Skip to content

Correct xi in SolverCore.solve! in LM and R2N solvers #239

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
MohamedLaghdafHABIBOULLAH wants to merge 5 commits into from
Closed

Correct xi in SolverCore.solve! in LM and R2N solvers #239

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
afdd274
Correct xi in SolverCore.solve! in LM and R2N solvers
MohamedLaghdafHABIBOULLAH Sep 30, 2025
cc03f69
Fix regularization parameter reference in SolverCore.solve! for LM an…
MohamedLaghdafHABIBOULLAH Sep 30, 2025
36299df
Fix regularization parameter initialization in LMSolver and R2NSolver
MohamedLaghdafHABIBOULLAH Sep 30, 2025
449b930
Remove redundant assignment of regularization parameter in SolverCore…
MohamedLaghdafHABIBOULLAH Sep 30, 2025
7b4748c
Set regularization parameter σ in SolverCore.solve! for LMSolver and …
MohamedLaghdafHABIBOULLAH Sep 30, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions .gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
Manifest.toml
6 changes: 3 additions & 3 deletions src/LM_alg.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,7 +66,7 @@ function LMSolver(
shifted(reg_nls.h, xk)

Jk = jac_op_residual(reg_nls.model, xk)
sub_nlp = LMModel(Jk, Fk, T(1), xk)
sub_nlp = LMModel(Jk, Fk, T(0), xk)
subpb = RegularizedNLPModel(sub_nlp, ψ)
substats = RegularizedExecutionStats(subpb)
subsolver = subsolver(subpb)
Expand Down Expand Up @@ -300,7 +300,7 @@ function SolverCore.solve!(
end

mk = let ψ = ψ, solver = solver
d -> obj(solver.subpb.model, d) + ψ(d)
d -> obj(solver.subpb.model, d) - solver.subpb.model.σ * dot(d, d) / 2 + ψ(d)
end

prox!(s, ψ, mν∇fk, ν)
Expand Down Expand Up @@ -331,7 +331,7 @@ function SolverCore.solve!(

while !done
sub_atol = stats.iter == 0 ? 1.0e-3 : min(sqrt_ξ1_νInv ^ (1.5), sqrt_ξ1_νInv * 1e-3)
solver.subpb.model.σ = σk

isa(solver.subsolver, R2DHSolver) && (solver.subsolver.D.d[1] = 1/ν)
if isa(solver.subsolver, R2Solver) #FIXME
solve!(solver.subsolver, solver.subpb, solver.substats, x = s, atol = sub_atol, ν = ν)
Expand Down
5 changes: 2 additions & 3 deletions src/R2N.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,7 @@ function R2NSolver(
shifted(reg_nlp.h, xk)

Bk = hess_op(reg_nlp.model, x0)
sub_nlp = R2NModel(Bk, ∇fk, T(1), x0)
sub_nlp = R2NModel(Bk, ∇fk, T(0), x0)
subpb = RegularizedNLPModel(sub_nlp, ψ)
substats = RegularizedExecutionStats(subpb)
subsolver = subsolver(subpb)
Expand Down Expand Up @@ -327,7 +327,7 @@ function SolverCore.solve!(
end

mk = let ψ = ψ, solver = solver
d -> obj(solver.subpb.model, d) + ψ(d)::T
d -> obj(solver.subpb.model, d) - solver.subpb.model.σ * dot(d, d) / 2 + ψ(d)::T
end

prox!(s1, ψ, mν∇fk, ν₁)
Expand Down Expand Up @@ -361,7 +361,6 @@ function SolverCore.solve!(
while !done
sub_atol = stats.iter == 0 ? 1.0e-3 : min(sqrt_ξ1_νInv ^ (1.5), sqrt_ξ1_νInv * 1e-3)

solver.subpb.model.σ = σk
Copy link
Collaborator Author

@MohamedLaghdafHABIBOULLAH MohamedLaghdafHABIBOULLAH Sep 30, 2025

Choose a reason for hiding this comment

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

@dpo is this ok for you?

Copy link
Collaborator

@MaxenceGollier MaxenceGollier Sep 30, 2025

Choose a reason for hiding this comment

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

Why should we remove this ?
solver.subpb.model.σwill not be updated anywhere ?

Copy link
Member

@dpo dpo Oct 1, 2025

Choose a reason for hiding this comment

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

I think that line needs to stay. It's the computation of $\xi_k$ that is inefficient. Please grab the model value from the stats and subtract the $\tfrac{1}{2} \sigma |s|^2$ to avoid recomputing it.

Copy link
Collaborator Author

@MohamedLaghdafHABIBOULLAH MohamedLaghdafHABIBOULLAH Oct 2, 2025

Choose a reason for hiding this comment

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

We should close this as this issue has been treated here PR by Maxence.

isa(solver.subsolver, R2DHSolver) && (solver.subsolver.D.d[1] = 1/ν₁)
if isa(solver.subsolver, R2Solver) #FIXME
solve!(
Expand Down
Loading