About BOSS Data from Percival

Got word from Will Percival over the weekend:
He said the mask for the data/randoms is here:

https://trac.sdss3.org/browser/repo/boss/bosslss/trunk/geometry/current_boss_geometry.ply

He said:
I used the current_boss_geometry, choosing sectors with > 70% completeness. I don't have this separately in a file I'm afraid, but you could easily create it by placing the random catalog within the current_boss_geometry file - taking the quoted numbers as the completeness.

I said:
Thanks for the response. In terms of the 'current_boss_geometry' file. Is this as old as the randoms/galaxy files? Because I could imagine a scenario where the geometry file covers MORE area than the data files (because more of the sky has been observed) and then when I apply this "mask" to my other data it would let in regions of the sky which are not in the BOSS data files.

He said:
The current_boss_geometry file will always cover a larger area than that of the galaxies observed with completeness > 70%. But you can only use those polygons that contain one of the randoms. This will cut it back.

I said:
Also, how are you dealing with the spectroscopic completeness when generating the randoms? Are you basically filtering for each sky polygon based on the completeness in the current_boss_geometry file? For instance if a polygon has 80% completeness, then only 80% of randoms generated in that polygon are kept?

He said:
Yes, this is what I'm doing, based on the number of observed galaxies over the number of targets.