TypeError in problem.py: unsupported operand type(s) for *: 'float' and 'NoneType'

[fracerma]

Description

I get this error frompymoo/core/problem.py when instantiating a subclass of Problem.
In my case I pass vars as a dictionary: vars = { 'x1': Choice(), 'x2' Choice(), 'x3': Integer(lb=1, ub=10) leaving xl and xu as None.

I think there is a problem in this part:

...
self.xl, self.xu = xl, xu

# if the variables are provided in their explicit form
if vars is not None:
    self.vars = vars
    self.n_var = len(vars)
    if self.xl is None:
        self.xl = {name: var.lb if hasattr(var, "lb") else None for name, var in vars.items()}
    if self.xu is None:
        self.xu = {name: var.ub if hasattr(var, "ub") else None for name, var in vars.items()}

...

# if it is a problem with an actual number of variables - make sure xl and xu are numpy arrays
if n_var > 0:

    if self.xl is not None:
        if not isinstance(self.xl, np.ndarray):
            self.xl = np.ones(n_var) * xl
        self.xl = self.xl.astype(float)

    if self.xu is not None:
        if not isinstance(self.xu, np.ndarray):
            self.xu = np.ones(n_var) * xu
        self.xu = self.xu.astype(float)

xl and xu are not checked to be not None and I'd changed it as:

if n_var > 0:

            if xl is not None:
                if not isinstance(self.xl, np.ndarray):
                    self.xl = np.ones(n_var) * xl
                self.xl = self.xl.astype(float)

            if xu is not None:
                if not isinstance(self.xu, np.ndarray):
                    self.xu = np.ones(n_var) * xu
                self.xu = self.xu.astype(float)

self.xl/xu should be a numpy array or a dictionary?
I also don't understand when this values are used, since the upper and lower bound are taken directly from each different variable type, right?

Environment

• OS: Ubuntu 20.04
• Python Version: 3.7
• Package Versions:
  • numpy: 1.21.6
  • pymoo: 0.6.0.1

Error Traceback

...
  File "/home/user/miniconda3/envs/fatai/lib/python3.7/site-packages/pymoo/core/problem.py", line 158, in __init__
    self.xl = np.ones(n_var) * xl
TypeError: unsupported operand type(s) for *: 'float' and 'NoneType'

[blankjul]

Can you please provide an example to reproduce the error?

The Integer variable takes bounds as an input. See the following example below:

from pymoo.algorithms.moo.nsga2 import NSGA2
from pymoo.core.mixed import MixedVariableSampling, MixedVariableMating, \
    MixedVariableDuplicateElimination
from pymoo.core.problem import ElementwiseProblem
from pymoo.core.variable import Choice, Integer
from pymoo.optimize import minimize


class MultiObjectiveMixedVariableProblem(ElementwiseProblem):

    def __init__(self, **kwargs):
        vars = {'x1': Choice(options=[0,1]), 'x2': Choice(options=[0,1]), 'x3': Integer(bounds=(1, 10))}
        super().__init__(vars=vars, n_obj=2, n_ieq_constr=0, **kwargs)

    def _evaluate(self, X, out, *args, **kwargs):
        out["F"] = [X["x1"], X["x2"] * X["x3"]]


problem = MultiObjectiveMixedVariableProblem()

algorithm = NSGA2(pop_size=50,
                  sampling=MixedVariableSampling(),
                  mating=MixedVariableMating(eliminate_duplicates=MixedVariableDuplicateElimination()),
                  eliminate_duplicates=MixedVariableDuplicateElimination(),
                  )


res = minimize(problem,
               algorithm,
               seed=1,
               verbose=True)

[fracerma]

Hello,
the same error happens when using the Problem class and inititializing with n_var setted:

from pymoo.algorithms.moo.nsga2 import NSGA2
from pymoo.core.mixed import MixedVariableSampling, MixedVariableMating, \
    MixedVariableDuplicateElimination
from pymoo.core.problem import Problem
from pymoo.core.variable import Choice, Integer
from pymoo.optimize import minimize


class MultiObjectiveMixedVariableProblem(Problem):

    def __init__(self, **kwargs):
        vars = {'x1': Choice(options=[0,1]), 'x2': Choice(options=[0,1]), 'x3': Integer(bounds=(1, 10))}
        super().__init__(vars=vars, n_obj=2, n_ieq_constr=0, n_var=3, **kwargs)

    def _evaluate(self, X, out, *args, **kwargs):
        out["F"] = [[x["x1"], x["x2"] * x["x3"]] for x in X]


problem = MultiObjectiveMixedVariableProblem()

algorithm = NSGA2(pop_size=50,
                  sampling=MixedVariableSampling(),
                  mating=MixedVariableMating(eliminate_duplicates=MixedVariableDuplicateElimination()),
                  eliminate_duplicates=MixedVariableDuplicateElimination(),
                  )


res = minimize(problem,
               algorithm,
               seed=1,
               verbose=True)