Selecting Order of Partition Columns in Pipeline Builder

In pipeline builder I am trying to perform a window function and when selecting the partition columns I need them to be in a very specific order for spark to optimize its plan correctly, however I am noticing that no matter the order in which I click the columns, they always appear in the same order in the selection box as shown below.

Is there a way to ensure the partitioning is done according to a specific order?

Hi @lorenzojb. I did some digging. It looks like the ordering you input is actually persisted. However, we incorrectly sort the columns when displaying them.

You can see the saved ordering in the minified view here:

Thank you for catching this bug! We can work to roll out a fix. Let me know if you’re seeing other issues with the partitioning.