In the case of boolean expressions (that return TRUE/FALSE), the dbl-neg
converts TRUE to 1 and FALSE to 0.

Example:
--ISNUMBER(56) converts TRUE to the number 1
--ISNUMBER("this is text") converts FALSE to the number 0

You could achieve the same results by multiplying a value by 1, but the
dbl-neg indicates to knowledgable users that a "type conversion" is being
effected.

Does that help?
--------------------------

Regards,

Ron
Microsoft MVP (Excel)
(XL2003, Win XP)

wrote in message
...
Just today I discovered that you can do some cool things w/
conditional sums of the form:

sumproduct(--(A1:A1000="West"),--(B1:B1000="Blue"),C1:C1000)

The above might give me sales in the Western region of Blue things
(column c might have the sales figures). This appears to avoid having
to enter arrary formulas (CTRL SHIFT ENTER as well as the likely
different specification for the formula itself) and would seem to be a
lot easier to work with.

What I haven't seen documented other than quick allusions to it is
exactly what the "--" is doing. Since "--" isn't easy to search for
on the web, I thought somebody here might be able to describe or refer
me to a documentation source online that describes what "--" (&
wrapping w/ () ) is doing. Seems to be a shorthand way to indicate
"do an array operation with this".

Thx!,

- TK