#!/usr/local/bin/python ''' pyMMA - A Python pyOpt interface to MMA. Copyright (c) 2008-2011 by pyOpt Developers All rights reserved. Revision: 1.4 $Date: 21/06/2010 21:00$ Tested on: --------- Linux with g77 Linux with gfortran Linux with pathf95 Win32 with g77 Mac with g95 Developers: ----------- - Mr. Andrew Lambe (AL) - Dr. Ruben E. Perez (RP) - Mr. Peter Jansen (PJ) History ------- v. 1.0 - Initial Class Creation (AL, 2005) v. 1.1 - Extensive Functionality Updates (RP, 2008) v. 1.2 - Migrate to pyOpt Framework (RP, 2008) v. 1.3 - History support (PJ,RP, 2010) v. 1.4 - Gradient Class Support (PJ,RP, 2010) ''' __version__ = '$Revision: $' ''' To Do: - ''' # ============================================================================= # MMA Library # ============================================================================= try: import mma except: raise ImportError('MMA shared library failed to import') #end # ============================================================================= # Standard Python modules # ============================================================================= import os, sys import copy, time # ============================================================================= # External Python modules # ============================================================================= import numpy # ============================================================================= # Extension modules # ============================================================================= from pyOpt import Optimizer from pyOpt import History from pyOpt import Gradient # ============================================================================= # Misc Definitions # ============================================================================= inf = 10.E+20 # define a value for infinity # ============================================================================= eps = 1.0 # define a value for machine precision while ((eps/2.0 + 1.0) > 1.0): eps = eps/2.0 #end eps = 2.0*eps #eps = math.ldexp(1,-52) # ============================================================================= # MMA Optimizer Class # ============================================================================= class MMA(Optimizer): ''' MMA Optimizer Class - Inherited from Optimizer Abstract Class ''' def __init__(self, pll_type=None, *args, **kwargs): ''' MMA Optimizer Class Initialization **Keyword arguments:** - pll_type -> STR: Parallel Implementation (None, 'POA'-Parallel Objective Analysis), *Default* = None Documentation last updated: Feb. 16, 2010 - Peter W. Jansen ''' # if (pll_type == None): self.poa = False elif (pll_type.upper() == 'POA'): self.poa = True else: raise ValueError("pll_type must be either None or 'POA'") #end # name = 'MMA' category = 'Local Optimizer' def_opts = { # MMA Options 'MAXIT':[int,1000], # Maximum Iterations 'GEPS':[float,1e-6], # Dual Objective Gradient Tolerance 'DABOBJ':[float,1e-6], # 'DELOBJ':[float,1e-6], # 'ITRM':[int,2], # 'IPRINT':[int,1], # Output Level (<0 - None, 0 - Screen, 1 - File) 'IOUT':[int,6], # Output Unit Number 'IFILE':[str,'MMA.out'], # Output File Name } informs = { 0 : 'The optimality conditions are satisfied.', 1 : 'The algorithm has been stopped after MAXIT iterations.', } Optimizer.__init__(self, name, category, def_opts, informs, *args, **kwargs) def __solve__(self, opt_problem={}, sens_type='FD', store_sol=True, disp_opts=False, store_hst=False, hot_start=False, sens_mode='', sens_step={}, *args, **kwargs): ''' Run Optimizer (Optimize Routine) **Keyword arguments:** - opt_problem -> INST: Optimization instance - sens_type -> STR/FUNC: Gradient type, *Default* = 'FD' - store_sol -> BOOL: Store solution in Optimization class flag, *Default* = True - disp_opts -> BOOL: Flag to display options in solution text, *Default* = False - store_hst -> BOOL/STR: Flag/filename to store optimization history, *Default* = False - hot_start -> BOOL/STR: Flag/filename to read optimization history, *Default* = False - sens_mode -> STR: Flag for parallel gradient calculation, *Default* = '' - sens_step -> FLOAT: Sensitivity setp size, *Default* = {} [corresponds to 1e-6 (FD), 1e-20(CS)] Additional arguments and keyword arguments are passed to the objective function call. Documentation last updated: February. 2, 2011 - Peter W. Jansen ''' # if ((self.poa) and (sens_mode.lower() == 'pgc')): raise NotImplementedError("pyMMA - Current implementation only allows single level parallelization, either 'POA' or 'pgc'") #end if self.poa or (sens_mode.lower() == 'pgc'): try: import mpi4py from mpi4py import MPI except ImportError: print 'pyMMA: Parallel objective Function Analysis requires mpi4py' #end comm = MPI.COMM_WORLD nproc = comm.Get_size() if (mpi4py.__version__[0] == '0'): Bcast = comm.Bcast elif (mpi4py.__version__[0] == '1'): Bcast = comm.bcast #end self.pll = True self.myrank = comm.Get_rank() else: self.pll = False self.myrank = 0 #end myrank = self.myrank # tmp_file = False def_fname = self.options['IFILE'][1].split('.')[0] if isinstance(store_hst,str): if isinstance(hot_start,str): if (myrank == 0): if (store_hst == hot_start): hos_file = History(hot_start, 'r', self) log_file = History(store_hst+'_tmp', 'w', self, opt_problem.name) tmp_file = True else: hos_file = History(hot_start, 'r', self) log_file = History(store_hst, 'w', self, opt_problem.name) #end #end self.sto_hst = True self.h_start = True elif hot_start: if (myrank == 0): hos_file = History(store_hst, 'r', self) log_file = History(store_hst+'_tmp', 'w', self, opt_problem.name) tmp_file = True #end self.sto_hst = True self.h_start = True else: if (myrank == 0): log_file = History(store_hst, 'w', self, opt_problem.name) #end self.sto_hst = True self.h_start = False #end elif store_hst: if isinstance(hot_start,str): if (hot_start == def_fname): if (myrank == 0): hos_file = History(hot_start, 'r', self) log_file = History(def_fname+'_tmp', 'w', self, opt_problem.name) tmp_file = True #end else: if (myrank == 0): hos_file = History(hot_start, 'r', self) log_file = History(def_fname, 'w', self, opt_problem.name) #end #end self.sto_hst = True self.h_start = True elif hot_start: if (myrank == 0): hos_file = History(def_fname, 'r', self) log_file = History(def_fname+'_tmp', 'w', self, opt_problem.name) tmp_file = True #end self.sto_hst = True self.h_start = True else: if (myrank == 0): log_file = History(def_fname, 'w', self, opt_problem.name) #end self.sto_hst = True self.h_start = False #end else: self.sto_hst = False self.h_start = False #end # gradient = Gradient(opt_problem, sens_type, sens_mode, sens_step, *args, **kwargs) #====================================================================== # MMA - Objective/Constraint Values and Gradients Function #====================================================================== def func(m,n,xval,f0val,df0dx,fval,dfdx): # Variables Groups Handling if opt_problem.use_groups: xg = {} for group in group_ids.keys(): if (group_ids[group][1]-group_ids[group][0] == 1): xg[group] = xval[group_ids[group][0]] else: xg[group] = xval[group_ids[group][0]:group_ids[group][1]] #end #end xn = xg else: xn = xval #end # Flush Output Files self.flushFiles() # Evaluate User Function (Real Valued) ifail = 0 if (myrank == 0): if self.h_start: [vals,hist_end] = hos_file.read(ident=['obj', 'con', 'fail']) if hist_end: self.h_start = False hos_file.close() else: [f,g,fail] = [vals['obj'][0][0],vals['con'][0],int(vals['fail'][0][0])] #end #end #end if self.pll: self.h_start = Bcast(self.h_start,root=0) #end if self.h_start and self.pll: [f,g,fail] = Bcast([f,g,fail],root=0) elif not self.h_start: [f,g,fail] = opt_problem.obj_fun(xn, *args, **kwargs) #end # Store History if (myrank == 0): if self.sto_hst: log_file.write(xval,'x') log_file.write(f,'obj') log_file.write(g,'con') log_file.write(fail,'fail') #end #end # Gradients if self.h_start: if (myrank == 0): [vals,hist_end] = hos_file.read(ident=['grad_obj','grad_con']) if hist_end: self.h_start = False hos_file.close() else: df = vals['grad_obj'][0].reshape((len(opt_problem._objectives.keys()),len(opt_problem._variables.keys()))) dg = vals['grad_con'][0].reshape((len(opt_problem._constraints.keys()),len(opt_problem._variables.keys()))) #end #end if self.pll: self.h_start = Bcast(self.h_start,root=0) #end if self.h_start and self.pll: [df,dg] = Bcast([df,dg],root=0) #end #end if not self.h_start: # df,dg = gradient.getGrad(xval, group_ids, [f], g, *args, **kwargs) #end # Store History if self.sto_hst and (myrank == 0): log_file.write(df,'grad_obj') log_file.write(dg,'grad_con') #end # Objective Assigment if isinstance(f,complex): f0val = f.astype(float) else: f0val = f #end # Constraints Assigment for i in xrange(len(opt_problem._constraints.keys())): if isinstance(g[i],complex): fval[i] = g[i].astype(float) else: fval[i] = g[i] #end #end # Gradients Assigment k = 0 for i in xrange(len(opt_problem._variables.keys())): if isinstance(df[0,i],complex): df0dx[i] = df[0,i].astype(float) else: df0dx[i] = df[0,i] #end for jj in xrange(len(opt_problem._constraints.keys())): if isinstance(dg[jj,i],complex): dfdx[k] = dg[jj,i].astype(float) else: dfdx[k] = dg[jj,i] #end k += 1 #end #end return f0val,df0dx,fval,dfdx # Variables Handling n = len(opt_problem._variables.keys()) xmin = [] xmax = [] xval = [] for key in opt_problem._variables.keys(): if (opt_problem._variables[key].type == 'c'): xmin.append(opt_problem._variables[key].lower) xmax.append(opt_problem._variables[key].upper) xval.append(opt_problem._variables[key].value) elif (opt_problem._variables[key].type == 'i'): raise IOError('MMA cannot handle integer design variables') elif (opt_problem._variables[key].type == 'd'): raise IOError('MMA cannot handle discrete design variables') #end #end xmin = numpy.array(xmin) xmax = numpy.array(xmax) xval = numpy.array(xval) # Variables Groups Handling group_ids = {} if opt_problem.use_groups: k = 0 for key in opt_problem._vargroups.keys(): group_len = len(opt_problem._vargroups[key]['ids']) group_ids[opt_problem._vargroups[key]['name']] = [k,k+group_len] k += group_len #end #end # Constraints Handling m = len(opt_problem._constraints.keys()) neqc = 0 #fval = [] fmax = [] if m > 0: for key in opt_problem._constraints.keys(): if opt_problem._constraints[key].type == 'e': raise IOError('MMA cannot handle equality constraints') #neqc += 1 #end #fval.append(opt_problem._constraints[key].value) fmax.append(opt_problem._constraints[key].upper) #end #end #fval = numpy.array(fval) fmax = numpy.array(fmax) # Objective Handling objfunc = opt_problem.obj_fun nobj = len(opt_problem._objectives.keys()) f0val = [] for key in opt_problem._objectives.keys(): f0val.append(opt_problem._objectives[key].value) #end f0val = numpy.array(f0val) # Setup argument list values xmma = numpy.zeros([n], numpy.float) # Space used internally by the program # for the asymptotes (xlow and xupp) and # computed bounds on x (alpha and beta) xlow = numpy.zeros([n], numpy.float) xupp = numpy.zeros([n], numpy.float) alfa = numpy.zeros([n], numpy.float) beta = numpy.zeros([n], numpy.float) # The objective and constraint function # values and space for the gradients fval = numpy.zeros([m], numpy.float) df0dx = numpy.zeros([n], numpy.float) dfdx = numpy.zeros([m*n], numpy.float) # Space for the coefficients and artificial # variables to be computed (set to default values) p = numpy.zeros([m*n], numpy.float) q = numpy.zeros([m*n], numpy.float) p0 = numpy.zeros([n], numpy.float) q0 = numpy.zeros([n], numpy.float) b = numpy.zeros([m], numpy.float) y = numpy.zeros([m], numpy.float) z = numpy.array([0.], numpy.float) a = numpy.zeros([m], numpy.float) c = 10000*numpy.ones([m], numpy.float) # Space for the Lagrange multipliers (ulam) # the gradient of the dual objective function, # search direction, and Hessian of the dual objective ulam = numpy.ones([m], numpy.float) gradf = numpy.zeros([m], numpy.float) dsrch = numpy.zeros([m], numpy.float) hessf = numpy.zeros([m*(m+1)/2], numpy.float) # Specify that all variables are free to move iyfree = numpy.ones([m], numpy.int) # iter = numpy.array([0], numpy.int) maxit = numpy.array([self.options['MAXIT'][1]], numpy.int) geps = numpy.array([self.options['GEPS'][1]], numpy.float) dabobj = numpy.array([self.options['DABOBJ'][1]], numpy.float) delobj = numpy.array([self.options['DELOBJ'][1]], numpy.float) itrm = numpy.array([self.options['ITRM'][1]], numpy.int) inform = numpy.array([0], numpy.int) if (myrank == 0): iprint = numpy.array([self.options['IPRINT'][1]], numpy.int) else: iprint = numpy.array([0], numpy.int) iout = numpy.array([self.options['IOUT'][1]], numpy.int) ifile = self.options['IFILE'][1] if (iprint >= 0): if os.path.isfile(ifile): os.remove(ifile) #end #end # nfunc = numpy.array([0], numpy.int) # Run MMA t0 = time.time() mma.mma(n,m,iter,maxit,geps,dabobj,delobj,itrm,inform, xval,xmma,xmin,xmax,xlow,xupp,alfa,beta,f0val,fval, fmax,df0dx,dfdx,p,q,p0,q0,b,y,z,a,c,ulam,gradf, dsrch,hessf,iyfree,iprint,iout,ifile,nfunc,func) sol_time = time.time() - t0 if (myrank == 0): if self.sto_hst: log_file.close() if tmp_file: hos_file.close() name = hos_file.filename os.remove(name+'.cue') os.remove(name+'.bin') os.rename(name+'_tmp.cue',name+'.cue') os.rename(name+'_tmp.bin',name+'.bin') #end #end #end if (iprint > 0): mma.closeunit(self.options['IOUT'][1]) #end # Store Results sol_inform = {} sol_inform['value'] = inform[0] sol_inform['text'] = self.getInform(inform[0]) if store_sol: sol_name = 'MMA Solution to ' + opt_problem.name sol_options = copy.copy(self.options) if sol_options.has_key('defaults'): del sol_options['defaults'] #end sol_evals = nfunc[0] sol_vars = copy.deepcopy(opt_problem._variables) i = 0 for key in sol_vars.keys(): sol_vars[key].value = xmma[i] i += 1 #end sol_objs = copy.deepcopy(opt_problem._objectives) i = 0 for key in sol_objs.keys(): sol_objs[key].value = f0val[i] i += 1 #end if m > 0: sol_cons = copy.deepcopy(opt_problem._constraints) i = 0 for key in sol_cons.keys(): sol_cons[key].value = fval[i] i += 1 #end else: sol_cons = {} #end sol_lambda = {} opt_problem.addSol(self.__class__.__name__, sol_name, objfunc, sol_time, sol_evals, sol_inform, sol_vars, sol_objs, sol_cons, sol_options, display_opts=disp_opts, Lambda=sol_lambda, Sensitivities=sens_type, myrank=myrank, arguments=args, **kwargs) #end return f0val, xmma, sol_inform def _on_setOption(self, name, value): ''' Set Optimizer Option Value (Optimizer Specific Routine) Documentation last updated: May. 07, 2008 - Ruben E. Perez ''' pass def _on_getOption(self, name): ''' Get Optimizer Option Value (Optimizer Specific Routine) Documentation last updated: May. 07, 2008 - Ruben E. Perez ''' pass def _on_getInform(self, infocode): ''' Get Optimizer Result Information (Optimizer Specific Routine) Keyword arguments: ----------------- id -> STRING: Option Name Documentation last updated: May. 07, 2008 - Ruben E. Perez ''' return self.informs[infocode] def _on_flushFiles(self): ''' Flush the Output Files (Optimizer Specific Routine) Documentation last updated: August. 09, 2009 - Ruben E. Perez ''' # iPrint = self.options['IPRINT'][1] if (iPrint >= 0): mma.pyflush(self.options['IOUT'][1]) #end #============================================================================== # MMA Optimizer Test #============================================================================== if __name__ == '__main__': # Test MMA print 'Testing ...' mma = MMA() print mma