> The exact details of what you would do if your proposal was accepted > are less important than showing that you can make substantial > improvements in general. It's very likely that if any proposal based > on DomainMatrix is accepted then the details of what would get > implemented would end up different from any proposal made now because > part of the work is in identifying what should be done.
Thanks for clarifying this. I am reasonably familiar with Jordan Canonical Form and Matrix exponential and I think I can implement them. I'll add these to my proposal instead of QR and Cholesky Decompositions. Thanks Kartik On Sunday, 21 March 2021 at 4:45:10 pm UTC+5:30 Oscar wrote: > On Sun, 21 Mar 2021 at 10:51, Kartik Sethi <kartik...@gmail.com> wrote: > > > > There is already a PR #21120 which is trying to implement DomainScalar > class > > which is why I deleted that message. > > Personally I read these messages as an email mailing list so deleting > a message is unnoticeable to me (you can't delete an email from my > inbox). > > > Should I add Jordan Normal Form to my > > Proposal instead assuming that DomainScalar would be implemented by then > or should > > I add improvements to DomainScalar as one of the objectives in my > proposal. > > I think that improving Jordan form and matrix exponential is > definitely something to do since currently the main usage of > DomainMatrix outside of linsolve is eigenvects and it clearly shows a > big improvement there. The Jordan form calculation is very similar and > is actually used more widely e.g. for computing the matrix exponential > and for solving systems of ODEs so applying the same improvement there > should be a high priority for DomainMatrix. > > > I am a little confused about this. > > The most important thing for a GSOC proposal is to demonstrate that: > > 1) You understand the parts of the codebase you are proposing to work > on and have ideas to improve things. > 2) There are substantial improvements to be made that are of value to > sympy. > 3) You individually are capable of implementing *some* of those > substantial improvements. > > The exact details of what you would do if your proposal was accepted > are less important than showing that you can make substantial > improvements in general. It's very likely that if any proposal based > on DomainMatrix is accepted then the details of what would get > implemented would end up different from any proposal made now because > part of the work is in identifying what should 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/e5a998cc-c272-4b70-93dd-109806c9cadcn%40googlegroups.com.