def __init__(self, n=10, derivs=False):
super(BruteForceSellarProblem, self).__init__(impl=impl)
root = self.root = Group()
if not derivs:
root.deriv_options['type'] = 'fd'
sellars = root.add('sellars', ParallelGroup())
for i in range(n):
name = 'sellar%i' % i
sellars.add(name, SellarDerivatives())
root.connect('dist_x', 'sellars.' + name + '.x', src_indices=[i])
root.connect('dist_z',
'sellars.' + name + '.z',
src_indices=[i * 2, i * 2 + 1])
root.connect('sellars.' + name + '.obj', 'collect.obj_%i' % i)
root.connect('sellars.' + name + '.con1', 'collect.con1_%i' % i)
root.connect('sellars.' + name + '.con2', 'collect.con2_%i' % i)
root.add('indep',
IndepVarComp([('x', 1.0), ('z', np.array([5.0, 2.0]))]),
promotes=['x', 'z'])
root.add(
'random',
Randomize(
n=n,
params=[
# name, value, std dev
('x', 1.0, 1e-2),
('z', np.array([5.0, 2.0]), 1e-2)
]),
promotes=['x', 'z', 'dist_x', 'dist_z'])
root.add('collect',
Collector(n=n, names=['obj', 'con1', 'con2']),
promotes=['obj', 'con1', 'con2'])
# top level driver setup
self.driver = ScipyOptimizer()
self.driver.options['optimizer'] = 'SLSQP'
self.driver.options['tol'] = 1.0e-8
self.driver.options['maxiter'] = 50
self.driver.options['disp'] = False
self.driver.add_desvar('z',
lower=np.array([-10.0, 0.0]),
upper=np.array([10.0, 10.0]))
self.driver.add_desvar('x', lower=0.0, upper=10.0)
self.driver.add_objective('obj')
self.driver.add_constraint('con1', upper=0.0)
self.driver.add_constraint('con2', upper=0.0)
def cobyla_optimize(iteration_count, dimensions):
top = Problem()
root = top.root = Group()
root.add('p', RosenbrockMultiDim(dimensions))
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'COBYLA'
top.driver.options['maxiter'] = iteration_count
top.driver.add_objective('p.f')
for i in range(dimensions):
componentName = 'p{0}'.format(i)
variableName = 'x{0}'.format(i)
portName = '{0}.{1}'.format(componentName, variableName)
root.add(componentName, IndepVarComp(variableName, 0.0))
root.connect(portName, 'p.{0}'.format(variableName))
top.driver.add_desvar(portName, lower=-50, upper=50)
top.setup()
top.run()
result_x = []
for i in range(dimensions):
result_x.append(top['p.x{0}'.format(i)])
print('\n')
print('Minimum of {0} found at {1}'.format(top['p.f'], result_x))
return top['p.f']
def test_opt_cylinder(self):
# this is just here to make sure that the cylinder optimization works normally so
# we know if the opt of the nested cylinder fails it's not due to some cylinder
# model issue.
prob = Problem(root=CylinderGroup())
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['disp'] = False
prob.driver.add_desvar("indep.r", lower=0.0, upper=1.e99)
prob.driver.add_desvar("indep.h", lower=0.0, upper=1.e99)
prob.driver.add_objective("cylinder.area")
prob.driver.add_constraint("cylinder.volume", equals=1.5)
prob.setup(check=False)
prob.run()
for name, opt in cylinder_opts:
self.assertAlmostEqual(prob[name], opt,
places=4,
msg="%s should be %s, but got %s" %
(name, opt, prob[name]))
self.assertAlmostEqual(prob['cylinder.volume'], 1.5,
places=4,
msg="volume should be 1.5, but got %s" %
prob['cylinder.volume'])
def main():
print("Bayesopt OpenMDAO Optimizer example")
top = Problem()
root = top.root = Group()
# Initial value of x and y set in the IndepVarComp.
root.add('p1', IndepVarComp('x', 13.0))
root.add('p2', IndepVarComp('y', -14.0))
root.add('p', Paraboloid())
root.connect('p1.x', 'p.x')
root.connect('p2.y', 'p.y')
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'COBYLA'
top.driver.add_desvar('p1.x', lower=-50, upper=50)
top.driver.add_desvar('p2.y', lower=-50, upper=50)
top.driver.add_objective('p.f_xy')
top.setup()
# You can also specify initial values post-setup
top['p1.x'] = 3.0
top['p2.y'] = -4.0
top.run()
print('\n')
print('Minimum of %f found at (%f, %f)' % (top['p.f_xy'], top['p.x'], top['p.y']))
def test_index_array_param(self):
prob = Problem()
prob.root = SellarStateConnection()
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1.0e-8
prob.driver.add_desvar('z',
lower=np.array([-10.0]),
upper=np.array([10.0]),
indices=[0])
prob.driver.add_desvar('x', lower=0.0, upper=10.0)
prob.driver.add_objective('obj')
prob.driver.add_constraint('con1', upper=0.0)
prob.driver.add_constraint('con2', upper=0.0)
prob.driver.options['disp'] = False
prob.setup(check=False)
prob['z'][1] = 5.0
prob.run()
assert_rel_error(self, prob['z'][0], 0.1005, 1e-3)
assert_rel_error(self, prob['z'][1], 5.0, 1e-3)
assert_rel_error(self, prob['x'], 0.0, 1e-3)
def test_sellar_state_connection(self):
top = Problem()
top.root = SellarStateConnection()
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'SLSQP'
top.driver.options['tol'] = 1.0e-8
top.driver.options['disp'] = False
top.driver.add_desvar('z',
lower=np.array([-10.0, 0.0]),
upper=np.array([10.0, 10.0]))
top.driver.add_desvar('x', lower=0.0, upper=10.0)
top.driver.add_objective('obj')
top.driver.add_constraint('con1', upper=0.0)
top.driver.add_constraint('con2', upper=0.0)
top.setup(check=False)
top.run()
assert_rel_error(self, top['z'][0], 1.977639, 1e-5)
assert_rel_error(self, top['z'][1], 0.0, 1e-5)
assert_rel_error(self, top['x'], 0.0, 1e-5)
assert_rel_error(self, top['obj'], 3.1833940, 1e-5)
def test_sellar_sand_architecture(self):
top = Problem()
top.root = SellarSAND()
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'SLSQP'
top.driver.options['tol'] = 1.0e-12
top.driver.options['disp'] = False
top.driver.add_desvar('z',
lower=np.array([-10.0, 0.0]),
upper=np.array([10.0, 10.0]))
top.driver.add_desvar('x', lower=0.0, upper=10.0)
top.driver.add_desvar('y1', lower=-10.0, upper=10.0)
top.driver.add_desvar('y2', lower=-10.0, upper=10.0)
top.driver.add_objective('obj')
top.driver.add_constraint('con1', upper=0.0)
top.driver.add_constraint('con2', upper=0.0)
top.driver.add_constraint('resid1', equals=0.0)
top.driver.add_constraint('resid2', equals=0.0)
top.setup(check=False)
top.run()
assert_rel_error(self, top['z'][0], 1.9776, 1e-3)
assert_rel_error(self, top['z'][1], 0.0000, 1e-3)
assert_rel_error(self, top['x'], 0.0000, 1e-3)
assert_rel_error(self, top['d1.y1'], 3.1600, 1e-3)
assert_rel_error(self, top['d1.y2'], 3.7553, 1e-3)
assert_rel_error(self, top['obj'], 3.1834, 1e-3)
def test_paraboloid_optimize_unconstrained(self):
top = Problem()
root = top.root = Group()
root.add('p1', IndepVarComp('x', 3.0))
root.add('p2', IndepVarComp('y', -4.0))
root.add('p', ParaboloidOptUnCon())
root.connect('p1.x', 'p.x')
root.connect('p2.y', 'p.y')
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'SLSQP'
top.driver.options['disp'] = False
top.driver.add_desvar('p1.x', lower=-50, upper=50)
top.driver.add_desvar('p2.y', lower=-50, upper=50)
top.driver.add_objective('p.f_xy')
top.setup(check=False)
top.run()
assert_rel_error(self, top['p.x'], 6.666667, 1e-6)
assert_rel_error(self, top['p.y'], -7.333333, 1e-6)
def test_beam_tutorial(self):
top = Problem()
top.root = BeamTutorial()
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'SLSQP'
top.driver.options['tol'] = 1.0e-8
top.driver.options['maxiter'] = 10000 #maximum number of solver iterations
top.driver.options['disp'] = False
#room length and width bounds
top.driver.add_desvar('ivc_rlength.room_length', lower=5.0*12.0, upper=50.0*12.0) #domain: 1in <= length <= 50ft
top.driver.add_desvar('ivc_rwidth.room_width', lower=5.0*12.0, upper=30.0*12.0) #domain: 1in <= width <= 30ft
top.driver.add_objective('d_neg_area.neg_room_area') #minimize negative area (or maximize area)
top.driver.add_constraint('d_len_minus_wid.length_minus_width', lower=0.0) #room_length >= room_width
top.driver.add_constraint('d_deflection.deflection', lower=720.0) #deflection >= 720
top.driver.add_constraint('d_bending.bending_stress_ratio', upper=0.5) #bending < 0.5
top.driver.add_constraint('d_shear.shear_stress_ratio', upper=1.0/3.0) #shear < 1/3
top.setup(check=False)
top.run()
assert_rel_error(self, -top['d_neg_area.neg_room_area'], 51655.257618, .01)
assert_rel_error(self, top['ivc_rwidth.room_width'], 227.277956, .01)
assert_rel_error(self,top['ivc_rlength.room_length'], 227.277904, .01)
assert_rel_error(self,top['d_deflection.deflection'], 720, .01)
assert_rel_error(self,top['d_bending.bending_stress_ratio'], 0.148863, .001)
assert_rel_error(self,top['d_shear.shear_stress_ratio'], 0.007985, .0001)
def test_scipy_optimizer_simple_paraboloid_unconstrained(self):
prob = Problem()
model = prob.model = Group()
model.add_subsystem('p1', IndepVarComp('x', 50.0), promotes=['*'])
model.add_subsystem('p2', IndepVarComp('y', 50.0), promotes=['*'])
model.add_subsystem('comp', Paraboloid(), promotes=['*'])
prob.set_solver_print(level=0)
prob.driver = ScipyOptimizer(optimizer='SLSQP', tol=1e-9, disp=False)
model.add_design_var('x', lower=-50.0, upper=50.0)
model.add_design_var('y', lower=-50.0, upper=50.0)
model.add_objective('f_xy')
prob.setup(check=False)
failed = prob.run_driver()
self.assertFalse(
failed,
"Optimization failed, result =\n" + str(prob.driver.result))
assert_rel_error(self, prob['x'], 6.66666667, 1e-6)
assert_rel_error(self, prob['y'], -7.3333333, 1e-6)
def test_paraboloid_optimize_constrained_explicit_infinite_bounds(self):
top = Problem()
root = top.root = Group()
root.add('p1', IndepVarComp('x', 3.0))
root.add('p2', IndepVarComp('y', -4.0))
root.add('p', ParaboloidOptCon())
# Constraint Equation
root.add('con', ExecComp('c = x-y'))
root.connect('p1.x', 'p.x')
root.connect('p2.y', 'p.y')
root.connect('p.x', 'con.x')
root.connect('p.y', 'con.y')
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'SLSQP'
top.driver.options['disp'] = False
top.driver.add_desvar('p1.x', lower=-inf_bound, upper=50)
top.driver.add_desvar('p2.y', lower=-50, upper=inf_bound)
top.driver.add_objective('p.f_xy')
top.driver.add_constraint('con.c', lower=15.0, upper=inf_bound)
top.setup(check=False)
top.run()
assert_rel_error(self, top['p.x'], 7.166667, 1e-6)
assert_rel_error(self, top['p.y'], -7.833333, 1e-6)
def test_feature_run_driver(self):
import numpy as np
from openmdao.api import Problem, NonlinearBlockGS, ScipyOptimizer
from openmdao.test_suite.components.sellar import SellarDerivatives
prob = Problem()
model = prob.model = SellarDerivatives()
model.nonlinear_solver = NonlinearBlockGS()
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
model.add_design_var('z', lower=np.array([-10.0, 0.0]), upper=np.array([10.0, 10.0]))
model.add_design_var('x', lower=0.0, upper=10.0)
model.add_objective('obj')
model.add_constraint('con1', upper=0.0)
model.add_constraint('con2', upper=0.0)
prob.setup()
prob.run_driver()
assert_rel_error(self, prob['x'], 0.0, 1e-5)
assert_rel_error(self, prob['y1'], 3.160000, 1e-2)
assert_rel_error(self, prob['y2'], 3.755278, 1e-2)
assert_rel_error(self, prob['z'], [1.977639, 0.000000], 1e-2)
assert_rel_error(self, prob['obj'], 3.18339395, 1e-2)
def test_unconstrainted(self):
from openmdao.api import Problem, ScipyOptimizer, IndepVarComp
# We'll use the component that was defined in the last tutorial
from openmdao.test_suite.components.paraboloid import Paraboloid
# build the model
prob = Problem()
indeps = prob.model.add_subsystem('indeps', IndepVarComp())
indeps.add_output('x', 3.0)
indeps.add_output('y', -4.0)
prob.model.add_subsystem('paraboloid', Paraboloid())
prob.model.connect('indeps.x', 'paraboloid.x')
prob.model.connect('indeps.y', 'paraboloid.y')
# setup the optimization
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'COBYLA'
prob.model.add_design_var('indeps.x', lower=-50, upper=50)
prob.model.add_design_var('indeps.y', lower=-50, upper=50)
prob.model.add_objective('paraboloid.f_xy')
prob.setup()
prob.run_driver()
# minimum value
assert_rel_error(self, prob['paraboloid.f_xy'], -27.33333, 1e-6)
# location of the minimum
assert_rel_error(self, prob['indeps.x'], 6.6667, 1e-4)
assert_rel_error(self, prob['indeps.y'], -7.33333, 1e-4)
def test_Sellar_SLSQP(self):
prob = Problem()
prob.root = SellarDerivatives()
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1.0e-8
prob.driver.add_desvar('z',
lower=np.array([-10.0, 0.0]),
upper=np.array([10.0, 10.0]))
prob.driver.add_desvar('x', lower=0.0, upper=10.0)
prob.driver.add_objective('obj')
prob.driver.add_constraint('con1', upper=0.0)
prob.driver.add_constraint('con2', upper=0.0)
prob.driver.options['disp'] = False
prob.setup(check=False)
prob.run()
assert_rel_error(self, prob['z'][0], 1.9776, 1e-3)
assert_rel_error(self, prob['z'][1], 0.0, 1e-3)
assert_rel_error(self, prob['x'], 0.0, 1e-3)
def test_simple_array_comp2D(self):
prob = Problem()
root = prob.root = Group()
root.add('p1', IndepVarComp('x', np.zeros((2, 2))), promotes=['*'])
root.add('comp', ArrayComp2D(), promotes=['*'])
root.add('con',
ExecComp('c = y - 20.0',
c=np.zeros((2, 2)),
y=np.zeros((2, 2))),
promotes=['*'])
root.add('obj',
ExecComp('o = y[0, 0]', y=np.zeros((2, 2))),
promotes=['*'])
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.add_desvar('x', lower=-50.0, upper=50.0)
prob.driver.add_objective('o')
prob.driver.add_constraint('c', equals=0.0)
prob.driver.options['disp'] = False
prob.setup(check=False)
prob.run()
obj = prob['o']
assert_rel_error(self, obj, 20.0, 1e-6)
def test_unsupported_equality(self):
prob = Problem()
model = prob.model = Group()
model.add_subsystem('p1', IndepVarComp('x', 50.0), promotes=['*'])
model.add_subsystem('p2', IndepVarComp('y', 50.0), promotes=['*'])
model.add_subsystem('comp', Paraboloid(), promotes=['*'])
model.add_subsystem('con', ExecComp('c = - x + y'), promotes=['*'])
prob.set_solver_print(level=0)
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'COBYLA'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = False
model.add_design_var('x', lower=-50.0, upper=50.0)
model.add_design_var('y', lower=-50.0, upper=50.0)
model.add_objective('f_xy')
model.add_constraint('c', equals=-15.0)
prob.setup(check=False)
with self.assertRaises(Exception) as raises_cm:
prob.run_driver()
exception = raises_cm.exception
msg = "Constraints of type 'eq' not handled by COBYLA."
self.assertEqual(exception.args[0], msg)
def test_beam_tutorial_viewtree(self):
top = Problem()
top.root = BeamTutorial()
top.driver = ScipyOptimizer()
top.driver.options['optimizer'] = 'SLSQP'
top.driver.options['tol'] = 1.0e-8
top.driver.options['maxiter'] = 10000 #maximum number of solver iterations
top.driver.options['disp'] = False
#room length and width bounds
top.driver.add_desvar('ivc_rlength.room_length', lower=5.0*12.0, upper=50.0*12.0) #domain: 1in <= length <= 50ft
top.driver.add_desvar('ivc_rwidth.room_width', lower=5.0*12.0, upper=30.0*12.0) #domain: 1in <= width <= 30ft
top.driver.add_objective('d_neg_area.neg_room_area') #minimize negative area (or maximize area)
top.driver.add_constraint('d_len_minus_wid.length_minus_width', lower=0.0) #room_length >= room_width
top.driver.add_constraint('d_deflection.deflection', lower=720.0) #deflection >= 720
top.driver.add_constraint('d_bending.bending_stress_ratio', upper=0.5) #bending < 0.5
top.driver.add_constraint('d_shear.shear_stress_ratio', upper=1.0/3.0) #shear < 1/3
top.setup(check=False)
from openmdao.api import view_tree
view_tree(top, show_browser=False)
import os.path
self.assertTrue(os.path.isfile('partition_tree_n2.html'))
os.remove('partition_tree_n2.html')
def test_simple_paraboloid_scaled_objective_rev(self):
prob = Problem()
model = prob.model = Group()
prob.set_solver_print(level=0)
model.add_subsystem('p1', IndepVarComp('x', 50.0), promotes=['*'])
model.add_subsystem('p2', IndepVarComp('y', 50.0), promotes=['*'])
model.add_subsystem('comp', Paraboloid(), promotes=['*'])
model.add_subsystem('con', ExecComp('c = x - y'), promotes=['*'])
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = False
model.add_design_var('x', lower=-50.0, upper=50.0)
model.add_design_var('y', lower=-50.0, upper=50.0)
model.add_objective('f_xy', ref=10.)
model.add_constraint('c', lower=10.0, upper=11.0)
prob.setup(check=False, mode='rev')
prob.run_driver()
# Minimum should be at (7.166667, -7.833334)
assert_rel_error(self, prob['x'] - prob['y'], 11.0, 1e-6)
def test(self):
import numpy as np
from openmdao.api import Problem, ScipyOptimizer
from openmdao.test_suite.test_examples.beam_optimization.beam_group import BeamGroup
E = 1.
L = 1.
b = 0.1
volume = 0.01
num_elements = 50
prob = Problem(model=BeamGroup(E=E, L=L, b=b, volume=volume, num_elements=num_elements))
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = True
prob.setup()
prob.run_driver()
print(prob['inputs_comp.h'])
def test_simple_paraboloid_equality(self):
prob = Problem()
model = prob.model = Group()
model.add_subsystem('p1', IndepVarComp('x', 50.0), promotes=['*'])
model.add_subsystem('p2', IndepVarComp('y', 50.0), promotes=['*'])
model.add_subsystem('comp', Paraboloid(), promotes=['*'])
model.add_subsystem('con', ExecComp('c = - x + y'), promotes=['*'])
prob.set_solver_print(level=0)
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = False
model.add_design_var('x', lower=-50.0, upper=50.0)
model.add_design_var('y', lower=-50.0, upper=50.0)
model.add_objective('f_xy')
model.add_constraint('c', equals=-15.0)
prob.setup(check=False)
prob.run_driver()
# Minimum should be at (7.166667, -7.833334)
# (Note, loose tol because of appveyor py3.4 machine.)
assert_rel_error(self, prob['x'], 7.16667, 1e-4)
assert_rel_error(self, prob['y'], -7.833334, 1e-4)
def test_simple_array_comp2D_dbl_sided_con_array(self):
prob = Problem()
model = prob.model = Group()
model.add_subsystem('p1',
IndepVarComp('widths', np.zeros((2, 2))),
promotes=['*'])
model.add_subsystem('comp', TestExplCompArrayDense(), promotes=['*'])
model.add_subsystem('obj',
ExecComp('o = areas[0, 0]', areas=np.zeros(
(2, 2))),
promotes=['*'])
prob.set_solver_print(level=0)
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = False
model.add_design_var('widths', lower=-50.0, upper=50.0)
model.add_objective('o')
model.add_constraint('areas', lower=20.0, upper=20.0)
prob.setup(check=False)
prob.run_driver()
obj = prob['o']
assert_rel_error(self, obj, 20.0, 1e-6)
def test_root_derivs_array(self):
prob = Problem()
prob.root = SellarDerivativesGrouped()
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1.0e-8
prob.driver.options['disp'] = False
prob.driver.add_desvar('z',
lower=np.array([-10.0, 0.0]),
upper=np.array([10.0, 10.0]))
prob.driver.add_desvar('x', lower=0.0, upper=10.0)
prob.driver.add_objective('obj')
prob.driver.add_constraint('con1', upper=0.0)
prob.driver.add_constraint('con2', upper=0.0)
prob.driver.add_recorder(self.recorder)
self.recorder.options['record_metadata'] = False
self.recorder.options['record_derivs'] = True
prob.setup(check=False)
prob.run()
prob.cleanup()
sout = open(self.filename)
lines = sout.readlines()
self.assertEqual(lines[14].rstrip(), 'Derivatives:')
self.assertTrue('9.61' in lines[15])
self.assertTrue('0.784' in lines[16])
self.assertTrue('1.077' in lines[17])
def test_simple_paraboloid_unconstrained(self):
prob = Problem()
model = prob.model = Group()
model.add_subsystem('p1', IndepVarComp('x', 50.0), promotes=['*'])
model.add_subsystem('p2', IndepVarComp('y', 50.0), promotes=['*'])
model.add_subsystem('comp', Paraboloid(), promotes=['*'])
prob.set_solver_print(level=0)
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = False
model.add_design_var('x', lower=-50.0, upper=50.0)
model.add_design_var('y', lower=-50.0, upper=50.0)
model.add_objective('f_xy')
prob.setup(check=False)
prob.run_driver()
assert_rel_error(self, prob['x'], 6.66666667, 1e-6)
assert_rel_error(self, prob['y'], -7.3333333, 1e-6)
def test_feature_basic(self):
from openmdao.api import Problem, Group, IndepVarComp, ScipyOptimizer
from openmdao.test_suite.components.paraboloid import Paraboloid
prob = Problem()
model = prob.model = Group()
model.add_subsystem('p1', IndepVarComp('x', 50.0), promotes=['*'])
model.add_subsystem('p2', IndepVarComp('y', 50.0), promotes=['*'])
model.add_subsystem('comp', Paraboloid(), promotes=['*'])
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1e-9
prob.driver.options['disp'] = True
model.add_design_var('x', lower=-50.0, upper=50.0)
model.add_design_var('y', lower=-50.0, upper=50.0)
model.add_objective('f_xy')
prob.setup()
prob.run_driver()
assert_rel_error(self, prob['x'], 6.66666667, 1e-6)
assert_rel_error(self, prob['y'], -7.3333333, 1e-6)
def test_driver_records_unknown_types_metadata(self):
prob = Problem()
root = prob.root = Group()
# Need an optimization problem to test to make sure
# the is_desvar, is_con, is_obj metadata is being
# recorded for the Unknowns
root.add('p1', IndepVarComp('x', 50.0))
root.add('p2', IndepVarComp('y', 50.0))
root.add('comp', Paraboloid())
root.connect('p1.x', 'comp.x')
root.connect('p2.y', 'comp.y')
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.add_desvar('p1.x', lower=-50.0, upper=50.0)
prob.driver.add_desvar('p2.y', lower=-50.0, upper=50.0)
prob.driver.add_objective('comp.f_xy')
prob.driver.options['disp'] = False
prob.driver.add_recorder(self.recorder)
self.recorder.options['record_metadata'] = True
prob.setup(check=False)
prob.cleanup() # close recorders
expected_params = list(iteritems(prob.root.params))
expected_unknowns = list(iteritems(prob.root.unknowns))
expected_resids = list(iteritems(prob.root.resids))
self.assertMetadataRecorded(
(expected_params, expected_unknowns, expected_resids))
def test_driver_param_indices_slsqp_force_fd(self):
""" Test driver param indices with ScipyOptimizer SLSQP and force_fd=True
"""
prob = Problem()
prob.root = SellarStateConnection()
prob.root.fd_options['force_fd'] = True
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1.0e-8
prob.driver.options['disp'] = False
prob.driver.add_desvar('z',
lower=np.array([-10.0]),
upper=np.array([10.0]),
indices=[0])
prob.driver.add_desvar('x', lower=0.0, upper=10.0)
prob.driver.add_objective('obj')
prob.driver.add_constraint('con1', upper=0.0)
prob.driver.add_constraint('con2', upper=0.0)
#prob.driver.options['disp'] = False
prob.setup(check=False)
prob['z'][1] = 0.0
prob.run()
assert_rel_error(self, prob['z'][0], 1.9776, 1e-3)
assert_rel_error(self, prob['z'][1], 0.0, 1e-3)
assert_rel_error(self, prob['x'], 0.0, 1e-3)
def test_simple_paraboloid_constrained_COBYLA_upper(self):
prob = Problem()
root = prob.root = Group()
root.add('p1', IndepVarComp('x', 50.0), promotes=['*'])
root.add('p2', IndepVarComp('y', 50.0), promotes=['*'])
root.add('comp', Paraboloid(), promotes=['*'])
root.add('con', ExecComp('c = y - x'), promotes=['*'])
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'COBYLA'
prob.driver.options['tol'] = 1.0e-8
prob.driver.add_desvar('x', lower=-50.0, upper=50.0)
prob.driver.add_desvar('y', lower=-50.0, upper=50.0)
prob.driver.add_objective('f_xy')
prob.driver.add_constraint('c', upper=-15.0)
prob.driver.options['disp'] = False
prob.setup(check=False)
prob.run()
# Minimum should be at (7.166667, -7.833334)
assert_rel_error(self, prob['x'], 7.16667, 1e-6)
assert_rel_error(self, prob['y'], -7.833334, 1e-6)
def test_Sellar_state_SLSQP(self):
""" Baseline Sellar test case without specifying indices.
"""
prob = Problem()
prob.root = SellarStateConnection()
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1.0e-8
prob.driver.add_desvar('z',
lower=np.array([-10.0, 0.0]),
upper=np.array([10.0, 10.0]))
prob.driver.add_desvar('x', lower=0.0, upper=10.0)
prob.driver.add_objective('obj')
prob.driver.add_constraint('con1', upper=0.0)
prob.driver.add_constraint('con2', upper=0.0)
prob.driver.options['disp'] = False
prob.setup(check=False)
prob.run()
assert_rel_error(self, prob['z'][0], 1.9776, 1e-3)
assert_rel_error(self, prob['z'][1], 0.0, 1e-3)
assert_rel_error(self, prob['x'], 0.0, 1e-3)
def test_sellar_opt(self):
from openmdao.api import Problem, ScipyOptimizer, ExecComp, IndepVarComp, DirectSolver
from openmdao.test_suite.components.sellar_feature import SellarMDA
prob = Problem()
prob.model = SellarMDA()
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
# prob.driver.options['maxiter'] = 100
prob.driver.options['tol'] = 1e-8
prob.model.add_design_var('x', lower=0, upper=10)
prob.model.add_design_var('z', lower=0, upper=10)
prob.model.add_objective('obj')
prob.model.add_constraint('con1', upper=0)
prob.model.add_constraint('con2', upper=0)
prob.setup()
prob.set_solver_print(level=0)
# Ask OpenMDAO to finite-difference across the model to compute the gradients for the optimizer
prob.model.approx_totals()
prob.run_driver()
print('minimum found at')
assert_rel_error(self, prob['x'][0], 0., 1e-5)
assert_rel_error(self, prob['z'], [1.977639, 0.], 1e-5)
print('minumum objective')
assert_rel_error(self, prob['obj'][0], 3.18339395045, 1e-5)
def test_simple_paraboloid_scaled_constraint_rev(self):
prob = Problem()
root = prob.root = Group()
root.add('p1', IndepVarComp('x', 50.0), promotes=['*'])
root.add('p2', IndepVarComp('y', 50.0), promotes=['*'])
root.add('comp', Paraboloid(), promotes=['*'])
root.add('con', ExecComp('c = x - y'), promotes=['*'])
prob.driver = ScipyOptimizer()
prob.driver.options['optimizer'] = 'SLSQP'
prob.driver.options['tol'] = 1.0e-8
prob.driver.options['disp'] = False
prob.driver.add_desvar('x', lower=-50.0, upper=50.0)
prob.driver.add_desvar('y', lower=-50.0, upper=50.0)
prob.driver.add_objective('f_xy')
prob.driver.add_constraint('c', lower=10.0, upper=11.0, scaler=1 / 10.)
root.ln_solver.options['mode'] = 'rev'
prob.setup(check=False)
prob.run()
# Minimum should be at (7.166667, -7.833334)
assert_rel_error(self, prob['x'] - prob['y'], 11.0, 1e-6)
def __init__(self, n=10, derivs=False):
super(BruteForceSellarProblem, self).__init__(impl=impl)
root = self.root = Group()
if not derivs:
root.deriv_options["type"] = "fd"
sellars = root.add("sellars", ParallelGroup())
for i in range(n):
name = "sellar%i" % i
sellars.add(name, SellarDerivatives())
root.connect("dist_x", "sellars." + name + ".x", src_indices=[i])
root.connect("dist_z", "sellars." + name + ".z", src_indices=[i * 2, i * 2 + 1])
root.connect("sellars." + name + ".obj", "collect.obj_%i" % i)
root.connect("sellars." + name + ".con1", "collect.con1_%i" % i)
root.connect("sellars." + name + ".con2", "collect.con2_%i" % i)
root.add("indep", IndepVarComp([("x", 1.0), ("z", np.array([5.0, 2.0]))]), promotes=["x", "z"])
root.add(
"random",
Randomize(
n=n,
params=[
# name, value, std dev
("x", 1.0, 1e-2),
("z", np.array([5.0, 2.0]), 1e-2),
],
),
promotes=["x", "z", "dist_x", "dist_z"],
)
root.add("collect", Collector(n=n, names=["obj", "con1", "con2"]), promotes=["obj", "con1", "con2"])
# top level driver setup
self.driver = ScipyOptimizer()
self.driver.options["optimizer"] = "SLSQP"
self.driver.options["tol"] = 1.0e-8
self.driver.options["maxiter"] = 50
self.driver.options["disp"] = False
self.driver.add_desvar("z", lower=np.array([-10.0, 0.0]), upper=np.array([10.0, 10.0]))
self.driver.add_desvar("x", lower=0.0, upper=10.0)
self.driver.add_objective("obj")
self.driver.add_constraint("con1", upper=0.0)
self.driver.add_constraint("con2", upper=0.0)
def __init__(self, n=10, derivs=False):
super(BruteForceSellarProblem, self).__init__(impl=impl)
root = self.root = Group()
if not derivs:
root.deriv_options['type'] = 'fd'
sellars = root.add('sellars', ParallelGroup())
for i in range(n):
name = 'sellar%i' % i
sellars.add(name, SellarDerivatives())
root.connect('dist_x', 'sellars.'+name+'.x', src_indices=[i])
root.connect('dist_z', 'sellars.'+name+'.z', src_indices=[i*2, i*2+1])
root.connect('sellars.'+name+'.obj', 'collect.obj_%i' % i)
root.connect('sellars.'+name+'.con1', 'collect.con1_%i' % i)
root.connect('sellars.'+name+'.con2', 'collect.con2_%i' % i)
root.add('indep', IndepVarComp([
('x', 1.0),
('z', np.array([5.0, 2.0]))
]),
promotes=['x', 'z'])
root.add('random', Randomize(n=n, params=[
# name, value, std dev
('x', 1.0, 1e-2),
('z', np.array([5.0, 2.0]), 1e-2)
]),
promotes=['x', 'z', 'dist_x', 'dist_z'])
root.add('collect', Collector(n=n, names=['obj', 'con1', 'con2']),
promotes=['obj', 'con1', 'con2'])
# top level driver setup
self.driver = ScipyOptimizer()
self.driver.options['optimizer'] = 'SLSQP'
self.driver.options['tol'] = 1.0e-8
self.driver.options['maxiter'] = 50
self.driver.options['disp'] = False
self.driver.add_desvar('z', lower=np.array([-10.0, 0.0]),
upper=np.array([ 10.0, 10.0]))
self.driver.add_desvar('x', lower=0.0, upper=10.0)
self.driver.add_objective('obj')
self.driver.add_constraint('con1', upper=0.0)
self.driver.add_constraint('con2', upper=0.0)
def setUp(self):
solver = 'SLSQP'
num_seg = 10
seg_ncn = 2
rel_lengths = 'lgl'
# Instantiate a problem and set it's root to an empty Trajectory
prob = Problem()
prob.add_traj(Trajectory("traj0"))
if solver == 'SNOPT' and pyOptSparseDriver is not None:
driver = pyOptSparseDriver()
driver.options['optimizer'] = solver
driver.opt_settings['Major iterations limit'] = 1000
driver.opt_settings['iSumm'] = 6
driver.opt_settings['Major step limit'] = 0.5
driver.opt_settings["Major feasibility tolerance"] = 1.0E-5
driver.opt_settings["Major optimality tolerance"] = 1.0E-5
driver.opt_settings["Minor feasibility tolerance"] = 1.0E-4
driver.opt_settings['Verify level'] = 3
else:
driver = ScipyOptimizer()
driver.options['tol'] = 1.0E-6
driver.options['disp'] = True
driver.options['maxiter'] = 500
prob.driver = driver
dynamic_controls = [{'name': 'theta', 'units': 'rad'},
{'name': 'psi', 'units': 'rad'}]
static_controls = [{'name': 'mass', 'units': 'kg'},
{'name': 'g', 'units': 'm/s/s'},
{'name': 'Cd', 'units': 'unitless'},
{'name': 'D_magnetic', 'units': 'N'},
{'name': 'R', 'units': 'J/(kg*K)'},
{'name': 'S', 'units': 'm**2'},
{'name': 'T_ambient', 'units': 'K'},
{'name': 'p_tube', 'units': 'Pa'}]
phase0 = CollocationPhase(name='phase0', rhs_class=MagnePlaneRHS,
num_seg=num_seg, seg_ncn=seg_ncn,
rel_lengths=rel_lengths,
dynamic_controls=dynamic_controls,
static_controls=static_controls)
prob.trajectories["traj0"].add_phase(phase0)
phase0.set_state_options('x', lower=0, upper=100,
ic_val=0, ic_fix=True,
fc_val=10, fc_fix=True,
scaler=10.0, defect_scaler=0.1)
phase0.set_state_options('y', lower=0, upper=0,
ic_val=0, ic_fix=True,
fc_val=0, fc_fix=True,
scaler=10.0, defect_scaler=0.1)
phase0.set_state_options('z', lower=-10, upper=10,
ic_val=-10, ic_fix=True,
fc_val=-5, fc_fix=True,
scaler=10.0, defect_scaler=0.1)
phase0.set_state_options('v', lower=0, upper=np.inf,
ic_val=0.0, ic_fix=True,
fc_val=10.0, fc_fix=False,
scaler=10.0, defect_scaler=0.1)
phase0.set_dynamic_control_options('theta', ic_val=-1.57, fc_val=0.2,
opt=True, lower=-1.58, upper=1.58,
scaler=0.01)
phase0.set_dynamic_control_options('psi', ic_val=0.0, fc_val=0.0,
opt=True, lower=0.0, upper=0.0,
scaler=0.01)
phase0.set_static_control_options(name='g', val=9.80665, opt=False)
phase0.set_static_control_options(name='mass', val=1200, opt=False)
phase0.set_static_control_options(name='Cd', val=0.0, opt=False)
phase0.set_static_control_options(name='S', val=0.0, opt=False)
phase0.set_static_control_options(name='p_tube', val=0.0, opt=False)
phase0.set_static_control_options(name='T_ambient', val=298.0,
opt=False)
phase0.set_static_control_options(name='R', val=287.0, opt=False)
# Set D_magnetic equal to the thrust value for now, no net forces
# on the pod except gravity.
phase0.set_static_control_options(name='D_magnetic', val=30000.0,
opt=False)
phase0.set_time_options(t0_val=0, t0_lower=0, t0_upper=0, tp_val=1.81,
tp_lower=0.5, tp_upper=10.0)
prob.trajectories["traj0"].add_objective(name="t", phase="phase0",
place="end", scaler=1.0)
# Do top-level FD
prob.root.deriv_options['type'] = 'fd'
self.prob = prob
class BruteForceSellarProblem(Problem):
""" Performs optimization on the Sellar problem.
Applies a normal distribution to the design vars and runs all of the
samples, then collects the values of all of the outputs, calculates
the mean of those and stuffs that back into the unknowns vector.
This is the brute force version that just stamps out N separate
sellar models in a parallel group and sets the input of each
one to be one of these random design vars.
Args
----
n : number of randomized points to generate for each input value
derivs : if True, use user-defined derivatives, else use Finite Difference
"""
def __init__(self, n=10, derivs=False):
super(BruteForceSellarProblem, self).__init__(impl=impl)
root = self.root = Group()
if not derivs:
root.deriv_options['type'] = 'fd'
sellars = root.add('sellars', ParallelGroup())
for i in range(n):
name = 'sellar%i' % i
sellars.add(name, SellarDerivatives())
root.connect('dist_x', 'sellars.'+name+'.x', src_indices=[i])
root.connect('dist_z', 'sellars.'+name+'.z', src_indices=[i*2, i*2+1])
root.connect('sellars.'+name+'.obj', 'collect.obj_%i' % i)
root.connect('sellars.'+name+'.con1', 'collect.con1_%i' % i)
root.connect('sellars.'+name+'.con2', 'collect.con2_%i' % i)
root.add('indep', IndepVarComp([
('x', 1.0),
('z', np.array([5.0, 2.0]))
]),
promotes=['x', 'z'])
root.add('random', Randomize(n=n, params=[
# name, value, std dev
('x', 1.0, 1e-2),
('z', np.array([5.0, 2.0]), 1e-2)
]),
promotes=['x', 'z', 'dist_x', 'dist_z'])
root.add('collect', Collector(n=n, names=['obj', 'con1', 'con2']),
promotes=['obj', 'con1', 'con2'])
# top level driver setup
self.driver = ScipyOptimizer()
self.driver.options['optimizer'] = 'SLSQP'
self.driver.options['tol'] = 1.0e-8
self.driver.options['maxiter'] = 50
self.driver.options['disp'] = False
self.driver.add_desvar('z', lower=np.array([-10.0, 0.0]),
upper=np.array([ 10.0, 10.0]))
self.driver.add_desvar('x', lower=0.0, upper=10.0)
self.driver.add_objective('obj')
self.driver.add_constraint('con1', upper=0.0)
self.driver.add_constraint('con2', upper=0.0)
class BruteForceSellarProblem(Problem):
""" Performs optimization on the Sellar problem.
Applies a normal distribution to the design vars and runs all of the
samples, then collects the values of all of the outputs, calculates
the mean of those and stuffs that back into the unknowns vector.
This is the brute force version that just stamps out N separate
sellar models in a parallel group and sets the input of each
one to be one of these random design vars.
Args
----
n : number of randomized points to generate for each input value
derivs : if True, use user-defined derivatives, else use Finite Difference
"""
def __init__(self, n=10, derivs=False):
super(BruteForceSellarProblem, self).__init__(impl=impl)
root = self.root = Group()
if not derivs:
root.deriv_options["type"] = "fd"
sellars = root.add("sellars", ParallelGroup())
for i in range(n):
name = "sellar%i" % i
sellars.add(name, SellarDerivatives())
root.connect("dist_x", "sellars." + name + ".x", src_indices=[i])
root.connect("dist_z", "sellars." + name + ".z", src_indices=[i * 2, i * 2 + 1])
root.connect("sellars." + name + ".obj", "collect.obj_%i" % i)
root.connect("sellars." + name + ".con1", "collect.con1_%i" % i)
root.connect("sellars." + name + ".con2", "collect.con2_%i" % i)
root.add("indep", IndepVarComp([("x", 1.0), ("z", np.array([5.0, 2.0]))]), promotes=["x", "z"])
root.add(
"random",
Randomize(
n=n,
params=[
# name, value, std dev
("x", 1.0, 1e-2),
("z", np.array([5.0, 2.0]), 1e-2),
],
),
promotes=["x", "z", "dist_x", "dist_z"],
)
root.add("collect", Collector(n=n, names=["obj", "con1", "con2"]), promotes=["obj", "con1", "con2"])
# top level driver setup
self.driver = ScipyOptimizer()
self.driver.options["optimizer"] = "SLSQP"
self.driver.options["tol"] = 1.0e-8
self.driver.options["maxiter"] = 50
self.driver.options["disp"] = False
self.driver.add_desvar("z", lower=np.array([-10.0, 0.0]), upper=np.array([10.0, 10.0]))
self.driver.add_desvar("x", lower=0.0, upper=10.0)
self.driver.add_objective("obj")
self.driver.add_constraint("con1", upper=0.0)
self.driver.add_constraint("con2", upper=0.0)
