The current implementation causes a clustering of the global ellipticities around the grid values, particularly for galaxies that are close to face-on. In order to increase the resolution of the grids and mitigate this issue, we have introduced a second pass to search for the optimal grid ellipticity. This approach results in a ~10 times increase in grid resolution, with only a ~10% increase in computing time during the initialization stage. Our tests on ~8000 galaxy images show that this second pass effectively removes the clustering, as can be seen in the attached plots.
The current implementation causes a clustering of the global ellipticities around the grid values, particularly for galaxies that are close to face-on. In order to increase the resolution of the grids and mitigate this issue, we have introduced a second pass to search for the optimal grid ellipticity. This approach results in a ~10 times increase in grid resolution, with only a ~10% increase in computing time during the initialization stage. Our tests on ~8000 galaxy images show that this second pass effectively removes the clustering, as can be seen in the attached plots.