Trying to use a non-additive metric to filter on a new metric

Hi!

I had a strange use case with metrics and Lightdash, and wanted to see if others had run into the same problem:

I have a summarised data source, and from it I’ve create a metric called metric_one_percentage (which is just sum(COL_A)/ sum(COL_B)), and the dimension is ENTITY_A.

So an example output for a monthly time grain would be:

I want to create a second metric which shows the percentage of entities, of all the entities, where the metric_one_percentage exceeds a certain threshold, x.

So in essence the formula would be:
Number of entities with a metric_one_percentage >= x / Number of entities

However I would also like this metric to have the flexibility with the time series that metric_one_percentage would have, that is, I don’t want to pre-define the time period for metric_one_percentage to be able to calculate the second metric. I would like the user to define the time grain when working in Lightdash.

Is this possible?