TM1 Forum

Hi ,

Can any one help me on how to manage cumilation in TM1?

Sruthy wrote:Can any one help me on how to manage cumilation in TM1?

Sure.

You start by reading the Request For Assistance Guidelines.

Then, having referred to the manuals to try to find the answer for yourself (and indicating exactly where you tried to find the answers, as the guidelines state) you post a detailed question explaining what it is that you're trying to do, what you're expecting to happen, and what is actually happening when you try it.

Then, you get an answer to the actual question... or more likely you won't need it, since you'll have already found it for yourself.

And most importantly, if you do still need to ask the question, you should do so using the correct word(s) (which, hazarding a guess, is actually accumulation in this case), for if you were to do a Google search on the word "cumilation" I think you will find that the word does not mean what you think it means. I did so, and now wish that I had not.

Since you're new here, though, I'll give you the key word that you're probably looking for, assuming that you look in the 10.1 or higher Reference Guide... the magic word is "CellIncrementN".

Thanku Alan,

Let me explain u th scenario.
Am having a cube which contains dimensions such as producs months and EMI Measures.
The purpose of the cube is to derive EMI and Accumilated EMI.
Eg:EMIs for the Month Jan Feb Mar are 100 100 100 i need accuilated emi for this.

You just need to look at and understand dimension element types.
Then you will be able to create YTD consolidations in the period dimension.

Performance modeler offers a dimensional function called 'Cumulate' when you create a Calculation dimension. Try that.

Muddler creates TM1 rules and feeders when you use dimensional functions like CUMULATE. Cumulation using dimensional consolidations is much much faster and doesn't require feeding.

I would be also very cautious with any of the built-in PM calculations and feeders...
Try to search the forum for 'YTD'

As generated code goes, I don't think the cumulation rule is that bad: However it is inefficent, as explained in my earlier post.
If you want to see really awful generated code, look at the scorecarding functions, then offer up a prayer that your client doesn't ask you how the calculation is done