Pyomo: implement build_maxcut_model; add support for quadratic objectives

miplearn/problems/maxcut.py

@@ -8,13 +8,14 @@ from typing import List, Union, Optional, Any
 import gurobipy as gp
 import networkx as nx
 import numpy as np
-from gurobipy import quicksum
+import pyomo.environ as pe
 from networkx import Graph
 from scipy.stats.distributions import rv_frozen
 
 from miplearn.io import read_pkl_gz
-from miplearn.problems import _gurobipy_set_params
+from miplearn.problems import _gurobipy_set_params, _pyomo_set_params
 from miplearn.solvers.gurobi import GurobiModel
+from miplearn.solvers.pyomo import PyomoModel
 
 
 @dataclass
@@ -99,7 +100,7 @@ def build_maxcut_model_gurobipy(
 
     # Add the objective function
     model.setObjective(
-        quicksum(
+        gp.quicksum(
             -data.weights[i] * x[e[0]] * (1 - x[e[1]]) for (i, e) in enumerate(edges)
         )
     )
@@ -108,6 +109,37 @@ def build_maxcut_model_gurobipy(
     return GurobiModel(model)
 
 
+def build_maxcut_model_pyomo(
+    data: Union[str, MaxCutData],
+    solver: str = "gurobi_persistent",
+    params: Optional[dict[str, Any]] = None,
+) -> PyomoModel:
+    # Initialize model
+    model = pe.ConcreteModel()
+
+    # Read data
+    data = _maxcut_read(data)
+    nodes = pe.Set(initialize=list(data.graph.nodes))
+    edges = list(data.graph.edges())
+
+    # Add decision variables
+    model.x = pe.Var(nodes, domain=pe.Binary, name="x")
+
+    # Add the objective function
+    model.obj = pe.Objective(
+        expr=pe.quicksum(
+            -data.weights[i] * model.x[e[0]]
+            + data.weights[i] * model.x[e[0]] * model.x[e[1]]
+            for (i, e) in enumerate(edges)
+        ),
+        sense=pe.minimize,
+    )
+    model.pprint()
+    pm = PyomoModel(model, solver)
+    _pyomo_set_params(model, params, solver)
+    return pm
+
+
 def _maxcut_read(data: Union[str, MaxCutData]) -> MaxCutData:
     if isinstance(data, str):
         data = read_pkl_gz(data)