Add grey box objectives to Pyomo.DoE #3606
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Issue on windows, uploading file to test on mac
Windows-specific issue with the code; will report issue in due time.
Used all phrases for ExternalGreyBoxModel. Need to make more robust the naming and bookkeeping for input names and subsequently the coo matrices.
Added a single version of grey box. Currently cannot solve due to the model not being a Pyomo Block. Will revisit next week.
Added greybox solver separate from regular solver (will call normal ipopt to solve subproblems without grey box model). Fixed bugs with grey box utilities to now update the input values correctly. Having determinism issue.
Trying to explore hessian option to make these things converge better. Need to explore more tensor calculus.
Need to try some other numerical things to encourage convergence before attacking Hessian problem...
Added option for condition number optimality. Last step will be to potentially change to triangular definition of inputs in grey-box function, then add Hessian computation.
Should improve computational performance. Seemed to help E and ME but D still has issues. Next steps would be to add the Hessian calculation.
Trying to get D-optimality to work without hessian first.
Adding functions for testing grey box derivative accuracy using numerical counterparts. Functions are working properly, just need to implement the derivatives/hessians in the grey box file.
This functionality is no longer planning to be supported. Therefore, we have removed this `TODO` comment.
|
@blnicho @mrmundt @michaelbynum @jsiirola I think this is ready for another review |
mrmundt
requested changes
Jun 26, 2025
|
|
||
| if self.objective_option == ObjectiveLib.trace: | ||
| Minv = np.linalg.pinv(M) | ||
| # Derivative formula of A-optimality |
There was a problem hiding this comment.
I really appreciate all of these inline comments - very helpful.
sscini
moved this from Testing
to Ready for final review
in ParmEst & Pyomo.DoE Development
mrmundt
reviewed
Jun 30, 2025
| @@ -826,7 +928,9 @@ def initialize_jac(m, i, j): | |||
| # Otherwise initialize to 0.1 (which is an arbitrary non-zero value) | |||
| else: | |||
| raise AttributeError( | |||
There was a problem hiding this comment.
You may want to have this be a DeveloperError or MouseTrap error instead of an AttributeError.
| @@ -1062,13 +1171,10 @@ def _generate_scenario_blocks(self, model=None): | |||
| model.scenarios = range(len(model.base_model.unknown_parameters) + 1) | |||
| else: | |||
| raise AttributeError( | |||
There was a problem hiding this comment.
You may want to have this be a DeveloperError or MouseTrap error instead of an AttributeError.
| @@ -1183,12 +1291,14 @@ def create_objective_function(self, model=None): | |||
| ObjectiveLib.zero, | |||
| ]: | |||
| raise AttributeError( | |||
There was a problem hiding this comment.
You may want to have this be a DeveloperError or MouseTrap error instead of an AttributeError.
|
|
||
| # Test whether or not cyipopt | ||
| # is appropriately calling the | ||
| # lienar solvers. |
There was a problem hiding this comment.
Ooo, typo. Reported upstream to crate-ci/typos
Suggested change
| # lienar solvers. | |
| # linear solvers. |
Comment on lines
+345
to
+347
| # Change linear solvers to mumps | ||
| # doe_object.solver.options["linear_solver"] = "mumps" | ||
| # doe_object.grey_box_solver.config.options["linear_solver"] = "mumps" |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary/Motivation:
Pyomo.DoE currently can only compute trace-optimal and D-optimal experimental designs. We are expanding the suite to be able to compute E-optimal and ME-optimal (or condition number optimal) designs. Given that these formulations require computing eigenvalues on the fly, we have updated Pyomo.DoE to make use of the
ExternalGreyBoxmodules.This PR adds grey box capabilities to Pyomo.DoE.
Changes proposed in this PR:
Legal Acknowledgement
By contributing to this software project, I have read the contribution guide and agree to the following terms and conditions for my contribution: