MatrixNetworks.jl
MatrixNetworks.jl copied to clipboard
Graph and Network algorithms in Julia
@JuliaRegistrator register() Release notes: This version uses GenericArpack.jl
Maybe can be faster with an implementation of the Hopcroft Karp algorithm but read that this is normally not faster than the basic BFS implementation. If this looks good I'll...
This should give something more reasonable ~~~~ scomponents(erdos_renyi_directed(0,0)) ~~~~
It'd be lovely to have Julia code for the metis hierarchy. ``` hier = metis_hierarchy(A;...) hier.nlevels hier.graphs # Arrary of SparseMatrix types... ``` along with some restriction and propagation functions...
I think compose gives a weird error with the current master branch. ``` using Compose using MatrixNetworks ``` ``` WARNING: New definition *(MatrixNetworks.SparseMatrixStochasticMult, Any) at /home/dgleich/.julia/v0.4/MatrixNetworks/src/diffusions.jl:296 is ambiguous with: *(#T
Now that we have Havel Hakimi, it'd be good to have analog for the bipartite graph and directed graph. (https://en.wikipedia.org/wiki/Gale%E2%80%93Ryser_theorem) and https://en.wikipedia.org/wiki/Fulkerson%E2%80%93Chen%E2%80%93Anstee_theorem Although the second seems very close to the...
The output from largest component should also have the indices of the reduced set, and vice versa, if you call "enrich" (e.g. find(p) gives the forward map, and we also...