Fig. 4: The Gini index of row/column sparsity and the hyperdegree distribution.

The row/column sparsity is calculated by the number of non-zero elements in the row/column divided by the row/column size. Each purple point represents the frequency at which the corresponding hyperdegree value occurs. a–e The left part indicates the Gini index of row sparsity and column sparsity for the incidence matrix of each hypergraph dataset. The right part shows the hyperdegree distribution of five empirical hypergraph datasets.