Performance Issue with Master Switch for Dynamic Product Selection in Polaris

Question

Hi All, We are facing a scenario where users must dynamically select specific list items for different planning scenarios. Example: * We have 5 products in total * Scenario A: User plans for 3 products * Scenario B: User plans for 4 products Because this selection is dynamic, we are not using subsets. Instead, we: * Use Boolean flags at the product level * Apply conditional logic in formulas (e.g., if Product Applicable? then calculate else 0) To control performance, we implemented the following approach: Master Switch Pattern We created a SYS Calculate? Boolean master switch to control high‑effort formulas and pause real‑time calculations. The idea is: * Enable calculations after selecting the products only at the final step, This works functionally, but we are encountering a performance issue: * When the master switch is turned ON, the Polaris model takes ~2–3 minutes to apply changes * It seems Polaris recalculates everything once the switch is enabled Questions: * Has anyone faced a similar issue with a master calculation switch in Polaris? * Are there better design patterns to handle dynamic product selection without triggering a full recalculation? * Any recommendations to improve performance or alternative approaches we should consider? Thanks in advance for your help! @Misbah @seb_mcmillan @Marina Ketelslegers Thanks Ashutosh

Ashutosh · Answer

Hi @rob_marshall , Thank you for your inputs.

Let me use numbers instead of Booleans & will check the performance.

Thanks

Ashutosh

rob_marshall · Answer

@Ashutosh

The double lookup could be causing an issue, so if you are getting numbers (total_volume), try to do that with a sum.  And yes, I am not looking for you to make structural changes, that was only when you are loading dimensional data where that calculate? could/should be used.

Ashutosh · Answer

Hi @rob_marshall ,

Thank you for your inputs. I agree that in Polaris, using additional checkboxes or Boolean logic can impact performance.

However, our requirement needs a flexible design where users can include or exclude specific list items, and calculations should run only for the selected items. This selection is scenario‑specific—for example, one scenario may include 10 items while another may include 20.

Additionally, downstream calculations such as allocations and ratio-based logic must adjust dynamically based on the selected items for each scenario. Because of this variability, the design needs to accommodate frequent changes without requiring structural updates, while still maintaining acceptable performance.

One of the Example formula:

IF    Applicable Versions THEN    'AGR0.3 '.total_volume[Lookup: A, Lookup: B] ELSE   0

But As per your suggestion, I will play with numbers instead of Booleans and will note the performance.

Thanks

Ashutosh