[Math] Why no "AnyVector" Interface?

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

[Math] Why no "AnyVector" Interface?

Bruce A Johnson-2
Hi,

I'm currently working on a scripted interface to Commons Math.  I support both Real and Field (Complex) matrices, and Real and Field (Complex) vectors.  Checking the size of the matrices is trivial because they both implement AnyMatrix which has row and column size getters, but because there is no AnyVector class I have to deal with the two vector types separately.  Since both RealVector and FieldVector already implement getDimension() it would seem trivial to add an AnyVector class with that method.

Is there a reason AnyVector doesn't already exist?  If not it would seem a worthwhile (and quick) thing to add for 3.0.

cheers,

Bruce


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Why no "AnyVector" Interface?

Gilles Sadowski
Hello.

>
> I'm currently working on a scripted interface to Commons Math.  I support both Real and Field (Complex) matrices, and Real and Field (Complex) vectors.  Checking the size of the matrices is trivial because they both implement AnyMatrix which has row and column size getters, but because there is no AnyVector class I have to deal with the two vector types separately.  Since both RealVector and FieldVector already implement getDimension() it would seem trivial to add an AnyVector class with that method.
>
> Is there a reason AnyVector doesn't already exist?  If not it would seem a worthwhile (and quick) thing to add for 3.0.

-1

I don't mean to argue that this proposal is wrong, just that no new feature
should be added at this point, as we try to get to a long overdue release.

Moreover, there has been recent dicussions over changing the design of the
matrix interface(s); it would be a contradiction to do as if everyone were
statisfied with the current design.


Best,
Gilles

P.S. Of course, you are very welcome to (re-)open the discussions after 3.0
     is out.

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]