https://github.com/amlalejini/signalgp-adventures
Capturing the event-driven programming paradigm in GP. Comparing results to traditional linear GP.
https://github.com/amlalejini/signalgp-adventures
Last synced: about 2 months ago
JSON representation
Capturing the event-driven programming paradigm in GP. Comparing results to traditional linear GP.
- Host: GitHub
- URL: https://github.com/amlalejini/signalgp-adventures
- Owner: amlalejini
- License: mit
- Created: 2017-10-02T21:47:50.000Z (about 8 years ago)
- Default Branch: master
- Last Pushed: 2018-05-29T15:22:14.000Z (over 7 years ago)
- Last Synced: 2025-08-23T10:05:11.408Z (about 2 months ago)
- Language: C++
- Size: 2.88 MB
- Stars: 0
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Overview
The signalgp-adventures repository holds a variety of general benchmarks/experiments specific to SignalGP.
## SignalGP
SignalGP is a genetic programming (GP) technique designed to give evolution direct access to the event-driven programming paradigm. [Check of the wiki for more details](https://github.com/amlalejini/signalgp-adventures/wiki).
As I write general SignalGP usage documentation, I'll try to keep this repo's wiki (linked above) up-to-date with the most recent stuff.
## The Pipeline
In general, I'll start exploratory SignalGP work in this repository. As a result, most 'adventures' will likely be unpolished, and potentiallly out of date with the most recent versions of Empirical.
I'll try to update things or make notes when that happens, but no promises.
Once an adventure makes it out of the exploratory stage and it becomes its own application/experiment for a publication, it'll likely get moved to it's own repository where I will document experiment/publication-specific details. Again, I'll try to make note when this happens.
# Adventures
For each documented adventure, see the README.md file in the adventure's directory for more specific details.
## Tests
These adventures are generally used to test a single aspect of SignalGP (typically a new feature while I'm developing it).
### Test: Function Regulation
This test can be found here: [./adventures/test__function_reg](adventures/test__function_reg)
## Experiments
These adventures are experiments. Experiments either aim to explore/demonstrate a particular SignalGP feature or aim to use SignalGP as an evolutionary substrate to explore a question or to solve a problem.
### Experiment: A-B Response
This experiment can be found here: [./adventures/ab_response](adventures/ab_response)
The A-B response problem is designed as a simple toy problem meant to test/demonstrate function regulation in SignalGP.
### Experiment: Environment Coordination
This experiment can be found here: [./adventures/env_coordination](./adventures/env_coordination)
The environment coordination experiment is designed to be a simple toy problem that tests the capacity for SignalGP agents to coordinate with external signals.
### Experiment: Division of Labor
This experiment can be found here: [./adventures/div_of_labor](adventures/div_of_labor)
**WARNING**: this experiment is unfinished/undocumented.
### Experiment: T-Maze
This experiment can be found here: [./adventures/t_maze](adventures/t_maze)
The T-maze problem is meant to test/demonstrate function regulation in SignalGP.