There is a newer version of the record available.

Published March 15, 2019 | Version v2.3
Software Open

COmparing Continuous Optimizers: numbbo/COCO on Github

Creators

  • 1. Inria
  • 2. Jožef Stefan Institute
  • 3. Veolia Research and Innovation
  • 4. Inria & Thales
  • 5. Hanoi National University of Education

Description

Numerical black-box optimization benchmarking framework.

New Features in v2.3:

  • support of discrete variables
  • new bbob-largescale test suite with 24 functions in dimensions 20 to 640
  • new bbob-mixint test suite, a discretized version of the bbob test suite
  • new bbob-biobj-mixint test suite, a discretized version of the bbob-biobj-ext test suite
  • the order of input arguments to cocopp.main is preserved as the order in which algorithms are displayed and in particular the order in which the colors are chosen.
  • improved COCODataArchive
  • revised example_experiment2.py

Solved Bugs

  • bug in SMS-EMOA example (issue #1853)

Others

  • adjustments to support matplotlib 3.0.0

Files

numbbo/coco-v2.3.zip

Files (14.6 MB)

Name Size Download all
md5:bdf6d087daa679d48343f21b51cb917e
14.6 MB Preview Download

Additional details

Related works

Is supplement to
https://github.com/numbbo/coco/tree/v2.3 (URL)