Add osy test problem #679
Add osy test problem #679
Conversation
facebook-github-bot
added
the
CLA Signed
Codecov Report
@@ Coverage Diff @@
## master #679 +/- ##
=========================================
Coverage 100.00% 100.00%
=========================================
Files 98 98
Lines 6596 6614 +18
=========================================
+ Hits 6596 6614 +18
Continue to review full report at Codecov.
|
| dim: int = 6, | ||
| num_objectives: int = 2, |
There was a problem hiding this comment.
looks like this is always in dim 6 with 2 objectives? Let's remove these args then (these are for test functions where one can select the dimensionality).
There was a problem hiding this comment.
In fact, you can just omit the constructor altogether then since the superclass constructor will do all you need
There was a problem hiding this comment.
I originally built the constructor to have negate=True as the default option, but that was breaking the unit tests.
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@Balandat has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
|
lgtm. Let's add this to botorch/test_functions/init.py |
|
@wjmaddox has updated the pull request. You must reimport the pull request before landing. |
|
@wjmaddox has updated the pull request. You must reimport the pull request before landing. |
facebook-github-bot
left a comment
facebook-github-bot
left a comment
There was a problem hiding this comment.
@Balandat has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
Summary: This PR adds support for input and outcome transforms for MTGPs as currently implemented in BoTorch. ## Motivation Helps improve bayesian optimization loops using MTGPs including contextual ones by allowing these types of transforms. ### Have you read the [Contributing Guidelines on pull requests](https://github.com/pytorch/botorch/blob/master/CONTRIBUTING.md#pull-requests)? yes. Pull Request resolved: meta-pytorch#681 Test Plan: Current unit tests; need to dive into these. ## Related PRs See meta-pytorch#679 as well. Reviewed By: Balandat Differential Revision: D26424473 Pulled By: wjmaddox fbshipit-source-id: e7b420879c964b268d6bbf32faad19f3c1fbcc45
Summary: This PR adds support for input and outcome transforms for MTGPs as currently implemented in BoTorch. ## Motivation Helps improve bayesian optimization loops using MTGPs including contextual ones by allowing these types of transforms. ### Have you read the [Contributing Guidelines on pull requests](https://github.com/pytorch/botorch/blob/master/CONTRIBUTING.md#pull-requests)? yes. Pull Request resolved: #681 Test Plan: Current unit tests; need to dive into these. ## Related PRs See #679 as well. Reviewed By: Balandat Differential Revision: D26424473 Pulled By: wjmaddox fbshipit-source-id: 07d9f2bcc471ce24db7b24caa0893da3880fb625
4 participants
This PR adds the OSY multi-objective test problem (two objectives, six constraints, six dimensions).
Motivation
It's a higher dimensional output multi-objective test problem that is useful for testing out MTGPs for constrained multi-objective bayes opt.
Have you read the Contributing Guidelines on pull requests?
yes.
Test Plan
unit tests. osy is added.
Related PRs
N/A. Pulled out of https://github.com/wjmaddox/botorch.