CHomP.jl

This package is a julia wrapper around the Computational Homology Project (CHomP) with python bindings. The wrapper is very barebones, really it just provides a way to call

using CHomP

and have pychomp2 be available for use (via PyCall.jl).

If you are a fellow mathematician or computer scientist and want to create a true julia implementation of CHomP, please contact April Herwig (via April-Hannah-Lena on GitHub or april.herwig@tum.de), I will happily transfer ownership rights of this package to you.

CHomP.jl requires julia 1.7 or later. To install CHomP.jl, call

] add https://github.com/April-Hannah-Lena/CHomP.jl.git

Upon first install, this will run a build process, which downloads and installs the python pakage pychomp2. After any updates to python, PyCall.jl, Conda.jl, or CHomP.jl, it is recommended (though probably not necessary) to rebuild:

build CHomP

One can now use CHomP.jl through PyCall.jl, which is automatically exported by CHomP.jl

using CHomP

(you do not need to call using PyCall)

The python module chompjl is then made available to the namespace (via pyimport).

julia> chompjl
PyObject <module 'pychomp2' from '/home/user/.local/lib/python3.8/site-packages/pychomp2/__init__.py'>

julia> betti_nrs = chompjl.CubicalHomology([
            # cubical set representing a 2-dimensional square with a hole punched in the middle
            [0, 0],
            [0, 1],
            [0, 2],
            [1, 0],
            [1, 2],
            [2, 0],
            [2, 1],
            [2, 2]
        ])

3-element Vector{Int64}:
 1
 1
 0

Ensure that python and pip are accessible in your systems PATH environment variable.

The python instance created by PyCall.jl should include the packages sys and subprocess. These are all the python packages required to build CHomP.jl. If pip is available in your PATH, then a failed build step is most likely an issue with PyCall.jl. See for reference PyCall.jl's torubleshooting section or the PyCalll.jl issues page.

If a problem occurs that seems to bee related to julia, open an issue in the package repository on GitHub.

If a problem occurs that seems to be related to chomp (resp. pychomp2 or the original C++ chomp), open an issue in the chomp repository on Github (resp. pychomp2 or chomp).