def __init__(self, nlpy_model, **kwargs):
"""
:parameters:
:nlpy_model: nonlinear problem from the NLPModel class or
from AmplModel class
"""
if nlpy_model.__module__ == 'nlpy.model.amplpy':
print 'AMPL model'
print "AMPL isn't handling complex values"
print "Choose 'FD' or 'opt_prob.grad_func' for sensitivity"
# Initialize model.
Optimization.__init__(self, nlpy_model.name, lambda x:
(self.nlpy_model.obj(x),
self.nlpy_model.cons(x).tolist(),0),
var_set={}, obj_set={}, con_set={},
use_groups=False, **kwargs)
self.nlpy_model = nlpy_model
self.addObj('f')
# Assigning lower and upper bounds on variables
for i in range(0,self.nlpy_model.n):
if i in self.nlpy_model.lowerB:
self.addVar("x"+ "%d"%(i+1), lower=self.nlpy_model.Lvar[i],
upper=numpy.inf, value=self.nlpy_model.x0[i])
elif i in self.nlpy_model.upperB:
self.addVar("x"+ "%d"%(i+1), lower=-numpy.inf,
upper=self.nlpy_model.Uvar[i], value=self.nlpy_model.x0[i])
elif i in self.nlpy_model.rangeB:
self.addVar("x"+ "%d"%(i+1), lower=self.nlpy_model.Lvar[i],
upper=self.nlpy_model.Uvar[i], value=self.nlpy_model.x0[i])
elif i in self.nlpy_model.freeB:
self.addVar("x"+ "%d"%(i+1), value=self.nlpy_model.x0[i],
lower=-numpy.inf, upper=numpy.inf)
# Assigning lower and upper bounds on constraints
for i in range(0,nlpy_model.m):
if i in nlpy_model.lowerC:
self.addCon("g"+"%d"%(i+1), 'i', lower=nlpy_model.Lcon[i],
upper=numpy.inf)
elif i in nlpy_model.upperC:
self.addCon("g"+"%d"%(i+1), 'i', lower=-numpy.inf,
upper=nlpy_model.Ucon[i])
elif i in nlpy_model.rangeC:
self.addCon("g"+"%d"%(i+1), 'i', lower=nlpy_model.Lcon[i],
upper=nlpy_model.Ucon[i])
elif i in nlpy_model.equalC:
self.addCon("g"+"%d"%(i+1), 'e', equal=nlpy_model.Lcon[i])
