Henrik Barthels

You mean a pattern where `f` is a variable that matches a function symbol, like `f_[a_, b_]` in Mathematica? Unfortunately, MatchPy doesn't support that yet. If the functions you are...

@wheerd should be able to give you a better answer than me. From the algorithmic point of view, right now I don't see a reason why it should be impossible....

I like the idea. Would you be interested in working on this? I currently don't have the time. My idea to implement this would be to define some special "neutral...

MatchPy does not exactly replicate the behavior of Mathematica. The reason is that there are some cases where Mathematica behaves in a way that could be considered inconsistent (Example: https://arxiv.org/pdf/1705.00907.pdf,...

Do you mean a combination of `SymbolWildcard` and sequence wildcards (the ones created with `Wildcard.dot()` and `Wildcard.star()`?

Looking at the code, I would say yes.

Thank you very much for letting us know! Are you planning to do any experiments to compare the performance of the generated C++ code to the generated Python code? I...

As a convention, operands in commutative functions are sorted in MatchPy. If I'm not mistaken, in that line we know that the operands come from a commutative functions because they...

> In order to reproduce this bug, using the branch in sympy/sympy#20612 Ah, now it makes a lot more sense that all the other stuff works. 😄 > is sorted(...

I don't think that removing or changing `sorted()` in `substitute()` is sufficient. I remember now that we sort operands in commutative functions to ensure that expressions that are equivalent modulo...