start routine

[Former Member]

Hi all;

I have this headcount cube and the headcount has value 2 instead of 1 while running the headcount report. the same person is assigned to the same costcenter twice, we can do this restriction on query level..but we have several reports on headcount so we would like to control this in cube level. we have start date, end date, 0costcenter..in the cube..; Anyone have good thoughts on this how to starts?

[Ashwin · Accepted solution]

Hi!

I think it depends on how you are loading data and the structure of your cube. as your title states you can use start routine to remove multiple assignements. but i think normally head count will normally be a constant "1"with overwrite option. so it must be one even when it is duplicate. and ya what is the content of cube and what is the key field. and how is your infosource structured.

with regards

ashwin

[Former Member · Accepted solution]

I am not experts in HR module; but ZHD_FA (Headcount Actuals) is my keyfigure, and I have 0empgroup, and 0costcenter are my char in the report. the same employee has current start and end date and future start and end date has also been assigned for the same employee under same cost center..; somehow..I am getting headcount 2 for that specific employee. I am thinking we can resolved this by restricting the start period in a start routine..I already tested doing restriction in report level and it works but we want to do this on cube level..; so pls give me your thoughts...how should I write a logic or anyother ways to handle this situations. thanks for the quick response.

[Former Member · Accepted solution]

Do you have exception aggregation turned on for the key figure? Headcount - exception aggregation is last value and reference characteristic is 0CALMONTH. Without this you will get total of all.

Try taking a drilldown in the report by calmonth. Check data for a particular month.

[Former Member · Accepted solution]

Aneesh;

This sounds good..where do I do this exception aggregation?

[Former Member · Accepted solution]

It should be set in the key figure properties on the second tab.

Other option is to do it in the query itself - create a clacualted key figure on the InfoProvider equal to the headcount key figure. In the properties screen, at the bottom is an enhance button - press that. Select last value in the first drop down and calmonth in the second.

Set it in the key figure though too - that will ensure that no one makes an accidental mistake in the future and default exception aggregation is turn on.

Have a look at the standard key figure for head count - i think it is 0HEAD_COUNT.

[Ashwin · Accepted solution]

I think your reports are logically correct acording to your cube designing and naturally there must be time restriction in all your reports to get the actual head count.

other wise remove all the future associations from cube while loading through start routine

example code can be

DELETE DATA_PACKAGE[] WHERE startdate >currentdate.

with regards

ashwin

[Former Member · Accepted solution]

I think it should work fine..make sense..but why do I need to defined reference char '0calmonth'? I am not using 0HEAD_COUNT, instead I have customer defined keyfigure 'ZHD_FA'. Also I am using this keyfigures in other cubes and reports also..do you think it would make any negatives affects! thanks

[Former Member · Accepted solution]

For the date range selected for the query, exception aggregation will take the lastest month for which data is available. Incase no selection is given, then the date range is everything - effectively returning the lastest month for which data has been posted.

Headcount cannot be aggregated. It should always be the last value.

If a user selects year 2004 then headcount should be reported as of 31/12/2004, i.e. last day of last month. If Q1 2004, then it should be as of 31/3/2004.

In the update rules, alst day of a month is used to determine the master data values for each employee.

[Former Member · Accepted solution]

Your custom defined key figure should work in a way similar to the standard.

[Former Member · Accepted solution]

Last question, I am using this keyfigure in other cubes like 'headcount & forcast, HR actual & forcast etc..; if I turn on the exception aggr. do you think I should be fine.! this is production issues..so I want to make sure..!

[Former Member · Accepted solution]

Handling headcount logic should be the same. Try it out in dev though - I do nto know how it is being used in other cubes.

For this particular one, turn on exception aggregation in the query only as described above.

[Former Member]

There is something amiss here. Headcount extractor brings one record per employee per month. In the update rules, cost center is read from the employee master for the last day of the month and updated in the cube.

Check if you have duplicate data update in the InfoCube through two reqeusts.