Am Sonntag, 21. Juni 2015 03:56:43 UTC+2 schrieb Nils Bruin:
>
> On Saturday, June 20, 2015 at 3:34:06 PM UTC-7, Martin R wrote:
>>
>>
>> What is the bug here? Do you think matrix(2,3,1) should succeed? Do you
>> think it should produce a different error message?
>>
>> No, I think that matrix(nrows, ncols, element) sould fail, unless element
>> is callable or iterable.
>>
>
> We'd need to do serious *extra* work and checking for that. Currently, if
> MatrixSpace(R,n,m)(L) works then matrix(R,n,m,L) will work too. Making sure
> that matrix(R,n,m,L) breaks for some L where MatrixSpace succeeds would
> require extra testing and explicit error raising.
>
That was my initial question in this thread: I do not understand where (or
why) MatrixSpace calls _matrix_constructor. I set a breakpoint in
_matrix_constructor and did M=MatrixSpace(SR, 2,2) and then trace("M(x)").
The breakpoint was never reached.
> Would that really be an improvement? To ensure that the following happens?
>
> sage: matrix(2,2,x+y)
> [0 1]
> [1 2]
>
> (assuming x+y would be recognized as callable before it is recognized as
> iterable)? I think the current answer is less surprising.
>
Yes, in my opinion certainly.
Martin
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.
