v1.0.0

[MetinSa]

This pull request will include all changes in the first new major ZodiPy release (v1.0.0) 🎉
 

Changes

• Implement suggestions from pyOpenSci review ZodiPy pyOpenSci/software-submission#161
• Create brand new interface with better Astropy integration
• Support specifying an obstime and obspos per coordinate to make ZodiPy better fit for time-ordered analysis.

The new public interface will look something like:
The idea is that at the end of this, ZodiPy should be used maybe as:

import astropy.coordinates as coords
from astropy.time import Time
import astropy.units as u

import zodipy

coordinates = coords.SkyCoord(
    20*u.deg, 
    30*u.deg, 
    obstime=Time.now(),
    frame=coords.BarycentricMeanEcliptic,
)

model = zodipy.Model(25*u.micron)

emission = model.evaluate(coordinates, obspos="earth")

This will require major changes to the public API.

TODO:

• Drop healpy as a dependency
• Move bandpass related calculations and the two arguments freq and weights to the ZodiPy initializer. This only needs to happen once when the model is initialized, and not every time the methods are run.
• Add new method which directly uses SkyCoord objects to unify theta, phi, lonlat, and obstime into one coord argument.
• Remove all other methods, so that we only have one evaluate method with a signature on the form of evaluate(coords, obspos). The previous get_emission_ang, ``get_binned_emission_ang, get_emission_pix`, and `get_binned_emission_pix` are all deleted. Rather than internally converting from healpix and `theta` `phi` to `SkyCoord` this now has to happen on the user end, but this is trivial with Astropy-healpix.
• Rename classes, methods, and arguments for a more a more consistent public API.
• Implement new interpolation since scipy.interp1d is depcrated Replace SciPy's interp1d with new intended interpolation API #27 Attempts to fix #27, currently fails due to keyword arguments in interp_kind #28
• Use astropy.models.physical_models.BlackBody instead of custom Plancks law implementation
• See if we can integrate synphot to do bandpass evaluations This was decided to be a bit overkill for the simple trapezoidal integration required by ZodiPy (see Use synphot package to perform bandpass integration #21 )
• Refactor the bandpass and blackbody internals to use astropy.modeling.physical_models.BlackBody
• Make sure ZodiPy is ready for Numpy 2.0 by running the new ruff rule ("NPY201")
• Rewrite tests to use new API and get back to 99% coverage
• Rewrite usage with new example on readme, index, and usage pages.
• Re-enable ruff on docs.
• Test if calling 'cartesian.xyz' is a cached property or a computation.
• Rewrite tabluate_number_density function.
• ~~Make the multiprocessing.Pool object be injected instead of created if desired. ~~This does currently not work as intended. Move to another release or drop feature until requested.

[nostutzer]

Looks good :)