Copied from office-comment list

Original author: "David King" <[email protected]>
Original date: 22 Dec 2008 23:28:57 -0000
Original URL: http://lists.oasis-open.org/archives/office-comment/200812/msg00045.html

> > TRIMMMEAN page 379
> > * "Let n denote the number of elements in the data set" - that's
> > too vague - does it mean elements including or excluding non-
> > numbers/blank cells? Calc ignores non-numbers.
>
> That is expressed by declaring the dataSet parameter as NumberSequence.

I disagree

This highlights my unease at having 'conversion types' - I would
be happier if all parameters were defined as 'types', so
NumberSequence would be a 'type' that triggered conversion. I'd
like to be able to look up any parameter type in the same place.

In this case it's not clear whether the number of elements is
counted before or after conversion.

"Let n denote the number of elements in dataSet after conversion
to NumberSequence" is a bit better, but still less than clear,
because dataSet is defined as a NumberSequence as well as being
converted to a NumberSequence.

Actually that's the essence of the prob for me - you can't define
a parameter as a NumberSequence and then convert it to a
NumberSequence....