Kaya Unalmis

> Here, the function does not produce incorrect results; rather, the domain is restricted compared to some versions of the function, and this restriction is not mentioned in the documentation....

also related: https://github.com/PlasmaControl/DESC/pull/1006#issuecomment-2091007113

> FAILED tests/test_plotting.py::TestPlotBoundary::test_plot_boundary_surface - pluggy.PluggyTeardownRaisedWarning: A plugin raised an exception during an old-style hookwrapper teardown. > > But the test works on my local system. > > Mention this during...

Figures are only compared with the `--mpl` flag

I would like to use a Hermite spline for |B|, but it looks like interpax doesn't support that. Should I add that?

> This should be close to the elliptic integrals (maybe differs by a constant factor). If you are busy, I can add this test on Tuesday or Wednesday. If you...

I've added more tests. I would be surprised if there are any bugs in the bounce integrals routine. The bounce points are tested to be computed correctly, after which it...

My jax version allows for excluding keyword arguments in `jnp.vectorize`. Apparently some don't, so it must be a new feature, and we’ll either need to update the minimum `jax` version,...

Just copying and pasting from a few different comments, hopefully context is still clear: > Since there will only ever be two types of bounce integrals, I think we should...

> When I run pytest -k test_bounce_averaged_drifts, I get the following error > > ``` > tests/test_bounce_integral.py:43: in > @np.vectorize(signature="(m)->()") > E TypeError: vectorize.__init__() missing 1 required positional argument: 'pyfunc'...