docs: `addConsIndicator` doc says cons must be `<=`, but `>=` works fine

[hbierlee]

In working the integration of SCIP (1.17.1) with CPMpy (CPMpy/cpmpy#412), I found that in the docs, it says indicator constraints have to be <=, but it seems >= does also work.

from pyscipopt import Model

m = Model()
m.hideOutput()

x = m.addVar(vtype="I", lb=0, ub=10, name="x")
b = m.addVar(vtype="B", name="b")

# This works despite docs saying only "<=" is accepted
m.addConsIndicator(x >= 5, binvar=b)
m.addCons(b == 1)
m.setObjective(x, sense="minimize")
m.optimize()

print(f"Status: {m.getStatus()}")
print(f"x = {m.getVal(x)}")
print(f"b = {m.getVal(b)}")
assert m.getStatus() == "optimal"
assert m.getVal(x) == 5

# Output:
# Status: optimal
# x = 5.0
# b = 1.0
# All assertions passed

Cheers,

Hendrik 'Henk' Bierlee

[Joao-Dionisio]

Thank you, Hendrik! Likely a relic from the past, I'll update it soon 😄

[hbierlee]

By the way @Joao-Dionisio I suspect the same may be true for addMatrixConsIndicator, although I didn't test that one.

[Joao-Dionisio]

🤦 I wonder when I will stop forgetting about the matrix API. Thank you again!