Obtain a_lms (spherical harmonic decomposition) of a cut sky in healpy

I am using HEALpy to work with some sky maps and I need to obtain the spherical harmonic decomposition of these. Usually I would use map2alm, but I am using partial sky maps, so I have some pixels which we have no information from. Is there a way to handle this with healpy?

Sometimes I have just given the value 0 to the pixels from which we have no information, but this is not accurate, since edge effects appear and the transformation is not accurate enough.

For an approximate solution, you can remove the mean of the map with healpy, so that you reduce the boundary effects:

m_masked = hp.ma(m)
m_masked -= m_masked.mean()

Then you can apodize your mask using hp.smoothing so that you reduce the effect of the discontinuity between the observed and unobserved sky.

For a proper analysis which corrects for the impact of the cut sky on your Spherical Harmonics transform you should use NaMaster instead of healpy:

https://namaster.readthedocs.io/en/latest/index.html