Are you the one that was saying the client wants you to do a t-test based on
using a Beta distribution rather than a Normal distribution. While that
doesn't compute for me, you might take this opportunity to ask Jerry - he has
the Phd in Statistics and may be able to figure out what you are talking
about.

Anyway, you might use dynamic range formulas to only do the Linest on
paired values. http://www.contextures.com/xlNames01.html#Dynamic

--
Regards,
Tom Ogilvy






" wrote:

On Jun 19, 3:16 pm, Tom Ogilvy
wrote:
Linest works on XY pairs. It isn't designed to work with missing data points.

In your example why wouldn't you use C3:C6, D3:D6

--
Regards,
Tom Ogilvy



" wrote:
The Excel function Linest does not return a value if your independant
and dependant variable count :i.e. Linest(C3:C18,D3:D6) do not match.
Is there a way to code this in VBA?- Hide quoted text -

- Show quoted text -

It's a problem when you have a data dump in columns (Linest at the
bottom cell) where there is varying rows between columns- trying to
automate the beta calculation whenever information is updated.