Figure 6
From: Selecting informative subsets of sparse supermatrices increases the chance to find correct trees
Examples of presence/absence B and edge-weighted B∗data used in simulations. Matrices of data availability in the upper panels, transformed edge-weighted matrices in the lower panels. All matrices are sorted. (A,B) 20% data availability, relative (A) low and (B) high heterogeneity of potential signal, missing data Gaussian distributed. (C,D) 10% data availability, relative (C) low and (D) high heterogeneity of potential signal, missing data following a power-law non-random distribution.