opt_prob = Optimization(args.logfile, optproblem.evalObjCon)
# Add functions
opt_prob.addObj('weight')
opt_prob.addCon('buckling-bar1', type='i')
opt_prob.addCon('failure-bar1', type='i')
opt_prob.addCon('failure-bar2', type='i')
# Add variables
opt_prob.addVar('area-1', type='c', value=1.0, lower=1.0e-3, upper=2.0)
opt_prob.addVar('area-2', type='c', value=1.0, lower=1.0e-3, upper=2.0)
opt_prob.addVar('height', type='c', value=4.0, lower=4.0, upper=10.0)
# Optimization algorithm
if args.algorithm == 'ALGENCAN':
opt = ALGENCAN()
opt.setOption('iprint', 2)
opt.setOption('epsfeas', 1e-6)
opt.setOption('epsopt', 1e-6)
else:
opt = SLSQP(pll_type='POA')
opt.setOption('MAXIT', 999)
opt(opt_prob,
sens_type=optproblem.evalObjConGradient,
disp_opts=True,
store_hst=True,
hot_start=False)
if optproblem.comm.Get_rank() == 0:
print opt_prob.solution(0)
jac = np.zeros([self.m, self.n])
jac[0][0] = 0.
jac[0][1] = 2.
jac[1][0] = 3.
jac[1][1] = 0.
return jac
# Translate this NLPy problem in a pyOpt problem
nlp = example()
opt_prob = PyOpt_From_NLPModel(nlp)
print opt_prob
# Instantiate Optimizer (ALGENCAN) & Solve Problem
algencan = ALGENCAN()
algencan.setOption('iprint', 0)
algencan(opt_prob)
print opt_prob.solution(0)
# Call the imported solver SNOPT
snopt = SNOPT()
# Choose sensitivity type for computing gradients with :
# opt_prob.grad_func : NLP.py is computing the gradients through the grad_func
# function
# 'FD' : finite difference
# 'CS' : complex step
SensType = opt_prob.grad_func
# Instantiate Optimizer (SOLVOPT) & Solve Problem
solvopt = SOLVOPT()
solvopt.setOption('iprint', -1)
solvopt(opt_prob, sens_type='FD')
print(opt_prob.solution(4))
# Instantiate Optimizer (KSOPT) & Solve Problem
ksopt = KSOPT()
ksopt.setOption('IPRINT', 0)
ksopt(opt_prob, sens_type='FD')
print(opt_prob.solution(5))
# Instantiate Optimizer (NSGA2) & Solve Problem
nsga2 = NSGA2()
nsga2.setOption('PrintOut', 0)
nsga2(opt_prob)
print(opt_prob.solution(6))
# Instantiate Optimizer (ALGENCAN) & Solve Problem
algencan = ALGENCAN()
algencan.setOption('iprint', 0)
algencan(opt_prob)
print(opt_prob.solution(7))
# Instantiate Optimizer (FILTERSD) & Solve Problem
filtersd = FILTERSD()
filtersd.setOption('iprint', 0)
filtersd(opt_prob)
print(opt_prob.solution(8))
def optimize(k, w1, w2):
# Physical problem
rho = 0.2836 # lb/in^3
L = 5.0 # in
P = 25000.0 # lb
E = 30.0e6 # psi
ys = 36260.0 # psi
fs = 1.5
dtruss = TwoBarTruss(rho, L, P, E, ys, fs)
struss = StochasticTwoBarTruss(dtruss)
# Optimization Problem
optproblem = TwoBarTrussOpt(MPI.COMM_WORLD, struss, k, w1, w2)
opt_prob = Optimization(args.logfile, optproblem.evalObjCon)
# Add functions
opt_prob.addObj('weight')
opt_prob.addCon('buckling-bar1', type='i')
opt_prob.addCon('failure-bar1' , type='i')
opt_prob.addCon('failure-bar2' , type='i')
# Add variables
opt_prob.addVar('area-1', type='c', value= 1.5, lower= 1.5, upper= 1.5)
opt_prob.addVar('area-2', type='c', value= 1.5, lower= 1.5, upper= 1.5)
opt_prob.addVar('height', type='c', value= 4.0, lower= 4.0, upper= 10.0)
# Optimization algorithm
if args.algorithm == 'ALGENCAN':
opt = ALGENCAN()
opt.setOption('iprint',2)
opt.setOption('epsfeas',1e-6)
opt.setOption('epsopt',1e-6)
else:
opt = SLSQP(pll_type='POA')
opt.setOption('MAXIT',999)
opt(opt_prob,
sens_type=optproblem.evalObjConGradient,
disp_opts=True,
store_hst=True,
hot_start=False)
if optproblem.comm.Get_rank() ==0:
print opt_prob.solution(0)
opt_prob.write2file(disp_sols=True)
x = optproblem.x_hist[-1]
f = optproblem.fvals[0]
print 'x', x
print 'f', f
return x, f