Check if we are cross-correlating a quasar with its own forest or not.

[alxogm]

Recently saw this warning appearing, when running picca_xcf.py on mocks,

computing xi: 1.49%
WARNING: 1 pairs have cos>=1.
WARNING: 1 pairs have cos>=1.
WARNING: 2 pairs have cos>=1.
WARNING: 1 pairs have cos>=1.
WARNING: 1 pairs have cos>=1.
WARNING: 1 pairs have cos>=1.

the specific case was resolved in #813, but it rises the question of whether we are cross-correlating a quasar with its own forest or not, so we should check.