The R blockcluster package provides algorithms for extracting a checkerboard structure in which the number of blocks in the checkerboard is equal to the number of row clusters times the number of column clusters. The analyst sets the number of row clusters and the number of column clusters. In the example code for the animals biclustering, we see the following code requesting a solution with four row clusters and four column clusters (a checkerboard with 16 blocks): set.seed(1) out = cocluster(animals_matrix, datatype = "continuous", nbcocluster = c(4, 4)) summary(out) How would you propose determining the number of row clusters and the number of column clusters in biclustering?