Creating constraints dynamically in pyomo abstract model

Question

I have a networkX graph with few nodes and these nodes have attributes such as "demand".

def mygraph():
    G = nx.Graph()
    G.add_nodes_from([("N1", {"demand": 10}),
    ("N2"{"demand": 12}),
    ("N3", {"demand": 25}),
    ("N4"{"demand": 18})])

I want my Pyomo Abstract model to create constraints and decision variables dynamically. Like,

def mymodel():
    model = AbstractModel()
    g=mygraph() #mygraph passed to abstract model
    model.nodes_range = RangeSet(1,len(g))# this creates a parameter with same size of the graph
    model.C = Param(model.nodes_range, within=pyo.NonNegativeIntegers) #one parameter for each node
def constraint_rule(model, i): #dummy constraint
     return sum(model.decision_var*demand_node1)&lt;=something
model.const1=Constraint(model.nodes_range, rule=constraint_rule)

model.obj1 = Objective(my objective) #my dummy objective

status = SolverFactory('glpk')
results = status.solve(model)
assert_optimal_termination(results)
model.display()

mymodel()

But the model.const1.pprint() command is giving 0 constraints. Can you guide me with the logic?

score 3 · Answer 1 · answered Sep 14 '22 at 14:11

You are missing a model.create_instance call to actually construct the abstract model into a concrete instance. Calling pprint on abstract model components will always return empty components because they haven't been constructed.

I don't see anything in your example model that would require an AbstractModel so you could also try using a ConcreteModel instead and then you could call pprint on components any time after declaration.

Creating constraints dynamically in pyomo abstract model

1 Answers1