Fix warnings

[mgeier]

I don't know if those are in any way harmful to us, but it's probably good to have a look.

E.g. at https://github.com/sfstoolbox/sfs-python/actions/runs/17990398126/job/51178925090:

/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:134: RuntimeWarning: invalid value encountered in divide
  d = d + 1/epsilon[m] * _np.exp(1j*nu*_np.pi/2) * _np.sin(nu*phi_s) \
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:134: RuntimeWarning: divide by zero encountered in divide
  d = d + 1/epsilon[m] * _np.exp(1j*nu*_np.pi/2) * _np.sin(nu*phi_s) \
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:134: RuntimeWarning: invalid value encountered in multiply
  d = d + 1/epsilon[m] * _np.exp(1j*nu*_np.pi/2) * _np.sin(nu*phi_s) \
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:294: RuntimeWarning: invalid value encountered in divide
  f = 1/epsilon[m] * _np.sin(nu*phi_s) * _np.cos(nu*phi) * nu/r
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:294: RuntimeWarning: divide by zero encountered in divide
  f = 1/epsilon[m] * _np.sin(nu*phi_s) * _np.cos(nu*phi) * nu/r
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:295: RuntimeWarning: invalid value encountered in multiply
  d[idx] = d[idx] + f[idx] * _jn(nu, k*r[idx]) * _hankel2(nu, k*r_s)
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:294: RuntimeWarning: invalid value encountered in divide
  f = 1/epsilon[m] * _np.sin(nu*phi_s) * _np.cos(nu*phi) * nu/r
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:294: RuntimeWarning: divide by zero encountered in divide
  f = 1/epsilon[m] * _np.sin(nu*phi_s) * _np.cos(nu*phi) * nu/r
/home/runner/work/sfs-python/sfs-python/sfs/fd/esa.py:295: RuntimeWarning: invalid value encountered in multiply
  d[idx] = d[idx] + f[idx] * _jn(nu, k*r[idx]) * _hankel2(nu, k*r_s)

 /home/runner/work/sfs-python/sfs-python/doc/examples/animations_pulsating_sphere.py:20: DeprecationWarning: __array_wrap__ must accept context and return_scalar arguments (positionally) in the future. (Deprecated NumPy 2.0)
  scat = sfs.plot2d.particles(grid + displacement, **kwargs)
/home/runner/work/sfs-python/sfs-python/doc/examples/animations_pulsating_sphere.py:23: DeprecationWarning: __array_wrap__ must accept context and return_scalar arguments (positionally) in the future. (Deprecated NumPy 2.0)
  position = np.real((grid + displacement * phasor**i))
MovieWriter imagemagick unavailable; using Pillow instead.
/home/runner/work/sfs-python/sfs-python/doc/examples/animations_pulsating_sphere.py:23: DeprecationWarning: __array_wrap__ must accept context and return_scalar arguments (positionally) in the future. (Deprecated NumPy 2.0)
  position = np.real((grid + displacement * phasor**i))
/home/runner/work/sfs-python/sfs-python/doc/examples/animations_pulsating_sphere.py:48: DeprecationWarning: __array_wrap__ must accept context and return_scalar arguments (positionally) in the future. (Deprecated NumPy 2.0)
  np.real(quiv.set_UVC(*(velocity[:2] * phasor**i)))
MovieWriter imagemagick unavailable; using Pillow instead.
/home/runner/work/sfs-python/sfs-python/doc/examples/animations_pulsating_sphere.py:48: DeprecationWarning: __array_wrap__ must accept context and return_scalar arguments (positionally) in the future. (Deprecated NumPy 2.0)
  np.real(quiv.set_UVC(*(velocity[:2] * phasor**i)))

[fs446]

I am aware of these nasty divide by zero warnings. They occur in the series expansion. For the shown examples they should not be critical. However, I'd suggest to find examples where these warnings do not occur. This might require to not set up a loudspeaker at exactly the edge position, which is the bad guy if I correctly remember.

The array_wrap warning needs probably fixed soon. I changed something here 8f63770
maybe that's related to it.