Hi All

I am coming across unusual behaviour with spreading and holds.
I have attached a little video showing this. 
Basically if i do a spread , with a number which is "held", the spreading total adds the held number to the total i want to spread.So if I want to spread a value of 300 across some cells, with a cell having a value of 10 being held, i get a total of 310.The update action I have chosen is to Replace values and not to add them.
Has anyone else experienced this ?. I am using PAW version 70, and a simple cube from SData.

Many Thanks

------------------------------
Armand Mizan
------------------------------

#PlanningAnalyticswithWatson

Hi Armand,
We'll verify the spreding in your scenario and post the answer.
Did you have different results in the same spread scenario in previous version of PAW?
Your scenario with Hold and Replace spreading is working the same way in Architect.


Best Regards,

------------------------------
Svetlana Pestsova
IBM Planning Analytics Product Manager
------------------------------

Hi Armand,
This is related to the making of a direction selection, there are two types of proportional spreading in play behind the proportional spread.

C Level : If I make a proportional spread without making a directional selection I get the expected behaviour, i.e. a full replacement of the C level value, with the hold respected.  (Assuming a C level selected as target)

N Level : If I make a direction selection then spread has totally different behaviour.  Firstly the spread is only working on what is visible on the screen and has to have N levels visible, to some extent the cell I start the spread from is meaningless except as a reference point for the direction.  So if I spread from the full year total but only have Apr-Dec visible then only Apr to Dec will be a targets.  In addition any cells that are held are excluded.

This probably just about makes sense (I think!) but it is confusing in your case becuase I think have set yourself up to do a C level proportional spread but are then executing an N level proportional spread.

HTH

------------------------------
Steven Rowe
Technical Director
www.infocat.co.uk
------------------------------