Skip to content
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.

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

Added filterslp preset #88

Merged
merged 1 commit into from
Nov 14, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
2 changes: 1 addition & 1 deletion README.md
Original file line number Diff line number Diff line change
Expand Up @@ -47,7 +47,7 @@ Uno implements the following strategies:

**Any strategy combination** can be automatically generated without any programming effort from the user. Note that all combinations do not necessarily result in sensible algorithms, or even convergent approaches. For more details, check out our [preprint](https://www.researchgate.net/publication/381522383_Unifying_nonlinearly_constrained_nonconvex_optimization) or my [presentation at the ICCOPT 2022 conference](https://www.researchgate.net/publication/362254109).

Uno implements three **presets**, that is strategy combinations that correspond to existing solvers (as well as hyperparameter values found in their documentations):
Uno implements **presets**, that is strategy combinations that correspond to existing solvers (as well as hyperparameter values found in their documentations):
* `filtersqp` mimics filterSQP (trust-region feasibility restoration filter SQP method);
* `ipopt` mimics IPOPT (line-search feasibility restoration filter barrier method);
* `byrd` mimics Byrd's S $\ell_1$ QP (line-search $\ell_1$ merit S $\ell_1$ QP method).
Expand Down
23 changes: 23 additions & 0 deletions uno/options/Presets.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@
#include <stdexcept>
#include "Presets.hpp"
#include "Options.hpp"
#include "solvers/LPSolverFactory.hpp"
#include "solvers/QPSolverFactory.hpp"
#include "solvers/SymmetricIndefiniteLinearSolverFactory.hpp"

Expand All @@ -19,13 +20,17 @@ namespace uno {
/** default preset **/
const auto QP_solvers = QPSolverFactory::available_solvers();
const auto linear_solvers = SymmetricIndefiniteLinearSolverFactory::available_solvers();
const auto LP_solvers = LPSolverFactory::available_solvers();

if (not QP_solvers.empty()) {
Presets::set(options, "filtersqp");
}
else if (not linear_solvers.empty()) {
Presets::set(options, "ipopt");
}
else if (not LP_solvers.empty()) {
Presets::set(options, "filterslp");
}
// note: byrd is not very robust and is not considered as a default preset
}
return options;
Expand Down Expand Up @@ -125,6 +130,24 @@ namespace uno {
options["funnel_switching_infeasibility_exponent"] = "2";
options["funnel_update_strategy"] = "2";
}
else if (preset_name == "filterslp") {
options["constraint_relaxation_strategy"] = "feasibility_restoration";
options["subproblem"] = "LP";
options["globalization_mechanism"] = "TR";
options["globalization_strategy"] = "fletcher_filter_method";
options["filter_type"] = "standard";
options["progress_norm"] = "L1";
options["residual_norm"] = "L2";
options["sparse_format"] = "CSC";
options["TR_radius"] = "10";
options["l1_constraint_violation_coefficient"] = "1.";
options["enforce_linear_constraints"] = "yes";
options["tolerance"] = "1e-5";
options["loose_tolerance"] = "1e-4";
options["TR_min_radius"] = "1e-8";
options["switch_to_optimality_requires_linearized_feasibility"] = "yes";
options["protect_actual_reduction_against_roundoff"] = "no";
}
else {
throw std::runtime_error("The preset " + preset_name + " is not known");
}
Expand Down