On Thu, Dec 17, 2009 at 6:07 AM, Dag Sverre Seljebotn
wrote:
> Patches up for a basic sparse fp matrix, #7723. That's the very basics,
> and if accepted I can move on to the fancier stuff -- I prefer getting
> corrections of course early...
>
> William Stein wrote:
>> I really hope that the Sage m
> I did think through whether these things should be cached or not and am
> prepared to write down my reasons for wanting them cached if anybody asks.
>
Putting at least some notes about this in docstrings/comments while
it's still in your head would probably be a good idea. (I haven't
looked at t
Dag Sverre Seljebotn wrote:
> Nick Alexander wrote:
>>> (In fact I'd wish for a convention that all matrix factorizations
>>> returned immutable matrices by default, since the factorizations are
>>> cached and one doesn't typically change them before using them.)
>> I think I wrote some (trivial!)
Nick Alexander wrote:
>> (In fact I'd wish for a convention that all matrix factorizations
>> returned immutable matrices by default, since the factorizations are
>> cached and one doesn't typically change them before using them.)
>
> I think I wrote some (trivial!) Cholesky decomposition code, an
Patches up for a basic sparse fp matrix, #7723. That's the very basics,
and if accepted I can move on to the fancier stuff -- I prefer getting
corrections of course early...
William Stein wrote:
> I really hope that the Sage matrix design is sufficiently flexible
> that it can be adapted to work
> (In fact I'd wish for a convention that all matrix factorizations
> returned immutable matrices by default, since the factorizations are
> cached and one doesn't typically change them before using them.)
I think I wrote some (trivial!) Cholesky decomposition code, and not
making the factorizat
On Thu, Dec 10, 2009 at 4:49 AM, Dag Sverre Seljebotn
wrote:
> I've started work on efficient sparse matrices over RDF/CDF, and here's
> my first round of questions/proposals. I commit to providing an
> implementation of these if accepted.
>
> 1) When multiplying sparse with dense, action.pyx conv
Dag Sverre Seljebotn wrote:
> 3) As mentioned earlier I'd like to implement explicitly diagonal and
> hermitian matrices (at least for RDF and CDF). Would this be OK?:
>
> sage: parent(hermitian_matrix(RDF, 3))
> Full MatrixSpace of 3 by 3 Hermitian matrices over Real Double Field
>
> sage: parent
I've started work on efficient sparse matrices over RDF/CDF, and here's
my first round of questions/proposals. I commit to providing an
implementation of these if accepted.
1) When multiplying sparse with dense, action.pyx converts sparse
matrices to dense matrices. I'm not sure about other use
