Option for formulas to ignore excluded rows

[shampton82]

I would like to see an option for formulas to ignore excluded rows without having to build this into the formula using the "exclude state" formula function.  I think the easiet way for this to be implemented is to have it as a column property option in the formula column property option box

shampton82_0-1670599787726.png

So just one more check box here that would say "treat excluded rows as missing"

Thanks for considering this request.

Steve

[amanda5]

This would be great for formulas and for curve fitting!

[MeanChris]

My 2-cents, I'm a relatively new JMP user and was about to submit this as a bug to tech support.  I thought there was no way excluded rows should be included in formula results.  That seemed to be the main point of excluding them.  That to me was what differentiated Exclude versus Hide.

I suppose a check box for each column would be better than nothing, but in my opinion the default should be that all excluded rows are excluded from everything.  Maybe a table-wide setting would be more straightforward.

Regardless, since one previous commenter noted that this 'gotcha' was one of the first things they warn new JMP users about then it's probably not the ideal approach and worthy of being addressed sooner rather than later.  I will need to go back through everything I've built in the last few months and see if I made any ill-informed decisions based on the intuitive yet incorrect assumption of excluded row behavior.

[hogi]

@MeanChris  most important: please support the wish with a vote : )
unfortunately, comments don't count.

Up to now, this wish has 4 Kudus.
For the Developers this is a clear sign that just a minority of the users has issues with the current behavior -> nothing to be fixed.

Why a minority:
The users have to be in a narrow time corridor between not being affected and not being vulnerable anymore.

From my point of view,

- a user needs some experience to suffer from the bug (uses who don't know Column Formulas: they are safe)

- a user needs some experience to notice the bug (any user expects a default mode like in Graph Builder - if you don't expect the bug the chance is lower that you notice it)

- on the other hand, a user needs some "bad luck" - most users at this stage learned from a colleague : this is very dangerous!

To protect users in this intermediate level, I set up a community post to collect such Places where Jmp does something unexpected.

Here is another post with the  Tiny Traps in Jmp and JSL .

[hogi]

With JMP19, there will be scoped formulas which will fix the issues for many applications.

For many cases, but not all cases:
At the moment, scoping is just planned for Transform Formulas in Reports.
The same approach *) could fix the issue with Column formulas as well.

 *) via: column formulas with default option @excluded

[MeanChris]

@hogi Thanks for the link and the 'voting' tip.  I didn't realize it was the Kudos count that mattered most.  I'll use my new power responsibly.

[babn]

I have the same issue and just tried numerous approaches around it. None of them worked.

[hogi]

Hi @babn , in this post Caution: Places where Jmp does something unexpected 
I collected some approaches (just ones that work).

The easiest approach:

Col Mean(:height, excluded())

The disadvantage:
 excluded rows get "their own result".

[Ressel]

7 Kudos - not a lot, but very much looking forward to have this resolved.

[jwiltsie]

I was trying to determine if a dimension is meeting tolerance and I have lots of excluded rows. This work around leaves excluded rows as blanks. 

If(
	Excluded(), .,
	:OUTTOL == 0, 0,
	1
)

[hogi]

Agree. 

Col Mean ( 
    If(	Excluded(), .,
   	:OUTTOL )
)

On one hand, slightly more complicated than using excluded() as a grouping variable.
But without the disadvantage of:

excluded rows get "their own result".