Representation of tuple set in Pyomo - tuples

Question

I am trying to solve a problem in Pyomo. I need to write down restrictions on a subset A for some arcs between nodes (I and II) for a Constraint1. I wrote a model:
from pyomo.environ import *
model = AbstractModel()
model.I = Set()
model.II = SetOf(model.I)
model.J = Set()
model.A = Set(model.I,model.II)
model.c = Param(model.I, model.J, default=0)
model.b = Param(model.I, default=0)
model.x = Var(model.I, model.J, within=NonNegativeReals)
model.y = Var(model.I, model.II, within=NonNegativeReals)
data = DataPortal()
data.load(filename='Data.yaml')
data.load(filename='Table.tab')
def objective_rule(model):
return (sum(model.c[i,j]*model.x[i,j] for i in model.I for j in model.J))
model.OBJ = Objective(rule=objective_rule, sense = minimize)
def B_rule(model,i):
Bb = sum(model.x[i,j] for j in model.J)-sum(model.y[i,ii] for ii in model.II if ii != i)+0.01*sum(model.y[ii,i] for ii in model.II if ii != i)
return model.b[i] == Bb
model.B1 = Constraint(model.I, rule=B_rule)
def constraint1_rule(model,i,ii):
if (i,ii) in model.A:
return model.y[i,ii] <= 10000
return Constraint.Skip
model.constraint1 = Constraint(model.I, model.II, rule = constraint1_rule)
instance = model.create_instance(data)
opt = SolverFactory('cplex')
opt.solve(instance)
instance.OBJ.display()
instance.x.display()
instance.y.display()
The data is presented in the file Data.yaml:
I: [1, 2, 3, 4]
J: [1, 2]
b : {1: 10000, 2: 20000, 3: 25000, 4: 22000}
c:
- index: [1, 1]
value: 550
- index: [2, 2]
value: 120
- index: [3, 1]
value: 650
- index: [4, 2]
value: 550
- index: [1, 1]
value: 120
- index: [2, 2]
value: 650
- index: [3, 1]
value: 650
- index: [4, 2]
value: 550
The two-dimensional set A is presented in the file Table.tab:
set A : 1 2 3 4 :=
1 - - + +
2 - - + +
3 + + - +
4 + + + - ;
After the solution I get the error:
Unspecified format and data option
How to correctly represent the set A?

Answer 1

First, your data and sets are so small, that I would really consider just putting them into the base file and making a concrete model. Much easier to troubleshoot. You could even use basic python to read from csv files or such.
Anyhow. You can find a good example here on working with .tab files:
https://pyomo.readthedocs.io/en/stable/working_abstractmodels/data/dataportals.html?highlight=.tab#loading-structured-data
2 things you need to do:
change your statement for model.A to just show that it is within the sets.
model.A = Set(within = model.I*model.I)
Clean up your tab file to show the set index in top left and clear out the punctuation:
I 1 2 3 4
1 - - + +
2 - - + +
3 + + - +
4 + + + -
update your read statement:
data.load(filename='table.tab', set=model.A, format='set_array')