There is certainly work to do with tensors in SymPy. Personally I would like to see more cohesion among tensor-like projects. There are a lot of tensor-like or indexed things in SymPy that all have valuable logic, none of which stem from some common framework. The original idea behind tensor core was to have some base layer from which all of the systems could benefit and from which pairs of systems could use to interact.
For example tensors as seen from a geometric point of view (sympy.tensor), matrix expressions as seen from a linear algebraic / matrix computations point of view (sympy.matrices.expressions) and arrays for code generation (sympy.tensor.indexed) all have some stuff in common. What is that stuff? Can we pull it out? On Wed, Mar 5, 2014 at 4:06 AM, F. B. <[email protected]> wrote: > > > On Wednesday, March 5, 2014 2:53:29 AM UTC+1, Charlie Paul wrote: >> >> I'm interested in doing the Tensor Core project for GSoC. >> > > That's good news. > > >> Other than the standard physics fare and a couple semesters of linear >> algebra, I've also taken differential geometry. >> > > SymPy currently also has a differential geometry module, with basic > support for tensors (covariant components only). > > http://docs.sympy.org/latest/modules/diffgeom.html > > >> >> Firstly, what is the status on SymPy's tensors compared to the Linear >> Algebra Vision document linked from the idea page? >> > > In sympy/tensor/indexed there is an old module. There is now a more > mathematically rigorous tensor module: > > http://docs.sympy.org/latest/modules/tensor/tensor.html > > It was initially meant to characterize abstract tensors (i.e. tensor > objects without components data). There is also support for data, but there > is still a lot of work to be done. > > The major inspiration for the new tensor module is *xAct*, a Wolfram > Mathematica package to do tensor manipulations. It has been used > extensively concerning symbolic math in General Relativity. xAct is open > source. > > It would be powerful to merge sympy/tensor/tensor tensors with those in > sympy/diffgeom. > > I see that there is a relatively recent tensor module (the aforementioned >> document is a couple years old). It looks like it isn't comprehensive, but >> I thought that I would check, as suggested on the GSoC page. >> > > I don't know if that page has been updated, in any case there have been > some additions. > > >> >> Also, I'd like to check the status of the Google Code and Github issues >> pages. Is one preferred and takes precedence over the other, or are they >> both used simultaneously? >> >> > I added a lot of issues concerning the new tensor module a month ago into > github. They are prefixed by *Tensor*. > > Thank you, >> >> > You're welcome. > > -- > 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 [email protected]. > To post to this group, send email to [email protected]. > Visit this group at http://groups.google.com/group/sympy. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sympy/3a260159-dee1-4d8a-9c54-9e69857e52f9%40googlegroups.com<https://groups.google.com/d/msgid/sympy/3a260159-dee1-4d8a-9c54-9e69857e52f9%40googlegroups.com?utm_medium=email&utm_source=footer> > . > > For more options, visit https://groups.google.com/groups/opt_out. > -- 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 [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAJ8oX-H8iRcQbYz8gnW8wV3WOKuFTr47BKveQ-Hc65CcdterwA%40mail.gmail.com. For more options, visit https://groups.google.com/groups/opt_out.
