Blank outputs are treated as zero

[CaseyL]

I have recently begun using yasai and I think its a great product.  Thank you for sharing it.  I have encountered only one issue/limitation.  In excel blank out puts are not included in counts and averages etc.  However blank outputs are included as a zero in Yasai output summaries.  One of my models outputs is the years till a fund balance is depleted. In approx. half of simulation runs the balance is not depleted which is reported as a blank in the function.  The simoutput function targets this cell which either reports years till depletion or blank if the fund was not depleted.  The simulation output summary reports blank cells as zero.  I would like the simulation output and resulting minimum, mean, median and standard deviation to reflect only instances when the fund was depleted.  Is there some type of workaround within excel or another solution within yasai?  Any thoughts would be appreciated.  

[CaseyL]

Disregard my post.  I read your initial article describing yasai which included the solution of inserting an if statement into the output name condition.  Thanks.

[Jonathan Eckstein]

Great. I was just about to e-mail you a suggestion to look at that
article, although I could not understand exactly what the problem was.

Prof. Eckstein

[Charles Grace]

Casey,
I had the same problem last year, where I wanted Yasai to ignore empty
cells, versus its default of treating them as zero values. I tried the
IF statement approach without any joy (although Prof E and the board
offered plenty of help). Let us know if you are successful and I'll give
my problem another try.

Good luck.

Charlie

[Jonathan Eckstein]

I am planning another release this spring and will think about making a
change to ignore blank values passed to SIMOUTPUT (I have to think it
through first).

If you want to play around with the source, I think this should be a
really easy change...

J E

[bar...@willamettepartnership.org]

Hi Jonathan,

Did you ever come up with anything to ignore blank values?

-Sam

[Jonathan Eckstein]

It's #1 on the to-do list, but we didn't address it yet.

However, it occurs to me there is a simple workaround that will work if
you have an output that is sometimes blank.

See the attached sheet. The only cells with formulas are A1:D1. D1 is
an example of how you deal with an output that can be blank. Basically,
you assign the output two different names depending on whether it's
blank or not, and it will get recorded as two different outputs, each
with its own sample size. The name assigned to non-blank value (in this
case, "Without blanks") will have its statistics unpolluted by zeroes
from the blanks.

Does that help?

Jonathan

> --
> You received this message because you are subscribed to the Google
> Groups "yasai-simulation" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to yasai-simulati...@googlegroups.com
> <mailto:yasai-simulati...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.

[bar...@willamettepartnership.org]

Thanks Jonathan,

As I wrote in haste, I realized we aren't quite talking about the same issue. I am dealing with a model variable that is sometimes blank. Because my subsequent formulas average and sum variables, blanks do not impact the output, whereas zeros do. However, in running the sensitivity analysis, I assume that blanks are interpretted as zero's. I am not quite sure if it is possible (and mathematically sound) to run a sensitivity analysis with sometimes blank variables.

-Sam

On Thursday, September 18, 2014 3:28:50 PM UTC-7, Jonathan wrote:

It's #1 on the to-do list, but we didn't address it yet.

However, it occurs to me there is a simple workaround that will work if
you have an output that is sometimes blank.

See the attached sheet.  The only cells with formulas are A1:D1.  D1 is
an example of how you deal with an output that can be blank.  Basically,
you assign the output two different names depending on whether it's
blank or not, and it will get recorded as two different outputs, each
with its own sample size.  The name assigned to non-blank value (in this
case, "Without blanks") will have its statistics unpolluted by zeroes
from the blanks.

Does that help?

    Jonathan

> <mailto:yasai-simulation+unsub...@googlegroups.com>.