OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

office-formula message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: Re: [office-formula] BETADIST parameter Cumulative

Hi Andreas,

On Tuesday, 2007-03-20 13:13:24 -0600, Andreas J. Guelzow wrote:

> > > Mathematically, the cumulative distribution and the density are defined
> > > for all values of x, so there is no justification to have an constraint
> > > such as a <= x <= b. The answer outside that range should just be 0 (or
> > > 1 in the case of cumulative if b < x).
> > 
> > This is also what our current semantics describe.
> > 
> > > The question is now mathematically correct the function definitions
> > > should be.
> > 
> > Well, as correct, exact and detailed as possible. In this case I think
> > we don't miss anything, or do we?
> Well, do we agree that BETADIST like the ...DIST functions above should
> have an optional "cumulative" parameter?

I don't see a reason to not have it except that so far no application
seems to implement it. So yes, I think we agree.

> About the constraint a <= x <= b that you suggested, what is its
> purpose?
> Sorry, I may have misunderstood your initial message.

That indeed may have been a misunderstanding. Currently, applications do
have that constraint, but our definition in the spec does not have it,
instead it specs to return 0 respectively 1 in case x is outside those
ranges. So, it seems we would be fine if we didn't change anything there.

I'm just not sure if and how users rely on the behavior that giving an
x outside the bounds [a,b] produces an error. Or may there be other
reasons why an error should be generated? The problem is, as usually,
that when migrating from "that other application" users expect to get
the same behavior, be it reasonable or not.

I'm fine without the constraints if we can say that there is no reason
to have them.


Automatic string conversions considered dangerous. They are the GOTO statements
of spreadsheets.  --Robert Weir on the OpenDocument formula subcommittee's list.

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]