def test_basic_infeasible_init(model):
model.c2 = Constraint(expr=model.y <= 50)
tc, prog, ni = homotopy(model, [model.x], [20])
assert tc == TerminationCondition.infeasible
assert prog == 0
assert ni == 0
def test_basic_max_iter(model):
tc, prog, ni = homotopy(model, [model.x], [20], max_eval=2)
assert pytest.approx(model.y.value, 1e-4) == 182.25
assert tc == TerminationCondition.maxEvaluations
assert prog == 0.35
assert ni == 2
def test_basic(model):
tc, prog, ni = homotopy(model, [model.x], [20])
assert model.y.value == 400
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 4
def test_basic_overshoot(model):
# Use a big step such that overshoot will occur if not caught
tc, prog, ni = homotopy(model, [model.x], [20], step_init=0.6)
assert model.y.value == 400
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 2
def test_basic_step_accel(model):
# With zero acceleration, should take 10 steps
tc, prog, ni = homotopy(model, [model.x], [20], step_accel=0)
assert model.y.value == 400
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 10
def test_basic_iter_target(model):
# Should take 5 steps
tc, prog, ni = homotopy(model, [model.x], [20], iter_target=2)
assert model.y.value == 400
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 5
def test_basic_max_step(model):
# With max_step = step_init = 0.1, should take 10 steps
tc, prog, ni = homotopy(model, [model.x], [20], max_step=0.1)
assert model.y.value == 400
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 10
def test_basic_constraint_violation(model):
# Add a constraint to limit y
model.c2 = Constraint(expr=model.y <= 300)
# Try to walk to a value of x that gives an infeasible value of y
tc, prog, ni = homotopy(model, [model.x], [20])
assert pytest.approx(model.y.value, 1e-4) == 293.27
assert tc == TerminationCondition.minStepLength
assert prog == 0.7125
assert ni == 12
def test_basic_step_cut_2(model):
# Add a constraint to limit y
model.c2 = Constraint(expr=model.y <= 196)
# Should take 7 steps
# 4 steps to reach 14, and 3 to cut back to min_step
tc, prog, ni = homotopy(model, [model.x], [20], step_init=0.1,
min_step=0.01, step_cut=0.25, step_accel=0)
assert model.y.value == 196
assert tc == TerminationCondition.minStepLength
assert prog == 0.4
assert ni == 7
def test_ideal_prop(model2):
tc, prog, ni = homotopy(model2, [model2.fs.state_block.temperature], [390])
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 5
# Check for VLE results
assert model2.fs.state_block.mole_frac_phase_comp["Liq", "benzene"].value \
== pytest.approx(0.291, abs=1e-3)
assert model2.fs.state_block.mole_frac_phase_comp["Liq", "toluene"].value \
== pytest.approx(0.709, abs=1e-3)
assert model2.fs.state_block.mole_frac_phase_comp["Vap", "benzene"].value \
== pytest.approx(0.5, abs=1e-5)
assert model2.fs.state_block.mole_frac_phase_comp["Vap", "toluene"].value \
== pytest.approx(0.5, abs=1e-5)
def test_ideal_prop_max_iter(model2):
tc, prog, ni = homotopy(
model2, [model2.fs.state_block.temperature], [390],
max_solver_iterations=3, min_step=0.01, step_accel=0)
assert tc == TerminationCondition.optimal
assert prog == 1
assert ni == 19
# Check for VLE results
assert model2.fs.state_block.mole_frac_phase_comp["Liq", "benzene"].value \
== pytest.approx(0.291, abs=1e-3)
assert model2.fs.state_block.mole_frac_phase_comp["Liq", "toluene"].value \
== pytest.approx(0.709, abs=1e-3)
assert model2.fs.state_block.mole_frac_phase_comp["Vap", "benzene"].value \
== pytest.approx(0.5, abs=1e-5)
assert model2.fs.state_block.mole_frac_phase_comp["Vap", "toluene"].value \
== pytest.approx(0.5, abs=1e-5)
def test_current_value_ub(model):
model.x.setub(5)
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20])
def test_var_not_fixed(model):
model.x.unfix()
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20])
def test_var_not_in_model(model):
m2 = ConcreteModel()
with pytest.raises(ConfigurationError):
homotopy(m2, [model.x], [20])
def test_not_var(model):
model.p = Param()
with pytest.raises(TypeError):
homotopy(model, [model.p], [20])
def test_invalid_model(model):
o = object()
with pytest.raises(TypeError):
homotopy(o, [model.x], [20])
def test_target_value_lb(model):
model.x.fix(50)
model.x.setlb(30)
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20])
def test_step_cut(model):
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], step_cut=0.09)
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], step_cut=0.91)
def test_step_accel(model):
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], step_accel=-1)
def test_iter_target(model):
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], iter_target=0)
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], iter_target=1.7)
def test_min_max_step(model):
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], min_step=0.1, max_step=0.05)
def test_init_step_min_max(model):
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], step_init=0.2, max_step=0.1)
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], step_init=0.05, min_step=0.1)
def test_max_eval(model):
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], max_eval=0)
with pytest.raises(ConfigurationError):
homotopy(model, [model.x], [20], max_eval=1.7)
