To determine the statistical significance of the over-densities we identify in Fig. 6, we bin the data in a 2D grid and count how often we observe as many or more points in the randomised realisations compared to the real data set. For clarity, only the bins with a frequency of such an occurrence of 1% or less are coloured. The left panel shows a 8 × 8 grid, while the right one shows a 16 × 16 grid.