etc, most of our Account displays correctly with proper aggregation (like Purchased Qty, Sold Qty is summation of all child months'). But in particular for Beginning Inventory Qty and Beginning Inventory Amount is where I am having trouble.
For instance, for Ending Inventory Qty, we assigned ACCTYPE AST, so the report ignores aggregation and carries the last month's value to the calculated time member. However, for Beginning Inventory Qty and Amount, I wish to assign the values of the first month. If it is 2008.H2, then 2008.H2 Beginning Inventory Qty should be the same as in 2008.07. Another example would be 2008.Q4's Beginning Inventory Amount being the same as 2008.10's.
Is this doable in BPC?
Thank you!
Brian

Hi Brian,
It may be possible to design the opening balances in the way that you propose here. To get the correct results in your approach, you will be relying in large part on the cube and Analysis Services to figure out -- for certain intersections of accounts & time levels -- that it should look to the opening period rather than the standard time dim aggregation. I assume that would require some fancy MDX programming of a custom measure.
I would recommend a different approach -- in particular if you plan to use the consolidation engine business rules -- that you store the opening balances in each of the 12 months. This makes it easy for the time-dim aggregation: for Q3, it looks at Sep; for H2, it looks at Dec, and so on -- and this is exactly what the cube does naturally. Sure, it may feel like some data duplication for no valid business purposes, but in some cases we master the technology, and in other cases the technology masters us. This is one of the latter cases.
So then the question is, which opening balances? This can be tricky. It comes down to the question, are opening balances for each month = the closing balance from prior year December (or whatever your final period of the fiscal year is), or are they = the closing balance of the prior month?
BPC can handle it either way, and for a legal consol I would always recommend the first approach (prior year December). For a planning app it depends on the business requirements for B/S planning. The copy opening balances business rule is designed for the former. The account transformation rules can be used to handle the latter (and also support redirection across the flow-type dimension), with a source period = -1.
It also depends in part on your application's data storage type (YTD or periodic), particularly if you expect the system to create a cash flow statement for you, where the P&L data comes into play too.
As Petar mentioned, you should also carefully consider using a flow / movement / accdetail dimension. It's not always helpful, but if there are lots of accounts that you're tracking movements on, it can be worthwhile -- even if it means a lot of mapping of the ERP COA to the different movements. It's not intuitive for people who haven't spent a lot of time with their heads in the OLAP cube, and are used to a flat COA concept. It it can be very powerful, since the consolidation engine takes enormous advantage of this. It makes the setup of the "copy opening balances" logic rules very, very simple. And it could either complicate, or simplify, some report layouts.
Usually it's one of those dimensions (like datasrc) that customers only really start to appreciate after 6 months on the project.
Hope that helps.... and if you do succeed in your initial proposed approach, I'd be very curious to hear how you did it.
Regards,
Tim

Sum value in aggregator

Similar Messages

Maybe you are looking for