Fig. 2From: On the inconsistency of ℓ 1-penalised sparse precision matrix estimationPerformances of different methods on the bipartite graph model with 100 samples. (Lower values are better)Back to article page