cancel
Showing results for 
Search instead for 
Did you mean: 
Read only

Data Inconsistency

Former Member
0 Likes
1,684

Hello experts,

I have created a generic data source to pull employee master data (time dependent) from info types PA9100 and PA0014 which have wage types and bonus.

For this I am doing full load into bw.But I am getting errors like " overlapping time intervals of data records 1 with 2(to/from)" for many records.When I checked the data i realized that it is because of the same from date and to date repeating.Like for example:

Record Perno From date To date wage type amount

1 111 03/04/2010 04/05/2010 AA 50

2 222 03/04/2010 04/05/2010 BB 100

Can some one help me over come this issue..

Thanks

Accepted Solutions (0)

Answers (1)

Answers (1)

Former Member
0 Likes

Hve u defined From and To date as the key?

Former Member
0 Likes

To date is defined as key.

Former Member
0 Likes

Thats why you are getting this erorr. Only date can never be a key. It has to e with combination with some other characterstic.

Former Member
0 Likes

No, date to and emp no are defined as keys.Date from is not defined as key.

Former Member
0 Likes

What i meant that date to and emp no combination might not be unique in the load. Try including some other characterstic which will make the 3 combinations unique.

Former Member
0 Likes

Can you please guide me how to do that..

Former Member
0 Likes

While creating Cube/DSO, you need to define which all characterstic will be your key.

Former Member
0 Likes

this is actually master data loading..employee master data..

Former Member
0 Likes

Hi,

For some employees there might be multiple records and the from date and to date might collide. Plz check.

Regards,

Raghavendra.

Former Member
0 Likes

Hi,

Actually the problem is for some employees the from date and to date are same but only the wage type and amount changes.If I do not tick the handle error records under update tab in DTP, the data load has errors overlapping dates and if I check that option then the repeated dates records are deleted.I need all the records bcoz for those records there is a change in the wage type.

Former Member
0 Likes

Hi,

How can an employee have multiple wage type in a same timeframe. Logically this is not possible? Plz check.

Regards,

Raghavendra.

Former Member
0 Likes

Hi Raghavendra,

I have checked, for the same dates the employee has different wage type(sub type) and amount.This data is taken from infotype pa0014.

Wagetype From date to date amount

6791 03/07/2010 03/20/2010 32

6755 03/07/2010 03/20/2010 8

Former Member
0 Likes

Hi,

If this is the case then you cannot load it in 0employee master, because your thing does not satisfy as master data criteria. You can use DSO for your purpose or you should compound wage type to ur master data(which i think is wrong) and load. Plz check with ur HR consultant whether a single employee can have multiple wage type for a given period.

Regards,

Raghavendra.

Former Member
0 Likes

ok.Here I have created a zemployee with only fields from date, to date, wage type, amount.And getting the data from a generic data source to get records from info types pa0014 and pa9100.Would making the data source delta capable solve the issue.

Former Member
0 Likes

Hi,

Ok. Let us say what is the use of your zemployee then? If all the fields are compounded then what are the attributes of zemployee? even if there are attributes then you have to have wagetype, amount and what ever field you have compounded in your transaction data to display the attributes in report. Plz check whether this is what u require?

Regards,

Raghavendra.

Former Member
0 Likes

I have taken wage type, amount as attributes of zemployee.

Former Member
0 Likes

Ok. Again we are back to square one :-). If these r attributes then for a particular time period you can have only one value for the employee.

Former Member
0 Likes

Basically I need a report which should show the transactions whose wage type(1sti/1lti) or bonus changes in a user selected time range and it should also show the previous wage type or bonus.So for that I am taking the data from the info types pa0014 and pa9100 into info object zemployee which have wage type, amount as attributes.

I have the wage types in one of the existing cubes but since it does not hold the history and also does nt show when it changed(from and to dates) I am goin for info object and reporting on master data.

Former Member
0 Likes

I hope you got what I am looking for..if you have any idea, please share..

0 Likes

Hi,

I think, if you put wage type in the compounding of ZEMPOYEE and make the InfoObject Time Dependent, this will solve your problem.

Roy

Former Member
0 Likes

Hi,

I tried that way but still the data is not correct.I figured out that if I can make wagetype, emp, frmdate and to date as primary keys I will be able to get the data correct.But im not sure how to do that.Now I have emp and to date as my keys.In the DTP also im not able to edit the semantic key.Can you please share if you have any idea.

Former Member
0 Likes

Hi,

Why cant you use a DSO for this purpose. Try to implement the logic in that.

Regards,

Raghavendra.

Former Member
0 Likes

I tried and i got the data correct.But I want to report on info object rather than DSO.Because I need to create an infoset for the 0employee and my zemployee and my report is based on the to and from dates.

If I use DSO i cannot use the time dependency factor(from and to dates).I have to create zfrom and zto dates but with this I cannot meet my reporting requirement.

Former Member
0 Likes

Can we make a DSO time dependent?

neelesh_jain3
Contributor
0 Likes

The Key Combination should be EMP ID, To Date and Wage Type.

You can have a full load without any issues if you are loading into an Info Object or DSO.

Of course if you can make your data source delta capable then that would be suggested.

Regards,

Neelesh