`plasmapy.formulary.quantum.chemical_potential` bug with powers

[StanczakDominik]

The current doctest error for plasmapy.formulary.quantum.chemical_formula() is a result of a floating point arithmetic error...

There is a large difference in n_e when we convert from cm ** -3 to m ** -3...

>>> (1.e27*u.m**-3) - (1.e21*u.cm**-3).to(u.m**-3)
<Quantity 1.37438953e+11 1 / m3>

Thus, even though the inputs are essentially identical, the following yields...

>>> chemical_potential(1.e27 ** u.m ** -3, 11000 * u.K)
<Quantity 2.00039985e-12>

>>> chemical_potential(1.e21 ** u.cm ** -3, 11000 * u.K)
<Quantity 5.1487703e-12>

@namurphy @StanczakDominik Thoughts??? If we can't figure out a quick fix, then I feel like we should exclude this code from the source until we determine a solution.

Originally posted by @rocco8773 in #722 (comment)

[rocco8773]

This issue has been raised on AstroPy (#9721).

[JaydenR2305]

@StanczakDominik I think this is working now, I get identical outputs for your given example case

[StanczakDominik]

Huh! Can you try removing @pytest.mark.xfail from the related tests in quantum to trigger CI runs for this?

[JaydenR2305]

I'm getting an assertion error for the known test case... Was the test case itself calculated incorrectly?

[StanczakDominik]

Could be! Can you post that as a PR?