Best Practice: PARENT() vs. LOOKUP ?

Zak_G · 2023-09-04T18:48:45+00:00

There was an error rendering this rich post.

I am looking for assistance to understand best practice/performance implications between 2 different approaches of getting a list item parent in a formula. Imagine I have 3 modules: X, Z, and System Module. The System Module contains Parent line item with formula PARENT(PARENT(ITEM('List Items'))). I want to return this same parent item in a new module and line item and I can use either method 1 or 2:

Method 1: PARENT(PARENT('Module Z'.'Child Line Item'))

Method 2: 'System Module'.'Parent'[LOOKUP: 'Module X'.'Child Line Item']

Is it better to use Method 2 as it follows the idea of "calculate once, refer to many times" , or is it better to use Method 1? Curious how the engine evaluates the parent() function and nested parent() functions.

Accepted answers

vishal3120

I guess you yourself gave the answer to your query at last. Still if I elaborate a little, in method 1: everytime that line item would be used it would have to first look into the list then find the parent of it and then engine will run again to calculate the parent of the parent.

But in method 2: You have a system module which has already done all the calculations and this time engine has to just refer the value(Parent) from system module and place it accordingly. Therefore, load on engine in method 1 »»» method 2.

PS: Anaplan engine is way too fast that you won't feel any difference after using either of Method 1 or Method 2 ;)

rob_marshall

@vishal3120 @Zak_G

You should never use parent(parent(item)) even in the SYS module. In the SYS module, break it up so it is this:

item: item(list)
parent: parent(item)
grandparent: parent(parent)

And yes, there is a performance difference, it is better to do the calculation once (in the SYS Module) and reference it in other modules than doing the same calculation in multiple modules.

All comments

vishal3120

PS: Anaplan engine is way too fast that you won't feel any difference after using either of Method 1 or Method 2 ;)

Zak_G

vishal3120 - thank you for your comment and indeed, in retrospect, I think I was probably overthinking things … I was concerned that using Method 2 would leave me open to daisy chaining , but as long as I remember to reference the system module instead of one of the line items looking up this value, I should be in the clear. Thank you!

rob_marshall

@vishal3120 @Zak_G

You should never use parent(parent(item)) even in the SYS module. In the SYS module, break it up so it is this:

item: item(list)
parent: parent(item)
grandparent: parent(parent)

And yes, there is a performance difference, it is better to do the calculation once (in the SYS Module) and reference it in other modules than doing the same calculation in multiple modules.

Quick Links

Can you help?

Unanswered questions

ADO File Upload - Semicolon Delimiter
Hi, Based on the regional settings, my default delimiter/separator for CSV files is a Semicolon. In the classic way of uploading a csv file, I was able to select the delimiter. When using ADO, I don't have the option to select a delimiter. Uploading a Semicolon delimited file returns an error, or an upload with only one…
Request for Workflow Design: End-User Report Finalization & Stakeholder Notification with Version Co
Hi everyone, I'm looking to design a workflow and would appreciate your input or guidance on how to set this up effectively in Anaplan. So, the goal is: Once end users complete their data input, they should be able to trigger a process that: Notifies stakeholders via email with a link to a report. The report includes: A…
OAUTH 2.0 vs Google Sheets Extension
Hi, My org is looking to explore leveraging Google Apps script to extract data out of Anaplan and update already built templates in Google Sheets. We use the Anaplan Extension currently but find it crashes often with the amount of tabs, data, etc. Has anyone used both Anaplan OAUTH 2.0 API and Goolge Apps Script to write…