Add unbounded objective tests #268
Merged
Conversation
This file contains 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
|
tmigot
requested changes
Apr 8, 2024
src/fomo.jl
Outdated
| set_status!(stats, :unbounded) | ||
| break | ||
| end | ||
| unbounded = isinf(fck) |
There was a problem hiding this comment.
I would go for a test like fx < min(-one(T), f0) / eps(T) instead of testing Inf.
…OSolvers.jl into unbounded-below-test
add objective value test in unbounded tests.
tmigot
approved these changes
Apr 9, 2024
d-monnet
added a commit
that referenced
this pull request
Apr 13, 2024
* update find_beta for the nonmonotone case * nonmonotone extension, remove useless norm computation * Add unbounded optimality in lbfgs * Add unbounded optimality for TRUNK * Add unbounded objective tests (#268) * add unbounded below obj test. Fix unbounded test in fomo. * add unbounded below obj test. Fix unbounded test in fomo. * standardize fomo :unbounded condition, add objective value test in unbounded tests. * rename: fk -> f0 * 🤖 Format .jl files (#270) Co-authored-by: d-monnet <[email protected]> * fomatting Co-authored-by: Tangi Migot <[email protected]> * update find_beta for the nonmonotone case * nonmonotone extension, remove useless norm computation * fomatting Co-authored-by: Tangi Migot <[email protected]> * fix rebase errors * add tests, replace `circshift` by index * fix allocs test * Update src/fomo.jl Co-authored-by: Tangi Migot <[email protected]> * update docstring * update docstring * Update src/fomo.jl Co-authored-by: Tangi Migot <[email protected]> --------- Co-authored-by: tmigot <[email protected]> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: d-monnet <[email protected]>
dpo
reviewed
Apr 13, 2024
| @@ -288,6 +289,9 @@ function SolverCore.solve!( | |||
| solver.α = init_alpha(norm_∇fk, step_backend) | |||
|
|
|||
| # Stopping criterion: | |||
| fmin = min(-one(T), f0) / eps(T) | |||
There was a problem hiding this comment.
If you work in Float64, this isn’t an especially large number in absolute value. Float64 can represent values up to about 1.0e+308. It would be better to use floatmax() here. Effectively, when f0 < -floatmax(Float64), it means that f0 == Inf. So you could declare unboundedness either if f0 == Inf or f0 < floatmax(eltype(x)) / 2 (for example).
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.
Looks like lbfgs doesn't detect unbounded objective.
Trunk subsolver errors (not a descent direction) before unbounded happen (if it happen). Maybe the subsolver crashes because under/overflow due to large value of the gradient.
fomoset status to unbounded whenever objective value overflow (-Inf) whiletronusesfx < min(-one(T), f0) / eps(T)conditions.I would be nice to have the same condition.