[Pkg-octave-devel] Bug#706376: Bug#706376: Bug#706376: Bug#706376: Bug#706376: Bug#706376: Bug#706376: octave: sparse matrix n*2^16

[David Bateman]

Le 20 juin 2013 à 15:01, "John W. Eaton" <jwe at octave.org> a écrit :

> On 06/19/2013 08:32 PM, David Bateman wrote:
>> On 06/20/2013 01:10 AM, David Bateman wrote:
>>> I'd like to add some tests first and see if any other bugs have turned
>>> up after this change. For example the changes use made to sprand and
>>> sprandn 2 years ago to call randperm also overflows. At the moment I'm
>>> getting 791 failed tests with "make check" is that normal ? David
>> 
>> Ok, it seems Jaroslav's code for idx_vector(Sparse<bool>  hasn't been
>> used much in the last 5 years as it was completely wrong and when I
>> started using it, it caused 791 failures in "make check". I've fixed his
>> code as it makes sense to use it and pushed my changeset at
>> 
>> http://hg.savannah.gnu.org/hgweb/octave/rev/8fce0ed4894a
> 
> This change seems OK to me, but is there some reason to not use
> 
>  dim_vector dv = dims ();
>  return (dv.any_zero ());
> 
> as the default definition for is_empty?
> 
> jwe

Just the principal of minimum impact on existing code. Making this change everywhere will make this slightly faster too, though I doubt the calculation of numel is the limiting factor in most codes

D.