Further refinement of IndexedBase and Idx

[Björn Dahlgren]

Hi!

Since I haven't opened any particular issues but only PR's I thought maybe I should announce them here so they won't go unnoticed.

https://github.com/sympy/sympy/pulls/bjodah

They are:

• Avoid superfluous CSEs from Idx and IndexedBase instances

• Makes IndexedBase consistent with Symbol wrt to latex printing

• Fix coeff for indexed

I made all three PRs from modifications I needed to do in order for me to productively use SymPy for code generation of finite difference
formulae in a larger C++ code base. Hence they are in the spirit of using tensor.Indexed, tensor.Idx and tensor.IndexedBase in a more general
setting than just tensors. (I could picture them being moved to core at some point even).

I have some more modifications in the pipline but I already forsee merge issues wrt to changing argument orders in docstrings and tests,
so before I submit any more PRs I think it would be good to have these reviewed and merged (or rejected if you disagree with me).

Looking forward to your feeback (those of you interested)!

Regards
/Björn

[Øyvind Jensen]

Hi Björn,
It is really cool that you are working on this, thanks for the PRs! I am not uptodate on the inner workings of SymPy any longer, but I will take a look at the patches and see if there is anything I can add.

Øyvind

[Björn Dahlgren]

Hi Øyvind,

Thank you for taking your time looking at the PR's.

Øyvind raised an important concern regarding ordering of terms and arguments in my test strings: https://github.com/sympy/sympy/pull/7355

Are these platform dependent (still)? If so: I might need to make those tests more robust not to unsettle users of other platforms
who might get test failures (even though those test failures would be benign since actual code generated would still work as expected)?
Unfortunately I do not have access to Mac / Windows atm.

[Aaron Meurer]

No, they shouldn't be. All the tests are run with hash randomization
enabled, so if they varied, you would notice right away. The
"platform" in question here is just 32-bit vs. 64-bit, which has
different hash values, but with python -R (or Python 3.3+) hash values
are randomized. But for some time now SymPy has been written so that
the output of anything doesn't depend on hash values.

Aaron Meurer

> --
> 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+un...@googlegroups.com.
> To post to this group, send email to sy...@googlegroups.com.
> 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/a0470860-bce5-4f22-9b6a-3131f629e585%40googlegroups.com.
>
> For more options, visit https://groups.google.com/d/optout.