S.Y. Lee, in #20987 <https://github.com/sympy/sympy/issues/20987> It is outlined that there is a need to implement a DomainScalar class, which would help speed up eigenvector computation. I think that would be a more fruitful endeavour instead of working on these decompositions.
What do you think? On Saturday, 20 March 2021 at 6:26:07 pm UTC+5:30 Kartik Sethi wrote: > > Mind me jumping in the discussion for this, but I have some comments > Not at all. > S.Y. Lee what do you suggest I work on instead of these decompositions. I > have added some fraction-free algorithms which were listed on #20987 > <https://github.com/sympy/sympy/issues/20987>. I could not find any > paper/website for > Paterson-Stockmayer inverse so I went with decompositions instead. > > I am open to any suggestions > On Saturday, 20 March 2021 at 5:53:46 pm UTC+5:30 syle...@gmail.com wrote: > >> Mind me jumping in the discussion for this, but I have some comments >> >> The important aspect of DomainMatrix is that the computations should be >> 'rational' (which can be generalized up to 'multivariate rational function >> field') >> So I don't think that you can implement cholesky or hessenberg or >> householder easily. >> I have also though about adjoining the new square roots as algebraic >> extensions, >> But I'm afraid that you can't anticipate the speed just like numeric >> cases even if this approach is taken. >> On Sunday, March 14, 2021 at 10:16:15 PM UTC+9 kartik...@gmail.com wrote: >> >>> Oscar Benjamin, Have you had a chance to review my GSoC application >>> draft? >>> I've posted it here >>> https://github.com/sympy/sympy/wiki/GSoC-2021-Current-Applications >>> >>> On Friday, 12 March 2021 at 11:44:49 am UTC+5:30 Kartik Sethi wrote: >>> >>>> Oscar Benjamin, could you review this first draft of my proposal. >>>> Please let me know if you feel that I should focus >>>> on some other topics/ algorithms in DomainMatrix instead of the ones >>>> chosen by me. >>>> >>>> GSoC 2021 Proposal >>>> <https://github.com/ks147/GSoC/wiki/Gsoc_2021_application> >>>> On Tuesday, 9 March 2021 at 6:49:29 am UTC+5:30 Oscar wrote: >>>> >>>>> On Mon, 8 Mar 2021 at 16:00, Kartik Sethi <kartik...@gmail.com> >>>>> wrote: >>>>> > >>>>> > Oscar Benjamin, what if someone else also puts up this same proposal >>>>> to improve the DomainMatrix module for Gsoc. How does sympy decide which >>>>> person will work on it. >>>>> >>>>> There is plenty more work to do on DomainMatrix than one person could >>>>> do in a single GSOC project so there could be two people doing related >>>>> projects on this although we do have limits on how many projects a >>>>> single mentor could supervise. I am not the only person who could >>>>> mentor a project on DomainMatrix though and it's also possible that I >>>>> would mentor a project on something else instead such as ODEs. >>>>> >>>>> We will look at all applications and decide which to accept based on a >>>>> number of factors. The two major factors are the priority of the >>>>> project and the quality of the application. I think that DomainMatrix >>>>> is high priority but if there are stronger proposals for other >>>>> projects then it is possible that no proposal for DomainMatrix would >>>>> be accepted. >>>>> >>>>> I suggest not to overthink this. The main thing is just to have a good >>>>> application that is well motivated and shows clear understanding of >>>>> the work to be done. >>>>> >>>>> -- >>>>> Oscar >>>>> >>>> -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/b966cd23-2e9f-4e98-a1be-65e640ad05dcn%40googlegroups.com.
