Cartesian Genetic Programming (CGP) implemented in Julia.
Switch branches/tags
Clone or download
Latest commit e74bc16 Mar 27, 2016
Type Name Latest commit message Commit time
Failed to load latest commit information.
examples Complete rename. Mar 27, 2016
src Complete rename. Mar 27, 2016
test Complete rename. Mar 27, 2016
.gitignore Add Vagrantfile and documentation. Mar 12, 2015
.travis.yml Rename package and fix test config. Mar 27, 2016
AUTHORS Add Vagrantfile and documentation. Mar 12, 2015 Complete rename. Mar 27, 2016 Update readme badge. Mar 27, 2016
REQUIRE Fix deprecation warnings and errors on 0.4. May 1, 2015
Vagrantfile Complete rename. Mar 27, 2016 Complete rename. Mar 27, 2016


Build Status

CartesianGP.jl is a library for using Cartesian Genetic Programming in Julia. It is being developed at the University of Montana in Missoula, MT for use in simulating the evolution of technology, though there is nothing specific to that application in the library so it is (will be) perfectly suitable for other applications as well.

Note that the code should be considered pre-alpha at the moment, though we are making rapid progress. We will tag a release when we feel it is useful. Until then, feel free to take a look at the code and send us questions or submit pull requests (though you might want to consult with us first just to make sure you aren't duplicating effort).

If you have questions or comments, please contact one of the authors (see AUTHORS).


The examples/ directory contains (or will soon contain) some examples of using CartesianGP.jl to evolve actual circuits. These can be used as blueprints for implementing your own simulations. Right now the example code is pretty rough, but it will improve over time and will eventually include comments and documentation to help with getting started.



You can run the test suite against the code currenting in the working directory of the repository by running the script. If you want to test the library once you have installed it through Julia's package manager (using Pkg.clone(...) or otherwise) you can use Pkg.test("CartesianGP") from within the Julia REPL. This will not, however, test any modifications you have made to the code that have not been committed, use the script for that.


There is a Vagrant configuration file (called Vagrantfile) in the repository root that will provide two properly configured development and test-running environments (using Virtualbox behind the scenes). One will run the release version of Julia, and the other will run the nightly version. This is especially helpful for Mac and Windows users for whom keeping Julia up-to-date can be a bit of a challenge.

Additionally, this method protects the developer's system Julia packages, which is ideal for people who are both using and developing CartesianGP.jl.

Once Vagrant is installed, bring up the VMs with the following command:

$ vagrant up

Optionally, and this applies to most of the vagrant commands, you can include either "releases" or "nightlies" after the command to apply the action to only one of the machines. So to bring up just the "releases" machine you would do:

$ vagrant up releases

An SSH session can then be opened with vagrant ssh <releases|nightlies>. However, the tests can be run on the VM without doing this. Just use ./ <releases|nightlies|all>. The VM must be running already for this to work.

To shut down the VMs:

$ vagrant halt

To destroy the VMs (and free the hard drive space used to store their disk images:

$ vagrant destroy