Closed fongkenrui closed 1 year ago
The changes look good. @ZedThree, do you think it's better to pass electron_mass or electron_temperature to functions like find_epsilon_para
? As an optional argument which is set to None
by default. This will mean we don't have to explicitly define a flag for relativistic corrections. I'm not sure which approach is best from a software engineering perspective.
Also, I've been thinking about ProfileFit. Do you think 1DProfileFit would be clearer? I'm on the fence about this too.
I wonder if your problems with pickle is a result of allclose
complaining about types. [Edit] I've looked at my old errors and it seems not.
Not had chance to look at this properly, but I'll try to do so tomorrow.
I suspect the error is due to something like trying to save an object to the npz file. I didn't see any likely candidates at first glance, but I'll dig into it tomorrow.
@valerian-chen Probably passing in temperature will be a bit cleaner
This looks great to me. @ZedThree, any final comments before we merge?
Ran into some conflicts with pytest test_simple_golden. Culprit seems to be under test_integrated line 449. Issue likely propagated from renaming of DensityFit to ProfileFit as well as the poloidal_flux_zero_density attribute being renamed -> some of the pytest modules importing from DensityFit have to be renamed. Not sure how it lead to the npz file being unable to be read without pickling enabled.
Error message dump: