Describe your changes

This adds a keyword argument to the compute_coeffs_discrete() function to allow specifying a parallel processing pool when computing the coefficients for a particle collection. This parallelizes over (n,l,m), so it is faster when the number of particles is large and nmax, lmax is large.

Checklist

• Did you add tests?
• Did you add documentation for your changes?
• Did you reference any relevant issues?
• Did you add a changelog entry? (see CHANGES.rst)
• Are the CI tests passing?
• Is the milestone set?