Due to the sensitivity of VoidFinder to the choice of grid origin, it would be good to fix the origin to [0,0,0] for all void-finding (as discussed in UofR cosmology meetings).
Right now, changing the grid origin for SDSS DR7 from the minimum galaxy coordinates to [0,0,0] results in a negative buffer length calculated in _voidfinder.py. This stems from line 412, box = coords_max - grid_origin, which creates an array with negative elements and propagates into the buffer calculation.
If we want to enforce a common grid origin, one quick solution might be to assign the grid origin as the floor of the grid_origin user input, which should snap the grid to be aligned with [0,0,0]. Otherwise, we should find a working method to account for the sensitivity of VoidFinder to the grid.
Due to the sensitivity of VoidFinder to the choice of grid origin, it would be good to fix the origin to
[0,0,0]for all void-finding (as discussed in UofR cosmology meetings).Right now, changing the grid origin for SDSS DR7 from the minimum galaxy coordinates to
[0,0,0]results in a negative buffer length calculated in_voidfinder.py. This stems from line 412,box = coords_max - grid_origin, which creates an array with negative elements and propagates into the buffer calculation.If we want to enforce a common grid origin, one quick solution might be to assign the grid origin as the floor of the
grid_originuser input, which should snap the grid to be aligned with[0,0,0]. Otherwise, we should find a working method to account for the sensitivity of VoidFinder to the grid.