Skip to content

Add RVEAa and final_step, and modify DTLZ, IGD and RVEA#231

Merged
BillHuang2001 merged 5 commits into
mainfrom
evoxtorch-dev-xbq
Apr 5, 2025
Merged

Add RVEAa and final_step, and modify DTLZ, IGD and RVEA#231
BillHuang2001 merged 5 commits into
mainfrom
evoxtorch-dev-xbq

Conversation

@XU-Boqing
Copy link
Copy Markdown
Collaborator

Add RVEAa

An implementation of the Reference Vector Guided Evolutionary Algorithm embedded with the reference vector regeneration strategy (RVEAa) for multi-objective optimization problems.

  • src/evox/algorithms/mo/rveaa.py
  • src/evox/algorithms/__ init __.py
  • src/evox/algorithms/mo/__ init __.py

Add final_step

Add the final_step() to Algorithm.

  • src/evox/core/components.py
  • src/evox/workflows/eval_monitor.py
  • src/evox/workflows/std_workflow.py

Modify DTLZ

Avoid changing the original population when evaluating.

  • src/evox/problems/numerical/dtlz.py
  • unit_test/problems/test_dtlz.py

Modify IGD

Make igd able to deal with populations with nan.

  • src/evox/metrics/igd.py

Modify RVEA

Switch .type(torch.int) to .to(dtype=torch.int)

  • src/evox/algorithms/mo/rvea.py

Others

Add import of basic functions, which will be used in LSMOP in Evomo.

  • src/evox/problems/numerical/init.py

Modify non_dominate_rank

  • src/evox/operators/selection/non_dominate.py

Copy link
Copy Markdown
Collaborator

@BillHuang2001 BillHuang2001 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please squash the unneeded commits, and then add RVEAa to test_moea.py.

Comment thread src/evox/core/components.py Outdated
Comment thread src/evox/metrics/igd.py Outdated
Comment thread src/evox/operators/selection/non_dominate.py Outdated
Comment thread src/evox/operators/selection/non_dominate.py Outdated
Comment thread src/evox/algorithms/mo/rvea.py Outdated
Comment thread src/evox/algorithms/mo/rveaa.py
Modify dtlz and add rveaa, zdt, maf

Fix cso algorithm

Manually import submodules

Fix std_workflow._evaluate cannot be captured by torch.compile bug

Remove unnecessary torch.jit.ignore

When recording batchedtensors, unwrap it to avoid leaking batchedtensors

Fix to_jax_array

Use normal assignment instead of inplace copy

Test vmapped workflow with monitor

Add default feature set

Update installation instruction on readme

Update install guide

Bump version to v1.1.1

add num_repeats and multi obj to hpo

Add RVEAa, modify RVEA, DTLZ and hpo_wrapper.
@XU-Boqing XU-Boqing force-pushed the evoxtorch-dev-xbq branch from 2ae5f25 to 1540708 Compare April 3, 2025 10:25
@XU-Boqing XU-Boqing requested a review from BillHuang2001 April 3, 2025 12:02
@BillHuang2001 BillHuang2001 merged commit eda1dba into main Apr 5, 2025
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants