How to ignore this filter in set analysis ?

[Derek_T]

I know the question have been asked in many times, but none of the solutions I've found give me the expected result.

I have 2 tables, and the only filter in my analysis is a date range: 2021-01-01 to 2022-01-01

I haven't shown all the date because their value are irrelevant, only the boundary matters.

What I'm trying to build is the table below, the trick is "MAX_DATE" and "AVG_AC_COUNT" should be unaffected by the ATA, only the "SUM_AMOUNT" should be impacted.

I've tried many combinations of set modifiers, but I always end up with the "MAX_DATE" and "AVG_AC_COUNT" being affected by the ATA where they should only be affected by the date range selected by the user.

Thanks !

[edwin]

in your expressions for AVG_AC_COUNT and MAX_DATE, add the set analysis modifier ATA=

this means for this expression ignore ATA filter

since you want SUM_AMOUNT to consider ATA filter, then dont add this into that expression

[Derek_T]

Thanks Edwin,

could you elaborate the code a bit more please, I've tried passing ATA but I don't know how to exactly, I've tried:
{1<ATA={*}>} but it does not work.

[edwin]

to ignore ATA the set analysis should be

{1<ATA=>} 

[edwin]

you may want to try 

{<ATA=>}

if you add 1, it means ignore ALL filters but you may want to ignore only specific filters

[Derek_T]

Thanks Edwin, I've tried both and it does not work.
Here's the formula for the cumulative amount:  Avg({1<ATA=>}AC_COUNT) / Avg({<ATA=>}AC_COUNT)

But I still get completely different number by ATA, which does not make sense since each AC (Aircraft) is composed of the same set of ATAs, so the count of AC / ATA should remains the same.

[Derek_T · Accepted solution]

The solution was to ignore the dimensions ATA:

Avg(TOTAL AC_COUNT)

[edwin]

you may want to test that.  without set analysis, there is no ignored filters

[Derek_T]

You're right, being new to Qlik my choice of words was incorrect, it's not a "filter" I was trying to ignore but a "dimensions" in a table. Thank you !

[edwin]

not intending to belabor this but filter and dimensions are interchangeable.  a dimension when selected IS a filter.  what i am saying is that without set analysis, you are not ignoring filters nor dimensions.  however, if this works for you then maybe you never had an issue in the first place.