単純な抽象モデルを作成しましたが、事前に解決したいと考えていました。これまでのモデルは次のとおりです。
from __future__ import division
from pyomo.environ import *
from pyomo.opt import SolverFactory
model=AbstractModel()
opt=SolverFactory('ipopt')
#Sets
model.i=Set(initialize=['conv','effic','hybrid','elec'],ordered=True,doc='car types')
model.iter=RangeSet(0,1000)
model.ir=Set(within=model.iter,ordered=True)
model.j=Set(initialize=model.i)
#Parameter initialize
def lamda_init(model):
return model.lamdalo
def mu_init(model):
return model.mulo
def shares_init(model,i):
return model.base[i]
def cost_init(model,i):
return model.pr[i]+ model.lamda*model.em[i]+ model.mu(1-model.type[i])
#Parameters
model.pr=Param(model.i,initialize= {'conv':12,'effic':15,'hybrid':25,'elec':45},mutable=True,doc='levelized cost of car EUR per 100km')
model.base=Param(model.i,initialize= {'conv':0.80,'effic':0.07,'hybrid':0.10,'elec':0.03},mutable=True,doc='shares of car types in base year')
model.em=Param(model.i,initialize={'conv':10,'effic':5,'hybrid':1,'elec':0},mutable=True,doc='emissions kgCO2 per 100km')
model.type=Param(model.i,initialize={'conv':0,'effic':0,'hybrid':0,'elec':1},mutable=True,doc='compliance with promoted car type')
model.gamma=Param(initialize=-2.5,mutable=True,doc='exponent of discrete choice')
model.lamdalo=Param(initialize=0,mutable=True,doc='min dual value of emission constraint measured as additional cost')
model.lamda=Param(initialize=lamda_init,mutable=True,doc='dual value of emission constraint measured as additional cost')
model.lamdaup=Param(initialize=1000,mutable=True,doc='max dual value of emission constraint measured as additional cost')
model.mulo=Param(initialize=0,mutable=True,doc='dual value of desired car type measured as additional cost')
model.mu=Param(initialize=mu_init,mutable=True,doc='dual value of desired car type measured as additional cost')
model.muup=Param(initialize=1000,mutable=True,doc='dual value of desired car type measured as additional cost')
model.CO2goal=Param(initialize=0.9,mutable=True,doc='target of emissions')
model.cargoal=Param(initialize=0.50,mutable=True,doc='target of share of promoted car type')
model.CO2perf=Param(mutable=True,doc='performance for emissions')
model.carperf=Param(mutable=True,doc='performance in terms of share of promoted car type')
model.count=Param(mutable=True,doc='counter of model runs -iterations')
model.count1=Param(mutable=True)
model.count2=Param(mutable=True)
model.conv=Param(mutable=True,doc='convergence criterion')
model.tolerance=Param(initialize=10**-3,mutable=True,doc='tolerance of convergence')
model.maxiter=Param(initialize=200,mutable=True)
model.report=dict()
#Variables
model.shares=Var(model.i,initialize=shares_init,within=NonNegativeReals,doc='shares of technologies')
model.cost=Var(model.i,initialize=cost_init,within=NonNegativeReals,doc='generalized unit cost')
def eqobj_rule(model):
return 0
model.obj=Objective(rule=eqobj_rule,doc='dummy objective')
def eq1_rule(model,i):
return model.shares[i]==model.base[i]*model.cost[i]**model.gamma/sum(model.base[j]*model.cost[j]**model.gamma for j in model.j)
model.eq1=Constraint(model.i,rule=eq1_rule)
def eq2_rule(model,i):
return model.cost[i]==model.pr[i]+model.lamda*model.em[i]+model.mu*(1-model.type[i])
model.eq2=Constraint(model.i,rule=eq2_rule)
model.count=0
model.conv=1000
instance=model.create_instance()
for i in instance.iter:
if i==instance.count:
instance.ir.add(i)
else:
pass
results= opt.solve(instance)
しかし、実行すると次のエラーが発生します。
TypeError Traceback (most recent call last)
<ipython-input-92-5dc9cea3a41f> in <module>()
75 display(instance)
76
---> 77 results= opt.solve(instance)
78
79 def twocarpolicies_sub(model):
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\opt\base\solvers.pyc in solve(self, *args, **kwds)
579 initial_time = time.time()
580
--> 581 self._presolve(*args, **kwds)
582
583 presolve_completion_time = time.time()
C:\Users\irini\Anaconda2\lib\site- packages\pyomo\solvers\plugins\solvers\ASL.pyc in _presolve(self, *args, **kwds)
176 self._instance = args[0]
177 xfrm = TransformationFactory('mpec.nl')
--> 178 xfrm.apply_to(self._instance)
179 if len(self._instance._transformation_data['mpec.nl'].compl_cuids) == 0:
180 # There were no complementarity conditions
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\core\base\plugin.pyc in apply_to(self, model, **kwds)
328 if not hasattr(model, '_transformation_data'):
329 model._transformation_data = TransformationData()
--> 330 self._apply_to(model, **kwds)
331
332 def create_using(self, model, **kwds):
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\mpec\plugins\mpec4.pyc in _apply_to(self, instance, **kwds)
42 free_vars = {}
43 id_list = []
---> 44 for vdata in instance.component_data_objects(Var, active=True, sort=SortComponents.deterministic):
45 id_list.append( id(vdata) )
46 free_vars[id(vdata)] = vdata
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\core\base\block.pyc in component_data_objects(self, ctype, active, sort, descend_into, descent_order)
1198 sort=sort,
1199 descend_into=descend_into,
-> 1200 descent_order=descent_order)
1201 else:
1202 block_generator = (self,)
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\core\base\block.pyc in block_data_objects(self, active, sort, descend_into, descent_order)
1269 active=active,
1270 sort=sort,
-> 1271 traversal=descent_order)
1272
1273 def _tree_iterator(self,
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\core\base\block.pyc in _tree_iterator(self, ctype, active, sort, traversal)
1289 if active is not None and self.active != active:
1290 return ().__iter__()
-> 1291 if self.parent_component().type() not in ctype:
1292 return ().__iter__()
1293
C:\Users\irini\Anaconda2\lib\site-packages\pyomo\core\base\param.pyc in __call__(self, exception)
927 if exception:
928 msg = 'Cannot compute the value of an array of parameters'
--> 929 raise TypeError(msg)
930
931 register_component(Param, "Parameter data that is used to define a model instance.")
TypeError: Cannot compute the value of an array of parameters
モデルのどの部分を参照しているのかわかりません。それがなぜなのか誰か知っていますか？