Oscar Benjamin

So there are some failures but they mostly seem to come down to this: ``` def _term_div(self): > raise NotImplementedError E NotImplementedError ``` I would have thought that the `_term_div`...

It is time to stop force pushing. We need to see a record of which tests were passing or failing at each commit.

```diff diff --git a/sympy/polys/ring_series.py b/sympy/polys/ring_series.py index 4c839eb0e2..64822d2ffd 100644 --- a/sympy/polys/ring_series.py +++ b/sympy/polys/ring_series.py @@ -1891,10 +1891,10 @@ def rs_hadamard_exp(p1, inverse=False): p = R.zero if not inverse: for exp1, v1 in p1.items():...

It is probably just a mistake in the docs. PRs welcome.

The fmpq_mpoly type is for multivariate polynomials over the rational numbers. If the docstring says otherwise then that it is probably because part of the docstring was copied over at...

Feel free to open a PR changing the docs.

> So why do the docs not say for `fmpq_mpoly` type that it "_represents sparse multivariate polynomials over the **rationals.**_ "? This has been fixed now in gh-320 although it...

See also gh-26743, gh-26740, gh-8552, gh-21343, gh-15753.

I think there was previous discussion that maybe this simplification should not happen automatically. When you look at something like gh-26816 it seems obvious that the simplification is desired. The...

> how about implementing `CRootOf.invert()` That seems reasonable to me. Also it would be good to have something that can rationalise the denominator to bring all RootOf into the numerator.