Same sub-sample definition for all 3D correlation

[londumas]

It would be nice, if possible, to have the same sub-sample definition across all calculation of the different 3D cross-correlation. I don't really complettly understand how the division is made currently, but maybe if the sub-sample division was based on the position on the sky of the mean of the pair, this could work.

We should think if there is a simple way and computing time inexpensive way to do such a thing.

[ngbusca]

I don't see an obvious way to do this, do you have any ideas?

[londumas]

sadly not yet, but we could think about it and thus see if there is one simple way.

[londumas]

@ngbusca, don't we have already a division on the same number of healpix ? The only problem is in witch pixel should we put a pair when the two line of sights don't belong to the same pixel.