in CVXPY: Give a name or ID to constraints

Question

In CVXPY, I have found that I am able to give a name or ID to the variables I define.

Now I would also like to give a string name the constraints I define, so that I can find them easily later. The reason is that I have many agents, which are defined in similar ways, but I would like to know which of the constraints belong to which agent. (so give the constraint the number of the agent for example).

I have looked in the documentation, where the Constraint class is defined as:

cvxpy.constraints.constraint.Constraint(args, constr_id=None)

Where the parameters are defined as

   args (list) – A list of expression trees.

   constr_id (int) – A unique id for the constraint.

So it seems that I can give an integer ID, but not a name.

Does anyone know whether it is possible to give a string name to a constraint?

Answer 1

I don't know if that's possible, but i would not try to dive that deep into it's internal-code.

Why not just wrap it away? This should be more safe and customizable.

Prototype (which does not check for duplicates for example):

import cvxpy as cp

class ConstraintBuilder:
    def __init__(self):
        self.constraintList = []
        self.str2constr = {}

    def addConstr(self, expr, str_):
        self.constraintList.append(expr)
        self.str2constr[str_] = len(self.constraintList)-1

    def get(self):
        return self.constraintList

    def getConstr(self, str_):
        return self.constraintList[self.str2constr[str_]]

####

cb = ConstraintBuilder()

x = cp.Variable(5)

cb.addConstr(x >= 0, 'nonnegative')
cb.addConstr(x[0] + x[1] >= 0.3, 'pair0')
cb.addConstr(x[0] + x[1] >= 0.6, 'pair1')

objective = cp.Minimize(cp.sum(x))
problem = cp.Problem(objective, cb.get())

problem.solve(verbose=True, solver=cp.ECOS)
print(problem.status)

print(cb.getConstr('pair1'))

which outputs:

ECOS 2.0.7 - (C) embotech GmbH, Zurich Switzerland, 2012-15. Web: www.embotech.com/ECOS

It     pcost       dcost      gap   pres   dres    k/t    mu     step   sigma     IR    |   BT
 0  +3.600e-01  +3.690e-01  +5e+00  7e-01  1e-02  1e+00  8e-01    ---    ---    1  1  - |  -  - 
 1  +4.151e-01  +4.114e-01  +8e-01  2e-01  2e-03  1e-01  1e-01  0.8590  3e-03   1  1  1 |  0  0
 2  +5.863e-01  +5.831e-01  +9e-02  3e-02  2e-04  1e-02  1e-02  0.9795  1e-01   1  1  1 |  0  0
 3  +5.998e-01  +5.998e-01  +1e-03  3e-04  2e-06  1e-04  2e-04  0.9887  1e-04   1  1  1 |  0  0
 4  +6.000e-01  +6.000e-01  +1e-05  3e-06  2e-08  2e-06  2e-06  0.9890  1e-04   1  0  0 |  0  0
 5  +6.000e-01  +6.000e-01  +1e-07  4e-08  2e-10  2e-08  2e-08  0.9890  1e-04   1  0  0 |  0  0
 6  +6.000e-01  +6.000e-01  +1e-09  4e-10  3e-12  2e-10  2e-10  0.9890  1e-04   1  0  0 |  0  0

OPTIMAL (within feastol=4.2e-10, reltol=2.5e-09, abstol=1.5e-09).
Runtime: 0.000236 seconds.

optimal
0.6 <= var0[0] + var0[1]