[prev in list] [next in list] [prev in thread] [next in thread] 

List:       koffice-devel
Subject:    Re: KSpread: A-functions
From:       "Michael Fair" <michael () daclubhouse ! net>
Date:       2005-08-06 22:46:06
Message-ID: 001301c59ad8$a4c05a10$6400a8c0 () maryfair
[Download RAW message or body]

The A functions include non-numerics as values.

Count will count all valid numbers in the list
CountA will count all valid Values in the list (only ignoring the null 
cells).

Count(1,"one",2,"two")=2
CountA(1,"one",2,"two")=4

I'm not sure what SumA does, but I know that Sum() does not
translate non-numeric values.
So Sum("1", TRUE,2) = 2 because the only the numeric values are considered.

Average ignores non-numeric values
AverageA treats non-numerics as 0 except true which is 1

If you can get a hold of it, the specifics of all this are the in the 
Microsoft
Office help files, perhaps even online.  Having made workbooks that
have to take in data from other vendors on one tab, and then combine
them all on another I see the needs for both versions.

-- Michael --

----- Original Message ----- 
From: "Tomas Mecir" <mecirt@gmail.com>
To: "For developer's discussion about KOffice" <koffice-devel@kde.org>
Sent: Thursday, July 28, 2005 2:14 AM
Subject: KSpread: A-functions


Hi everyone !

Not sure if you've noticed, but I've just enabled the new formula
engine (yay!), thus lotsa things have changed :D

One thing that I'm a bit uncertain of, is the following one:
Some range functions exist in two versions, with one of tham having
the A suffix. SUM vs. SUMA, COUNT vs. COUNTA, and so on.
In excel and friends, the difference is that the A version treats
boolean values as 1s and 0s, whereas the non-A version ignores them or
fails on them or I-don't-know-what. With all the new technology in
KSpread, however, boolean values are automatically converted to
numbers when needed (and so are any other data types). Hence, the
A-functions are a bit redundant.

The only possible use for these would be so that the A-versions would
ignore empty cells, whereas the non-A ones would include them.
COUNT/COUNTA already behave like this.

So, the question is - any idea what to do ? Should we stay
excel-compatible even in these minor cases ? Note that full
compatibility is not possible, due to excel's value handling being
much less powerful than ours ( - feeling a bit of personal pride :D
-). Or dump A-versions entirely?

Hmpf. Stupid excel.

/ Tomas
_______________________________________________
koffice-devel mailing list
koffice-devel@kde.org
https://mail.kde.org/mailman/listinfo/koffice-devel

_______________________________________________
koffice-devel mailing list
koffice-devel@kde.org
https://mail.kde.org/mailman/listinfo/koffice-devel

[prev in list] [next in list] [prev in thread] [next in thread]