Harlan Grove wrote:
> . . .
> But more to the point, if all one wants to do is *COUNT*
> the distinct entries in a range, why would one ever want
> to use a slow udf rather than the much faster
>
> =SUMPRODUCT((A1:A15<>"")/COUNTIF(A1:A15,A1:A15&""))
>
> ?
>
Because the range may contain error values.
Alan Beban
Bookmarks