Qnum and calculated columns

Hi! I’ve been working with Qnum columns, and I noticed that they are not properly handled by standard operators, for instance:

Comparison:
image

Casting:

The first one will certainly hit the users as they wouldn’t understand that “<25.0” is actually internally stored as 25.000000000000004 and as such greater than 25.0.

What would be handy too, is a set of three additional functions similar to DG.Qnum.getQ() == DG.QNUM_GREATER, QNUM_LESS and QNUM_EXACT (for instance IsQualifierGreater(Qnum), IsQualifierLess(Qnum), and IsQualifierExact(Qnum)).

Hi @gregori.gerebtzoff.novartis.com!
Thank you for providing the information on the problem.
@skalkin I think it has to do with the way QNum mantissa least significant bits are involved in storing QNum. I have created an issue for us to follow up.
@gregori.gerebtzoff.novartis.com we will update you once this one is addressed.