Sergey B Kirpichev

https://dlmf.nist.gov/ (not free) http://functions.wolfram.com/ (same) http://ddmf.msr-inria.inria.fr/1.9.1/ddmf http://fredrikj.net/blog/2019/05/a-grimoire-of-functions/

https://proofwiki.org/ ?

https://planetmath.org/ ?

Yeah, probably there should be ``_eval_expand_mul`` helper for matrix expressions.

> Is there a way to re-use the same code that expands a*(b+c) in the scalar case? I doubt it's possible with current code organization. Basically, we have two Mul...

Correct. With current code organization, there should be a similar helper for MatMul. I don't like this approach, there is something wrong. Perhaps, symbolic arithmetics should be implemented like in...

Probably, porting of https://github.com/sympy/sympy/pull/15373 may solve this issue. But see https://github.com/sympy/sympy/issues/15665.

@jksuom, looking on [your comment](https://github.com/sympy/sympy/issues/20484#issuecomment-734106550) - it seems you are suggesting to merge eval/compose/subs/\_\_call\_\_ methods (or eval/compose/\_\_call\_\_ in the Diofant case). Or not?

Thank you for clarification. I also has doubts, that more generic evaluation rules will have application. So, the plan is: rewrote `compose()` to handle more generic substitutions of generators and...

Another concern with `p(1, 2)` syntax is that it's difficult to get partial evaluation "out the box". functools.partial seems to be too restrictive (default values, i.e. for which we get...