TM1 Forum

Hi All,

I'm trying to import a csv file into a cube with 6 dimensions, whereas the source CSV file only has data for 4 dimensions.
The business logic is load the data to all the members in the remaining two dimensions.

For this I'm using a clustered while loop, the first while loop to go over the leaf level elements of fifth dimension and the second while to go over the leaf level elements of sixth dimension. Then I use "CELLPUTN" to input data to measure.

I find this hinders the performance very much. For even a small amount of data, the TI process takes a minute to complete.

Is there any other workaround for this? If yes, please do let me know. Thank you very much.


-Vino

Please explain the logic to put the same value on every cell on a quadrant, being the multiplication of x * y cells. Not: x + y cells but x * y cells.
x leaf level elements from the 5th dimension, y leaf level elements from the 6th dimension.
What is the reasoning ?
You could also load the value in 1 spot and have a rule retrieve that value to show it in the other cells ?

Whoever made the decision to load to all the members of the non-represented dimensions does not understand TM1 and the consequences of that. Standard practice is to create a member in each of these dimensions called "No account", or "No Cost Center", etc., and load the data there.

Wim Gielis wrote: ↑Fri Dec 14, 2018 4:20 pm Please explain the logic to put the same value on every cell on a quadrant, being the multiplication of x * y cells. Not: x + y cells but x * y cells.
x leaf level elements from the 5th dimension, y leaf level elements from the 6th dimension.
What is the reasoning ?
You could also load the value in 1 spot and have a rule retrieve that value to show it in the other cells ?

@Wim Gielis ... I like the idea of loading the value in one dummy combination and load the same for other combinations. But is this possible using a rule file?

I know if I go for TI, again I will have to use WHILE loop.

Hopefully vino will put us out of our misery and tell us the logic behind this!

I saw something like this once where within TM1 there was a need to reconcile 2 different TM1 models because data was transferred between the 2 models and for whatever reason the organisations chart of accounts looked different in the 2 models (both models had say 6 account key dimension but only 4 of them matched each other and only one of them matched the actual general ledger)

In order to reconcile the 2 models the % of the 6 part to the 4 part was calculated for one model and the values were then apportioned back to the 6 part in the other model! Yes the world out there can be that crazy!

I think in the end the values were exported to excel and the reconciliation was done there instead of TM1.

I hasten to add that I had nothing to do with this model, other than looking at it incredulously!

Hi,

I did not post the above comment into this post yesterday! I am pretty sure I posted this back in December!

Is there a TM1forum forum!

cheers, Mark