https://bugs.documentfoundation.org/show_bug.cgi?id=97513
Heiko Tietze changed:
What|Removed |Added
Keywords|needsUXEval |
Component|LibreOffice
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
--- Comment #9 from Jaka ---
Just agreeing and noting that the wish to have unique counting as an option is
tracked here: https://bugs.documentfoundation.org/show_bug.cgi?id=90665
I wouldn't say it's a dependency though, since even
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
Pénzes Dávid changed:
What|Removed |Added
Blocks||103381
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
--- Comment #5 from Heiko Tietze ---
I was looking for a reason to change it in general. The more variants we have
in the program, like ifthen sum or count, the harder is it for users to figure
out why things
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
--- Comment #6 from raal ---
(In reply to Heiko Tietze from comment #5)
> So the question is why not always (row/col, with/without NAN) calculate the
> number of values?
Because when you change the default (from SUM to
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
--- Comment #7 from Milan Bouchet-Valat ---
That's right. Another point is that users already have to take into account
different behaviors depending on whether the values are numeric or not: sum
produces an empty
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
--- Comment #4 from Milan Bouchet-Valat ---
(In reply to Heiko Tietze from comment #3)
> Sounds reasonable. First step in an analysis should always be the checking
> of raw data, here the count of valid numbers,
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
Milan Bouchet-Valat changed:
What|Removed |Added
CC|
https://bugs.documentfoundation.org/show_bug.cgi?id=97513
--- Comment #3 from Heiko Tietze ---
Sounds reasonable. First step in an analysis should always be the checking of
raw data, here the count of valid numbers, followed by the computation of
descriptive or summative