Solution hint is not passed to the HiGHS solver using mathopt #4345
Description
What version of OR-Tools and what language are you using?
Version: 9.10
Language: Python
Which solver are you using (e.g. CP-SAT, Routing Solver, GLOP, BOP, Gurobi)
HiGHS, SCIP
What operating system (Linux, Windows, ...) and version?
Ubuntu 22.04
What did you do?
Using the following simple model I am trying to pass to the solver a solution hint in order to warm start it.
from ortools.math_opt.python import mathopt, model_parameters
model = mathopt.Model()
x = model.add_integer_variable(name="x")
y = model.add_integer_variable(name="y")
model.add_linear_constraint(2 * x + 2 * y >= 7)
model.add_linear_constraint(3 * x + 4 * y >= 12)
model.add_linear_constraint(2 * x + y >= 6)
model.minimize_linear_objective(8 * x + 10 * y)
params = mathopt.SolveParameters()
params.enable_output = True
solution_hint = model_parameters.SolutionHint(variable_values={x: 1, y: 25})
mathopt.solve(
model,
mathopt.SolverType.HIGHS,
params=params,
model_params=model_parameters.ModelSolveParameters(solution_hints=[solution_hint]),
)The output below clearly indicates that the solution hint was not taken into account, as first BestSol is inf:
Coefficient ranges:
Matrix [1e+00, 4e+00]
Cost [8e+00, 1e+01]
Bound [0e+00, 0e+00]
RHS [6e+00, 1e+01]
Presolving model
3 rows, 2 cols, 6 nonzeros 0s
3 rows, 2 cols, 6 nonzeros 0s
Objective function is integral with scale 0.5
Solving MIP model with:
3 rows
2 cols (0 binary, 2 integer, 0 implied int., 0 continuous)
6 nonzeros
Nodes | B&B Tree | Objective Bounds | Dynamic Constraints | Work
Proc. InQueue | Leaves Expl. | BestBound BestSol Gap | Cuts InLp Confl. | LpIters Time
0 0 0 0.00% -inf inf inf 0 0 0 0 0.0s
T 0 0 0 0.00% -inf 32 Large 0 0 0 3 0.0s
Solving report
Status Optimal
Primal bound 32
Dual bound 32
Gap 0% (tolerance: 0.01%)
Solution status feasible
32 (objective)
0 (bound viol.)
1.7763568394e-15 (int. viol.)
0 (row viol.)
Timing 0.00 (total)
0.00 (presolve)
0.00 (postsolve)
Nodes 1
LP iterations 3 (total)
0 (strong br.)
0 (separation)
0 (heuristics)
The same model but solved with SCIP shows that solution hint was passed to the solver. The output for SCIP is below:
1/1 feasible solution given by solution candidate storage, new primal bound 2.580000e+02
presolving:
(round 1, fast) 0 del vars, 0 del conss, 0 add conss, 0 chg bounds, 1 chg sides, 0 chg coeffs, 0 upgd conss, 0 impls, 0 clqs
(round 2, exhaustive) 0 del vars, 0 del conss, 0 add conss, 0 chg bounds, 1 chg sides, 0 chg coeffs, 3 upgd conss, 0 impls, 0 clqs
Deactivated symmetry handling methods, since SCIP was built without symmetry detector (SYM=none).
Deactivated symmetry handling methods, since SCIP was built without symmetry detector (SYM=none).
presolving (3 rounds: 3 fast, 2 medium, 2 exhaustive):
0 deleted vars, 0 deleted constraints, 0 added constraints, 0 tightened bounds, 0 added holes, 1 changed sides, 0 changed coefficients
0 implications, 0 cliques
presolved problem has 2 variables (0 bin, 2 int, 0 impl, 0 cont) and 3 constraints
3 constraints of type <varbound>
transformed objective value is always integral (scale: 2)
Presolving Time: 0.00
transformed 2/2 original solutions to the transformed problem space
time | node | left |LP iter|LP it/n|mem/heur|mdpt |vars |cons |rows |cuts |sepa|confs|strbr| dualbound | primalbound | gap | compl.
i 0.0s| 1 | 0 | 0 | - | oneopt| 0 | 2 | 3 | 3 | 0 | 0 | 0 | 0 | -- | 4.800000e+01 | Inf | unknown
* 0.0s| 1 | 0 | 2 | - | LP | 0 | 2 | 3 | 3 | 0 | 0 | 0 | 0 | 3.200000e+01 | 3.200000e+01 | 0.00%| unknown
0.0s| 1 | 0 | 2 | - | 590k | 0 | 2 | 3 | 3 | 0 | 0 | 0 | 0 | 3.200000e+01 | 3.200000e+01 | 0.00%| unknown
SCIP Status : problem is solved [optimal solution found]
Solving Time (sec) : 0.01
Solving Nodes : 1
Primal Bound : +3.20000000000000e+01 (4 solutions)
Dual Bound : +3.20000000000000e+01
Gap : 0.00 %
What did you expect to see
The solution is passed to the HiGHS solver by mathopt interface. HiGHS indeed accepts the solution hint, checked using highspy. The mode using highspy:
from highspy import Highs, HighsSolution, HighsVarType
h = Highs()
x = h.addVariable(obj=8, name="x", type=HighsVarType.kInteger)
y = h.addVariable(obj=10, name="y", type=HighsVarType.kInteger)
h.addConstr(2 * x + 2 * y >= 7)
h.addConstr(3 * x + 4 * y >= 12)
h.addConstr(2 * x + y >= 6)
h.setMinimize()
sol = HighsSolution()
sol.col_value = [1, 25]
h.setSolution(sol)
h.solve()
And the output:
Running HiGHS 1.7.2 (git hash: 184e327): Copyright (c) 2024 HiGHS under MIT licence terms
Coefficient ranges:
Matrix [1e+00, 4e+00]
Cost [8e+00, 1e+01]
Bound [0e+00, 0e+00]
RHS [6e+00, 1e+01]
Assessing feasibility of MIP using primal feasibility and integrality tolerance of 1e-06
Solution has num max sum
Col infeasibilities 0 0 0
Integer infeasibilities 0 0 0
Row infeasibilities 0 0 0
Row residuals 0 0 0
Presolving model
3 rows, 2 cols, 6 nonzeros 0s
3 rows, 2 cols, 6 nonzeros 0s
MIP start solution is feasible, objective value is 258
Objective function is integral with scale 0.5
Solving MIP model with:
3 rows
2 cols (0 binary, 2 integer, 0 implied int., 0 continuous)
6 nonzeros
Nodes | B&B Tree | Objective Bounds | Dynamic Constraints | Work
Proc. InQueue | Leaves Expl. | BestBound BestSol Gap | Cuts InLp Confl. | LpIters Time
0 0 0 0.00% 0 258 100.00% 0 0 0 0 0.0s
T 0 0 0 0.00% 0 32 100.00% 0 0 0 3 0.0s
Solving report
Status Optimal
Primal bound 32
Dual bound 32
Gap 0% (tolerance: 0.01%)
Solution status feasible
32 (objective)
0 (bound viol.)
1.7763568394e-15 (int. viol.)
0 (row viol.)
Timing 0.01 (total)
0.00 (presolve)
0.00 (postsolve)
Nodes 1
LP iterations 3 (total)
0 (strong br.)
0 (separation)
0 (heuristics)
What did you see instead?
The solution hint was not passed HiGHS using mathopt interface.