Incorrect Time based disaggregation

[Former Member]

Hi All,

I need some consultation on Disaggregation of a value.

We have our historical data in weeks, where as the planning bucket is in months.

I want the monthly forecast data to be disaggregated in weeks based upon the historical data.

In the planning area, for the forecast key figure we set the Time based disaggregation as ‘K’ –Based on another key figure and selected APODPDANT as the time based disaggregation key figure.

For generating the proportional factors, I have selected ‘History’ key figure as basis for proportion calculation. I have selected the ‘detailed proportions’ and the periodicity as ‘week’.

I’ll explain it in an example.

Say this is the data in my history key figure for May’07

w18.2007 15000

w19.2007 2100

w20.2007 11000

w21.2007 6500

w22.2007 20000

Total 54600

Based upon this, I’m disaggregating the monthly forecast quantity to weeks.

The monthly forecast for May’08 is 54600. As I have generated the proportional factors for time based disaggregation, I’m expecting the same proportions as of history but the results from the system are like this :

w18.2008 7045

w19.2008 12329

w20.2008 12329

w21.2008 12329

w22.2008 10568

Total 54600

I have deleted the proportion factors as well as forecast quantity before generating new proportions.

Structural disaggregation is working perfectly.

Please guide me where am I going wrong. Your help will be greatly appreciated.

Regards,

Raj

[Former Member · Accepted solution]

Hi Raj,

The data is first disaggregated by time and then by proportions.

Time based disaggregation also depends on the number of days in that week. If a week has 7 days it would have more qty. than week with 5 days.

This sap help [example|Time based disaggregation also depends on the number of days in that week. If a week has 7 days it would have more qty. than week with 5 days. ] explains you in case there are no proportions. The system does this first and then apply proportions.

For number of days, please check the factory calendar/ time stream id you are using.

Try with these and let us know if that's working.

[Former Member · Accepted solution]

Hi Visu,

Thanks for your reply.

The storage bucket profile for the planning area has days, weeks and months.

The link you have given is not opening.

If i want to replicate the same weekly proportions of the historical data to the forecast data, what are the steps do i need to maintain.

plz guide me.

Thankyou

Regards,

Raj

Edited by: RAJ S on Mar 19, 2008 11:19 AM

Edited by: RAJ S on Mar 19, 2008 3:39 PM

[Former Member]

.

[Former Member]

.

[Former Member]

Hello Rajesh,

Please post the solution if you have figured out the solution for this time based dis aggregation. I am doing exactly the same thing except that i trying to achieve this from Quarters to Months.

Your solution would be really valuable to me.Thanks in advance.

Sai