Best Of

😁 And two minutes later I found the answer myself.....

In the ETL, at the Input Data Set going to Settings for Integer and change to en-US....Does the trick...

Hi @eriena, is your use case related to getting a list of products that don't have an assigned shop and then you (or other users) can go back to that dataset 2 to correct/update it? If so, you can have a full outer join (i.e. giving you all the records from both tables and joining matching records where possible) setup and then add a Filter tile to only include any rows where you have a product but no shop. So it would go: Join tile -> Filter tile -> Output tile.

I'd recommend using a formula tile to input your CASE statements which are inside your SUM aggregation as two separate values. Then do a group by to get the SUM then another formula tile to do the outer portion of the beast mode (checking if your denominator is 0)

That seemed to work. Thanks so much guys!

I figured out what I was doing wrong - thanks for your help!

You're missing a single quote and a few back ticks (these designate column names).

CASE
WHEN `Journey-Pathway Content Progress` = 'Completed' THEN 'Completed'
ELSE `Content Progress`
END

Doing this as a beast mode will create a second column in your dataset as that's how beast modes function. You can rename your column's Display Name value clicking on the column name and changing Label value. Then you can use this new column / beast mode field in your report instead of the old column so it's only displayed once.

Hi @micheleb

You can use a case statement to conditionally set column B to completed if column A is completed

CASE WHEN `Progress Column A` = 'Completed' THEN `Progress Column A` ELSE `Progress Column B` END

If you don't want an extra column in your dataset from this beast mode you can use the same formula in a Formula tile in the new Magic ETL and set the column name to be the same as the existing one and it'll overwrite the value for you.

You wouldn’t necessarily need a full outer join as the left join would work just fine. After your join just add a filter tile to say where shop is null and then feed that into another output dataset