Skip to content

A Genetic Programming platform for Python with TensorFlow for wicked-fast CPU and GPU support.

License

Notifications You must be signed in to change notification settings

kstaats/karoo_gp

Repository files navigation

Karoo GP

Karoo GP is an evolutionary algorithm, a genetic programming application suite written in Python which supports both symbolic regression and classification data analysis. It has been used in radio astronomy, gravitational wave detector characterisation and synthetic supernovae detection, and a variety of other use cases in a diversity of fields.

You need only prepare your dataset according to the User Guide. No programming required. Karoo is multicore and GPU enabled by means of the powerful library TensorFlow. Karoo has three text cases built-in: Iris dataset, Kepler's law of planetary motion, and a maths problem you can modify to various degrees of challenge.

Karoo is launched from the command line with an intuitive user interface or with arguments for full automation from bash or another Python script. The output of each run is automatically archived and includes the configuraiton, a summary, and the full suite of GP trees saved as .csv files for your review and edit such that you can hand-build the starting block for your next run.

Be certain to read the User Guide for a starter's guide to Genetic Programming and examples of all you can do with this unique body of code.

For an interesting read on scalar vs vector and CPU vs GPU performance with Karoo GP: https://arxiv.org/abs/1708.03157 or to learn how Karoo applied to supernova detection at LIGO: https://arxiv.org/abs/2002.04591

Learn more at kstaats.github.io/karoo_gp/ ...

About

A Genetic Programming platform for Python with TensorFlow for wicked-fast CPU and GPU support.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages