Implement CMA-MEGA - Githubissues

icaros-usc / pyribs

A bare-bones Python library for quality diversity optimization.

https://pyribs.org

MIT License

208 stars 33 forks source link

Implement CMA-MEGA #240

Closed itsdawei closed 2 years ago

itsdawei commented 2 years ago

Description

This PR implements CMA-MEGA algorithm presented in this paper.

TODO

[x] Implement CMA-MEGA
[x] Add CMA-MEGA benchmark
[x] Add CMA-MEGA tests
[x] Run sphere.py with CMA-MEGA
[x] Add comments
[x] Create DQDEmitterBase
[x] Add test for call order check

Question

Behavior when trying to call ask_dqd on non-DQD emitter. (a) skip the call in scheduler.

Status

[x] I have read the guidelines in CONTRIBUTING.md
[x] I have formatted my code using yapf
[x] I have tested my code by running pytest
[x] I have linted my code with pylint
[x] I have added a one-line description of my change to the changelog in HISTORY.md
[x] This PR is ready to go

itsdawei commented 2 years ago

A comparison between our results and the results from the original paper. Note that the LP (sphere) results are obtained with dimensions=1000, iterations=10000, and archive_dim=(100,100).

Original Paper

Our Results for LP (sphere)

CMA-MEGA: Mean QD Score: 75.292, Coverage: 100%, Time: 30s

Hence, we conclude that our implementation of CMA-MEGA is consistent with the origin paper.