OMMX adapter for Gurobi

This package provides an adapter for Gurobi from OMMX. It allows you to solve optimization problems defined in OMMX format using Gurobi's powerful solver.

Features

• Support for various variable types:
  • Binary variables
  • Integer variables
  • Continuous variables
• Support for different optimization problems:
  • Linear Programming (LP)
  • Mixed Integer Linear Programming (MILP)
  • Quadratic Programming (QP)
  • Mixed Integer Quadratic Programming (MIQP)
• Support for both minimization and maximization problems

Prerequisites

• Python >= 3.9
• Gurobi Optimizer and valid license
• gurobipy >= 10.0.0
• ommx >= 1.8.4

Installation

First, ensure you have Gurobi installed and properly licensed. Then install the OMMX Gurobi adapter using pip:

pip install ommx-gurobipy-adapter

Usage

Here's a simple example of how to use the adapter:

from ommx_gurobipy_adapter import OMMXGurobipyAdapter
from ommx.v1 import Instance, DecisionVariable

# Create decision variables
x1 = DecisionVariable.integer(1, lower=0, upper=5)
x2 = DecisionVariable.continuous(2, lower=0, upper=5)

# Create OMMX instance
instance = Instance.from_components(
    decision_variables=[x1, x2],
    objective=x1 + 2*x2,
    constraints=[
        x1 + x2 <= 5,  # Linear constraint
    ],
    sense=Instance.MAXIMIZE,
)

# Solve using Gurobi
solution = OMMXGurobipyAdapter.solve(instance)

# Access the results
print(f"Objective value: {solution.objective}")
print(f"x1 = {solution.state.entries[1]}")
print(f"x2 = {solution.state.entries[2]}")

Controlling Gurobi Parameters

If you need more control over the Gurobi solver parameters, you can use the adapter in two steps:

from ommx_gurobipy_adapter import OMMXGurobipyAdapter

# Create adapter
adapter = OMMXGurobipyAdapter(instance)

# Get Gurobi model
model = adapter.solver_input

# Set Gurobi parameters
model.setParam('TimeLimit', 60)  # Set time limit to 60 seconds
model.setParam('MIPGap', 0.01)   # Set relative MIP gap tolerance to 1%

# Solve
model.optimize()

# Get solution
solution = adapter.decode(model)

Error Handling

The adapter provides specific error types for different situations:

• OMMXGurobipyAdapterError: Base error class for adapter-specific errors
• InfeasibleDetected: Raised when the problem is infeasible
• UnboundedDetected: Raised when the problem is unbounded

Example of error handling:

from ommx_gurobipy_adapter import OMMXGurobipyAdapterError
from ommx.adapter import InfeasibleDetected, UnboundedDetected

try:
    solution = OMMXGurobipyAdapter.solve(instance)
except InfeasibleDetected:
    print("Problem is infeasible")
except UnboundedDetected:
    print("Problem is unbounded")
except OMMXGurobipyAdapterError as e:
    print(f"Adapter error: {e}")

Testing

• Install uv

• Run pytest and markdown-code-runner:

  uv run pytest -vv
  uv run markdown-code-runner README.md

License

This project (the source code contained in this repository) is licensed under the Apache License 2.0.

Warning

The package ommx-gurobipy-adapter depends on gurobipy, which is a proprietary software. Please make sure you have a valid license for Gurobi Optimizer when using this package.

Reference

For more information about OMMX, please visit: https://github.com/Jij-Inc/ommx

For Gurobi documentation, please visit: https://www.gurobi.com/documentation/

Contributing

Contributions are welcome! Please feel free to submit issues and pull requests.