Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Is the orange output considered expected behaviour?
The straight table for the top data set counts records where 'ss' is not null - as expected.
The straight table for the bottom data set counts records where 'dd' is null - this only happens when the column upon which the condition is placed contains ONLY nulls. This seems highly inconsistent from a mathematical standpoint - is it an expected behaviour of set analysis?
Thanks,
J
Now that I see your data, I understand what happens: There are no values at all in the field "dd". Then the selection {<dd={"*"}>} fails, and the count function will count all records.
However, if you add a value to "dd", everything works OK:
This is a known limitation, that only occurs for empty fields with no values at all.
No, it shouldn't count NULLs, so I suspect that something else is wrong. What happens if you add a second expression "=IsNull(dd)" ?
Hi.
I'm also attaching the file(s).
I wish I had an idea to contribute, but I'm just plain unsure of what's what. Any insights will be of help.
Thanks,
J.
Now that I see your data, I understand what happens: There are no values at all in the field "dd". Then the selection {<dd={"*"}>} fails, and the count function will count all records.
However, if you add a value to "dd", everything works OK:
This is a known limitation, that only occurs for empty fields with no values at all.
Thank you for confirming. I was not aware.
Do you know of a location where this and possibly other limitations of similar nature are documented?