def test_current_model(self, two_reservoir_problem):
"""Test PywrRandomGenerator inserts the current model configuration in to the population."""
generator = PywrRandomGenerator(wrapper=two_reservoir_problem)
algorithm = NSGAII(two_reservoir_problem.problem,
population_size=10,
generator=generator)
algorithm.initialize()
# Ensure the first solution in the population has variable values from the model
solution = algorithm.population[0]
np.testing.assert_allclose(solution.variables, np.zeros(12))
def test_other_solutions(self, two_reservoir_problem, use_current):
"""Test PywrRandomGenerator inserts other solutions into the population."""
# Create some alternative initial solutions
solutions = [
{
"control_curve": {
"doubles": [1] * 12
}
},
{
"control_curve": {
"doubles": [2] * 12
}
},
]
generator = PywrRandomGenerator(wrapper=two_reservoir_problem,
solutions=solutions,
use_current=use_current)
algorithm = NSGAII(two_reservoir_problem.problem,
population_size=10,
generator=generator)
algorithm.initialize()
# Ensure the first solution in the population has variable values from the model
if use_current:
np.testing.assert_allclose(algorithm.population[0].variables,
np.zeros(12))
np.testing.assert_allclose(algorithm.population[1].variables,
np.ones(12))
np.testing.assert_allclose(algorithm.population[2].variables,
np.ones(12) * 2)
else:
np.testing.assert_allclose(algorithm.population[0].variables,
np.ones(12))
np.testing.assert_allclose(algorithm.population[1].variables,
np.ones(12) * 2)
