One way to do this would be to run two profiling steps in single measure,
register results using out.dataframe.name, and then do custom spark-sql DSL
rule to compare results from previous steps. It's not possible to create
such a measure using UI wizards, but it's possible to express that in
measure JSON.

On Mon, Apr 22, 2019, 19:55 secfree teng <[email protected]> wrote:

> Hi all,
>
> I want to compare the count of two tables based on some conditions. But
>
>    - “Accuracy” does not support calculating the count
>    - “Profiling” does not support comparison
>
> Any recommendations?
>
> Thank you.
>
>

Reply via email to