Skip to Content

Period Split Profile question

Hello gurus:

I have a period split profile on my mc90 release of DP forecast to SNP.

However, I'm not sure it is serving a function. DP is in monthly

buckets and SNP is in weekly buckets. I have a Time Bucket Profile

that I'm using on the variant in mc90 for the entire horizon in weekly

buckets. Should a Period Split Profile be used on these variants? If

so, how should it be designed? It seems like the forecast is released

to SNP no matter what Period Split Profile I use.

Thanks.

Andrea

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

2 Answers

  • Jun 19, 2012 at 03:45 PM

    Andrea,

    If there is no period split profile, your forecast in monthly bucket is proportionately distributed into weekly buckets and released to SNP. For example, your forecast for a month could be 100 and considering 4 weeks for the month, the forecast release to SNP will be 25 for each week i.e. uniform distribution. However, based on your configuration, the calculation also considers the working days and 'Period Split' parameter value (/SAPAPO/MAT1 - location product master - SNP2 tab) also influence the time based disaggregation.

    Period split profile helps you to alter the uniform distribution flexibly to a custom distribution based on proportions or quantities. For instance, you may want to give more emphasis on first 2 weeks of the month i.e. weekly forecast release proportions are 45%, 35%, 15%, and 5%.


    To use period split profile, first you need to define a distribution function using t.code /SAPAPO/DFCT

    Then, define the period split profile using t.code /SAPAPO/SDP_SPLIT

    Thanks,
    Rajesh

    Add comment
    10|10000 characters needed characters exceeded

  • avatar image
    Former Member
    Jun 19, 2012 at 07:35 PM

    Andrea

    What is your requirement ? Are you trying to avoid forecast values from getting released in the past ? If so you have to maintain the value of period split as 1 in your period split profile.

    Thanks

    Aparna

    Add comment
    10|10000 characters needed characters exceeded