Skip to content

moved Proxstorm and TypeP rework PR from old repo#9

Open
rjbaraldi wants to merge 15 commits into
developfrom
proxstorm_typep
Open

moved Proxstorm and TypeP rework PR from old repo#9
rjbaraldi wants to merge 15 commits into
developfrom
proxstorm_typep

Conversation

@rjbaraldi
Copy link
Copy Markdown
Collaborator

Implementation of proxstorm in typeP (need to move to SOL) and rework of trust-region algorithms to readable subproblem solvers, similar to typeu.

@rjbaraldi rjbaraldi requested a review from aj463 November 18, 2025 04:10
@rjbaraldi rjbaraldi changed the title moved PR from old repo over moved Proxstorm and TypeP rework PR from old repo Nov 18, 2025
@dridzal
Copy link
Copy Markdown
Collaborator

dridzal commented Dec 8, 2025

@rjbaraldi , @aj463 : Is there a time constraint with this PR, or is it ok that it has not been reviewed in three weeks?

@rjbaraldi
Copy link
Copy Markdown
Collaborator Author

@dridzal - Brian was working on moving ProxStorm to sol in this branch + writing some tests for it. If you want, we can merge this and he can make another. Let me know.

@dridzal
Copy link
Copy Markdown
Collaborator

dridzal commented Dec 8, 2025

@rjbaraldi , there is no rush; I was just making sure that the PR is current. Thanks!

…r algorithms in SOL. Moved dualsvm and logistic examples from example/proxstorm to example/ and use the moved proxstorm.

Added test/sol/test_17.cpp, based on test/algorithm/TypeP/test_08.cpp and adjusted to use proxstorm (with no stochasticity)

Todo: Add stochasticity.
@rjbaraldi rjbaraldi requested a review from dpkouri March 6, 2026 00:18
@rjbaraldi
Copy link
Copy Markdown
Collaborator Author

@bchen0 - looks like some prox-storm test failed. Maybe we can remove that for now?

bchen0 and others added 8 commits March 6, 2026 17:42
Signed-off-by: bchen0 <bchen121@gmail.com>
2) Fixed bug in Monte Carlo error calculation
3) Updated RiskNeutralObjective value and gradient to work with Monte Carlo samples
1) A MonteCarloDataGenerator object, that is based off MonteCarloGenerator but takes samples from a text file instead of a distribution.
2) Add in logistic regression example that uses the stochastic proxstorm and the above MonteCarloDataGenerator
…2) Change nullptr to ROL::nullPtr to play nice with Teuchos
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