This package provides a series of functions to optimize the modularity of bipartite networks, test its significance, and plot the results. The approach to using this package is as follows:

1. Prepare the data as an adjacency matrix (type Array{Int64,2}) containing 0 and 1
2. Assign an initial community partition using one of the different methods
3. Refine this structure using the recursive Brim approach
4. (Optional) Measure the roles of all nodes
5. (Optional) Test the significance of the modular structure using swaps
6. (Optional) Visualize the results

Functions that create an initial partition (partition_*) work on an adjancency matrix. All (almost all) other function work on a Modular object, which is returned by the starting partition functions. The Modular types stores both the adjacency matrix, and the module matrix.