Bill Barker wrote:
> Currently StatUtils.sum(double []) (as well as most other statistics
> that operate on arrays) returns NaN on a zero length array. As pointed
> out in the Jira issue, this is not the conventional treatment of a
> summation over the empty set (which would return zero) in the
>
On 10/06/2010, Ted Dunning wrote:
> I would count this as a bug fix rather than a compatibility break. (my vote
> is non-binding, of course)
>
I agree that the current behaviour is wrong.
Unfortunately, the current behaviour is documented in the Javadoc, so
it would be a compatibility break.
- "Bill Barker" a écrit :
> Currently StatUtils.sum(double []) (as well as most other statistics
> that
> operate on arrays) returns NaN on a zero length array. As pointed out
> in
> the Jira issue, this is not the conventional treatment of a summation
> over
> the empty set (which woul
I would count this as a bug fix rather than a compatibility break. (my vote
is non-binding, of course)
On Wed, Jun 9, 2010 at 7:08 PM, Bill Barker wrote:
> Currently StatUtils.sum(double []) (as well as most other statistics that
> operate on arrays) returns NaN on a zero length array. As poi
Currently StatUtils.sum(double []) (as well as most other statistics that
operate on arrays) returns NaN on a zero length array. As pointed out in
the Jira issue, this is not the conventional treatment of a summation over
the empty set (which would return zero) in the mathematical world. I wo
