solve.h

Go to the documentation of this file.

// Copyright 2010-2021 Google LLC
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
 
// Functions and classes used to solve a Model.
//
// The main entry point is the Solve() function.
//
// For users that need incremental solving, there is the IncrementalSolver
// class.
#ifndef OR_TOOLS_MATH_OPT_CPP_SOLVE_H_
#define OR_TOOLS_MATH_OPT_CPP_SOLVE_H_
 
#include <functional>
#include <iostream>
#include <memory>
#include <optional>
#include <string>
#include <utility>
#include <vector>
 
#include "absl/status/statusor.h"
#include "absl/strings/string_view.h"
#include "ortools/base/source_location.h"
#include "ortools/math_opt/core/model_storage.h"
#include "ortools/math_opt/core/non_streamable_solver_init_arguments.h"  // IWYU pragma: export
#include "ortools/math_opt/core/solve_interrupter.h"  // IWYU pragma: export
#include "ortools/math_opt/core/solver.h"
#include "ortools/math_opt/cpp/callback.h"  // IWYU pragma: export
#include "ortools/math_opt/cpp/model.h"
#include "ortools/math_opt/cpp/model_solve_parameters.h"  // IWYU pragma: export
#include "ortools/math_opt/cpp/parameters.h"  // IWYU pragma: export
#include "ortools/math_opt/cpp/solve_result.h"  // IWYU pragma: export
#include "ortools/math_opt/cpp/streamable_solver_init_arguments.h"  // IWYU pragma: export
#include "ortools/math_opt/parameters.pb.h"  // IWYU pragma: export
 
namespace operations_research {
namespace math_opt {
 
// Callback function for messages callback sent by the solver.
//
// Each message represents a single output line from the solver, and each
// message does not contain any '\n' character in it.
//
// Thread-safety: a callback may be called concurrently from multiple
// threads. The users is expected to use proper synchronization primitives to
// deal with that.
using MessageCallback = std::function<void(const std::vector<std::string>&)>;
 
// Returns a message callback function that prints its output to the given
// output stream, prefixing each line with the given prefix.
//
// For each call to the returned message callback, the output_stream is flushed.
//
// Usage:
//
//   SolveArguments args;
//   args.message_callback = PrinterMessageCallback(std::cerr, "solver logs> ");
MessageCallback PrinterMessageCallback(std::ostream& output_stream = std::cout,
                                       absl::string_view prefix = "");
 
 
// Arguments passed to Solve() and IncrementalSolver::New() to control the
// instantiation of the solver.
//
// For convenience, constructors with streamable or/and non-streamable arguments
// are provided. The non-streamable arguments are cloned so any change made
// after passing them to this class are ignored.
//
// Usage with streamable arguments:
//
//   Solve(model, SOLVER_TYPE_GUROBI, /*solver_args=*/{},
//         SolverInitArguments({
//           .gurobi = StreamableGurobiInitArguments{
//             .isv_key = GurobiISVKey{
//               .name = "some name",
//               .application_name = "some app name",
//               .expiration = -1,
//               .key = "random",
//             }
//           }
//         });
//
// Usage with non-streamable arguments:
//
//   NonStreamableGurobiInitArguments gurobi_args;
//   gurobi_args.master_env = master_env.get();
//
//   Solve(model, SOLVER_TYPE_GUROBI, /*solver_args=*/{},
//         SolverInitArguments(gurobi_args));
//
struct SolverInitArguments {
  SolverInitArguments() = default;
 
  // Initializes this class with a copy of the provided streamable arguments.
  explicit SolverInitArguments(StreamableSolverInitArguments streamable);
 
  // Initializes this class with a clone of the provided non-streamable
  // arguments.
  //
  // Note that since this constructors calls Clone() to initialize the
  // non_streamable_solver_init_arguments field, changes made after calling it
  // to the input non_streamable are ignored.
  explicit SolverInitArguments(
      const NonStreamableSolverInitArguments& non_streamable);
 
  // Initializes this class with both the provided a copy streamable arguments
  // and a clone of the non-streamable ones.
  SolverInitArguments(StreamableSolverInitArguments streamable,
                      const NonStreamableSolverInitArguments& non_streamable);
 
  // Initializes this class as a copy of the provided arguments. The
  // non_streamable field is cloned if not nullptr.
  SolverInitArguments(const SolverInitArguments& other);
 
  // Sets this class as a copy of the provided arguments. The non_streamable
  // field is cloned if not nullptr.
  SolverInitArguments& operator=(const SolverInitArguments& other);
 
  SolverInitArguments(SolverInitArguments&&) = default;
  SolverInitArguments& operator=(SolverInitArguments&&) = default;
 
  StreamableSolverInitArguments streamable;
 
  // This should either be the solver specific class or nullptr.
  //
  // Solvers will fail (by returning an absl::Status) if called with arguments
  // for another solver.
  std::unique_ptr<const NonStreamableSolverInitArguments> non_streamable;
};
 
// Arguments passed to Solve() and IncrementalSolver::Solve() to control the
// solve.
struct SolveArguments {
  // Model independent parameters, e.g. time limit.
  SolveParameters parameters;
 
  // Model dependent parameters, e.g. solution hint.
  ModelSolveParameters model_parameters;
 
  // An optional callback for messages emitted by the solver.
  //
  // When set it enables the solver messages and ignores the `enable_output` in
  // solve parameters; messages are redirected to the callback and not printed
  // on stdout/stderr/logs anymore.
  //
  // See PrinterMessageCallback() for logging to stdout/stderr.
  //
  // Usage:
  //
  //   // To print messages to stdout with a prefix.
  //   ASSIGN_OR_RETURN(
  //     const SolveResult result,
  //     Solve(model, SOLVER_TYPE_GLOP,
  //           { .message_callback = PrinterMessageCallback(std::cout,
  //                                                        "logs| "); });
  //
  //   // To print messages to the INFO log.
  //   ASSIGN_OR_RETURN(
  //     const SolveResult result,
  //     Solve(model, SOLVER_TYPE_GLOP,
  //           { .message_callback = InfoLoggerMessageCallback("[solver] "); });
  //
  //   // To print messages to the VLOG(1) log.
  //   ASSIGN_OR_RETURN(
  //     const SolveResult result,
  //     Solve(model, SOLVER_TYPE_GLOP,
  //           { .message_callback = VLoggerMessageCallback(1, "[solver] "); });
  //
  MessageCallback message_callback = nullptr;
 
  // Callback registration parameters. Usually `callback` should also be set
  // when these parameters are modified.
  CallbackRegistration callback_registration;
 
  // The callback. The `callback_registration` parameters have to be set, in
  // particular `callback_registration.events`.
  Callback callback = nullptr;
 
  // An optional interrupter that the solver can use to interrupt the solve
  // early.
  //
  // Usage:
  //   auto interrupter = std::make_shared<SolveInterrupter>();
  //
  //   // Use another thread to trigger the interrupter.
  //   RunInOtherThread([interrupter](){
  //     ... wait for something that should interrupt the solve ...
  //     interrupter->Interrupt();
  //   });
  //
  //   ASSIGN_OR_RETURN(const SolveResult result,
  //                    Solve(model, SOLVER_TYPE_GLOP,
  //                          { .interrupter = interrupter.get() });
  //
  SolveInterrupter* interrupter = nullptr;
};
 
// Solves the input model.
//
// A Status error will be returned if there is an unexpected failure in an
// underlying solver or for some internal math_opt errors. Otherwise, check
// SolveResult::termination.reason to see if an optimal solution was found.
//
// Memory model: the returned SolveResult owns its own memory (for solutions,
// solve stats, etc.), EXPECT for a pointer back to the model. As a result:
//  * Keep the model alive to access SolveResult,
//  * Avoid unnecessarily copying SolveResult,
//  * The result is generally accessible after mutating the model, but some care
//    is needed if variables or linear constraints are added or deleted.
//
// Asserts (using CHECK) that the inputs solve_args.model_parameters and
// solve_args.callback_registration only contain variables and constraints from
// the input model.
//
// See callback.h for documentation on solve_args.callback and
// solve_args.callback_registration.
//
// Thread-safety: this method is safe to call concurrently on the same Model.
//
// Some solvers may add more restrictions regarding threading. Please see
// SolverType::kXxx documentation for details.
absl::StatusOr<SolveResult> Solve(const Model& model, SolverType solver_type,
                                  const SolveArguments& solve_args = {},
                                  const SolverInitArguments& init_args = {});
 
// Incremental solve of a model.
//
// This is a feature for advance users. Most users should only use the Solve()
// function above.
//
// Here incremental means that the we try to reuse the existing underlying
// solver internals between each solve. There is no guarantee though that the
// solver supports all possible model changes. Hence there is not guarantee that
// performances will be improved when using this class; this is solver
// dependent. Typically LPs have more to gain from incremental solve than
// MIPs. In both cases, even if the solver supports the model changes,
// incremental solve may actually be slower.
//
// The New() function instantiates the solver and setup it from the current
// state of the Model. Calling Solve() will update the underlying solver with
// latest model changes and solve this model.
//
// Usage:
//   Model model = ...;
//   ASSIGN_OR_RETURN(
//     const std::unique_ptr<IncrementalSolver> incremental_solve,
//     IncrementalSolver::New(model, SOLVER_TYPE_XXX));
//
//   ASSIGN_OR_RETURN(const SolveResult result1, incremental_solve->Solve());
//
//   model.AddVariable(...);
//   ...
//
//   ASSIGN_OR_RETURN(const SolveResult result2, incremental_solve->Solve());
//
//   ...
//
// Thread-safety: The New(), Solve() and Update() methods must not be called
// while modifying the Model() (adding variables...). The user is expected to
// use proper synchronization primitives to serialize changes to the model and
// the use of this object. Note though that it is safe to call methods from
// different IncrementalSolver instances on the same Model concurrently.
//
// There is no problem calling SolveWithoutUpdate() concurrently on different
// instances of IncrementalSolver or while the model is being modified (unless
// of course the underlying solver itself is not thread-safe and can only be
// called from a single-thread).
//
// Note that Solve(), Update() and SolveWithoutUpdate() are not reentrant so
// they should not be called concurrently on the same instance of
// IncrementalSolver.
//
// Some solvers may add more restrictions regarding threading. Please see
// SolverType::kXxx documentation for details.
class IncrementalSolver {
 public:
  struct UpdateResult {
    UpdateResult(const bool did_update, std::optional<ModelUpdateProto> update)
        : did_update(did_update), update(std::move(update)) {}
 
    // True if the solver has been successfully updated or if no update was
    // necessary (in which case `update` will be nullopt). False if the solver
    // had to be recreated.
    bool did_update;
 
    // The update that was attempted on the solver. Can be nullopt when no
    // update was needed (the model was not changed).
    std::optional<ModelUpdateProto> update;
  };
 
  // Creates a new incremental solve for the given model. It may returns an
  // error if the parameters are invalid (for example if the selected solver is
  // not linked in the binary).
  //
  // The returned IncrementalSolver keeps a copy of `arguments`. Thus the
  // content of arguments.non_streamable (for example pointers to solver
  // specific struct) must be valid until the destruction of the
  // IncrementalSolver.
  static absl::StatusOr<std::unique_ptr<IncrementalSolver>> New(
      Model& model, SolverType solver_type, SolverInitArguments arguments = {});
 
  // Updates the underlying solver with latest model changes and runs the solve.
  //
  // A Status error will be returned if there is an unexpected failure in an
  // underlying solver or for some internal math_opt errors. Otherwise, check
  // SolveResult::termination.reason to see if an optimal solution was found.
  //
  // Memory model: the returned SolveResult owns its own memory (for solutions,
  // solve stats, etc.), EXPECT for a pointer back to the model. As a result:
  //  * Keep the model alive to access SolveResult,
  //  * Avoid unnecessarily copying SolveResult,
  //  * The result is generally accessible after mutating this, but some care
  //    is needed if variables or linear constraints are added or deleted.
  //
  // Asserts (using CHECK) that the inputs arguments.model_parameters and
  // arguments.callback_registration only contain variables and constraints from
  // the input model.
  //
  // See callback.h for documentation on arguments.callback and
  // arguments.callback_registration.
  absl::StatusOr<SolveResult> Solve(const SolveArguments& arguments = {});
 
  // Updates the model to solve.
  //
  // This is an advanced API, most users should use Solve() above that does the
  // update and before calling the solver. Calling this function is only useful
  // for users that want to access to update data or users that need to use
  // SolveWithoutUpdate() (which should not be common).
  //
  // The returned value indicates if the update was possible or if the solver
  // had to be recreated from scratch (which may happen when the solver does not
  // support this specific update or any update at all). It also contains the
  // attempted update data.
  //
  // A status error will be returned if the underlying solver has an internal
  // error.
  absl::StatusOr<UpdateResult> Update();
 
  // Same as Solve() but does not update the underlying solver with the latest
  // changes to the model.
  //
  // This is an advanced API, most users should use Solve().
  absl::StatusOr<SolveResult> SolveWithoutUpdate(
      const SolveArguments& arguments = {}) const;
 
 private:
  IncrementalSolver(SolverType solver_type, SolverInitArguments init_args,
                    const ModelStorage* expected_storage,
                    std::unique_ptr<UpdateTracker> update_tracker,
                    std::unique_ptr<Solver> solver);
 
  const SolverType solver_type_;
  const SolverInitArguments init_args_;
  const ModelStorage* const expected_storage_;
  const std::unique_ptr<UpdateTracker> update_tracker_;
  std::unique_ptr<Solver> solver_;
};
 
}  // namespace math_opt
}  // namespace operations_research
 
#endif  // OR_TOOLS_MATH_OPT_CPP_SOLVE_H_