Gurobi fails to find feasible solutions for a simple optimization problem with sin function

Answered

Tong Han

August 08, 2024 04:46

• 
• 
• 

I am testing the nonconvex optimization capability of Gurobi. I found Gurobi even fails to find a feasible solution for a very simple problem with sin function.Can anyone help me with this problem? 

The optimization problem is: min x ; s.t. sin(x) <= -0.1, -pi <= x <= pi. The code is given as follows: 

import numpy as np

import gurobipy as gp

from gurobipy import GRB

 

model = gp.Model()

x = model.addMVar(shape=(1, 1), vtype=GRB.CONTINUOUS, name="x")

y = model.addMVar(shape=(1, 1), vtype=GRB.CONTINUOUS, name="y")

gc1 = model.addGenConstrSin(x, y)

model.addConstr(y <= -0.1 )

model.addConstr(x >= -np.pi)

model.addConstr(x <= np.pi)

model.setObjective(x, GRB.MINIMIZE)

model.optimize()

print(x.X)

 

The output is given below: 

Set parameter Username

Academic license - for non-commercial use only - expires 2025-06-03

Gurobi Optimizer version 10.0.1 build v10.0.1rc0 (linux64)

CPU model: 13th Gen Intel(R) Core(TM) i9-13900HX, instruction set [SSE2|AVX|AVX2]

Thread count: 32 physical cores, 32 logical processors, using up to 32 threads

Optimize a model with 3 rows, 2 columns and 3 nonzeros

Model fingerprint: 0x56424c07

Model has 1 general constraint

Variable types: 2 continuous, 0 integer (0 binary)

Coefficient statistics:

Matrix range [1e+00, 1e+00]

Objective range [1e+00, 1e+00]

Bounds range [0e+00, 0e+00]

RHS range [1e-01, 3e+00]

Presolve removed 1 rows and 0 columns

Presolve time: 0.00s

Explored 0 nodes (0 simplex iterations) in 0.00 seconds (0.00 work units)

Thread count was 1 (of 32 available processors)

Solution count 0

Model is infeasible

Best objective -, best bound -, gap -

0

• 

  Riley Clement

  • Gurobi Staff

  August 08, 2024 05:05

  Hi Tong,

  Gurobi variables have a default lower bound of 0.  If you want x to be able to take negative values you will need to specify a lower bound when creating the variables.

  - Riley

  0
• 

  Tong Han

  August 08, 2024 05:59

  • 
  • 
  • 

  Hi Riley,

  I got the right result after adding bounds for both x and y. Thanks so much. 

  - Tong 

  0

Please sign in to leave a comment.