Opinions and Best Practices for Allocation Model
Posted: Mon Nov 29, 2021 9:01 pm
Good Afternoon Everyone,
I am looking to solicit some advice on creating a TI process to replace an existing allocation model. Currently, Allocation Rates are generated within Excel by pulling data from GL MAIN (cube) then uploading the computed rates into a separate cube called, "Production Allocations." Once the rates have been loaded, I run a separate TI process to clear Product Line Adjustments (measure) in GL MAIN, followed by another TI process to repopulate Product Line Adjustments in GL MAIN using the newly updated rates multiplied by balances from GL MAIN.
Would it make more sense to replace/remove the Product Allocation cube and store (access) the Allocation Rates in GL MAIN, or would the existing model be sufficient for the Allocation process? Note: Even manually, the full process takes less than a few minutes to run, so this new TI process would move us toward running Chores as opposed to manual intervention.
Provided I keep the existing model in-tact and replace the Excel-model with a TI process, I'd love to hear some different strategies on ways to replace over 800 DBR calculations with a reasonable number of calculations within TI. There are only three dimensions that could change based on their relative row/column position.
I was leaning toward creating a new attribute in two (maybe three) of the "variable" dimensions and write the formula to fire on only the intersection points where Allocation Flag = Y.
Thank you in advance,
Michael
I am looking to solicit some advice on creating a TI process to replace an existing allocation model. Currently, Allocation Rates are generated within Excel by pulling data from GL MAIN (cube) then uploading the computed rates into a separate cube called, "Production Allocations." Once the rates have been loaded, I run a separate TI process to clear Product Line Adjustments (measure) in GL MAIN, followed by another TI process to repopulate Product Line Adjustments in GL MAIN using the newly updated rates multiplied by balances from GL MAIN.
Would it make more sense to replace/remove the Product Allocation cube and store (access) the Allocation Rates in GL MAIN, or would the existing model be sufficient for the Allocation process? Note: Even manually, the full process takes less than a few minutes to run, so this new TI process would move us toward running Chores as opposed to manual intervention.
Provided I keep the existing model in-tact and replace the Excel-model with a TI process, I'd love to hear some different strategies on ways to replace over 800 DBR calculations with a reasonable number of calculations within TI. There are only three dimensions that could change based on their relative row/column position.
I was leaning toward creating a new attribute in two (maybe three) of the "variable" dimensions and write the formula to fire on only the intersection points where Allocation Flag = Y.
Thank you in advance,
Michael