See full documentation here, including tutorials and examples.

CoexistenceHoles is a julia and R package that originally was made for the project "Coexistence holes characterize the assembly and disassembly of multispecies sytems". However, the package has the potential for a variety of applications. In short, it provides efficient tools for analyzing the homology of general hypergraphs.

This package is not registered (yet). You can install it via the Julia REPL like this:

julia> using Pkg
julia> Pkg.add(PackageSpec(url="https://github.com/SyntheticDynamics/CoexistenceHoles.jl.git", rev="master"))

Or you can install it via the Pkg REPL like this:

(v1.3) pkg> add https://github.com/SyntheticDynamics/CoexistenceHoles.jl.git#master

If you already have R installed then you'll need to download install julia. You can check if julia is installed correctly by running the julia command in a terminal. If this command is not found, you will need to add it to your path following the proper instructions for your operating system.

In R use JuliaCall is used to interface between languages. For function summaries see this document. However studying these functions is not necessary since CoexistenceHoles's shows the proper functions to use from JuliaCall in the tutorial and examples.

The follwoing are steps to install CoexistenceHoles in R. See the examples or tutorials for more specific instructions.

install.packages("JuliaCall")

library(JuliaCall)
julia <- julia_setup()

# only need to run this once
julia_install_package("https://github.com/SyntheticDynamics/CoexistenceHoles.jl.git#master")

# add the library every time you open a new session of R and want to use CoexistenceHoles
julia_library("CoexistenceHoles")

using CoexistenceHoles






N = 8 # number of species in our ecosystem

# create a random community matrix
σA = 0.1 # standard deviation for entries
C = 0.1 # success rate of Bernoulli distribution used to populate matrix
A = random_communitymatrix(N, σA, C)

# create a random growth vector
μ = 0.3 # mean of LogNormal distribution used to generate each value
σr = 0.2 # standard deviation of LogNormal distribution used to generate each value
r = random_growthvector(N, μ, σr)

# create assembly and disassembly hypergraph
reg = 0
H = assembly_hypergraph_GLV(A, r; method = "localstability", regularization = reg)
R = disassembly_hypergraph(H)

# save these for later if you want
save_hypergraph_dat("~/hypergraphs/assembly_hypergraph.dat", H)
save_hypergraph_dat("~/hypergraphs/disassembly_hypergraph.dat", R)

# get the betti numbers
betti_H = betti_hypergraph_ripscomplex(H; max_dim = max_dim)

julia_library("CoexistenceHoles")

opt <- julia_pkg_import("CoexistenceHoles", func_list = c("random_communitymatrix",
                                                           "random_growthvector",
                                                           "assembly_hypergraph_GLV",
                                                           "dissassembly_hypergraph",
                                                           "save_hypergraph_dat"))
N = 8 # number of species in our ecosystem

# create a random community matrix
sA = 0.1 # standard deviation for community matrix
C = 0.1 # success rate of Bernoulli distribution used to populate matrix
A = opt$random_communitymatrix(N, sA, C)

# create a random growth vector
mr = 0.1
sr = 0.1
r = opt$random_growthvector(N, mr, sr)

# create assembly and disassembly hypergraph
reg = 0
max_dim  = 4
H = opt$assembly_hypergraph_GLV(A,R; method="localstability", regularization=reg)
M = opt$disassembly_hypergraph(H)

# save these for later if you want
save_hypergraph_dat("~/hypergraphs/assembly_hypergraph.dat", H)
save_hypergraph_dat("~/hypergraphs/disassembly_hypergraph.dat", R)

# get the betti numbers
betti_H = betti_hypergraph_ripscomplex(H; max_dim = max_dim)

If you use CoexistenceHoles for academic research, please cite the following paper:

Angulo, M.T., Kelley, A., Montejano, L. et al. Coexistence holes characterize the assembly and disassembly of multispecies systems. Nat Ecol Evol 5, 1091–1101 (2021). https://doi.org/10.1038/s41559-021-01462-8

• Marco Tulio
• Aaron Kelley