Esempio n. 1
0
    def line_search(cls):
        optimize.initialize_search()

        while True:
            cls.evaluate_function()
            optimize.update_status()

            if optimize.isdone:
                optimize.finalize_search()
                break

            elif optimize.step_count < PAR.STEPMAX:
                optimize.compute_step()
                continue

            else:
                retry = optimize.retry_status
                if retry:
                    print ' Line search failed... retry'
                    optimize.restart()
                    cls.line_search()
                    break
                else:
                    print ' Line search failed... abort'
                    sys.exit(-1)
Esempio n. 2
0
    def line_search(cls):
        optimize.initialize_search()

        while True:
            cls.evaluate_function()
            optimize.update_status()

            if optimize.isdone:
                optimize.finalize_search()
                break

            elif optimize.step_count < PAR.STEPMAX:
                optimize.compute_step()
                continue

            else:
                retry = optimize.retry_status
                if retry:
                    print ' Line search failed... retry'
                    optimize.restart()
                    cls.line_search()
                    break
                else:
                    print ' Line search failed... abort'
                    sys.exit(-1)
    def iterate_search(self):
        """ First, calls self.evaluate_function, which carries out a forward 
          simulation given the current trial model. Then calls
          optimize.update_status, which maintains search history and checks
          stopping conditions.
        """
        if PAR.VERBOSE > 0:
            print " trial step", optimize.step_count + 1

        self.evaluate_function()
        optimize.update_status()
Esempio n. 4
0
    def iterate_search(self):
        """ First, calls self.evaluate_function, which carries out a forward 
          simulation given the current trial model. Then calls
          optimize.update_status, which maintains search history and checks
          stopping conditions.
        """
        if PAR.VERBOSE:
            print " trial step", optimize.step_count+1

        self.evaluate_function()
        optimize.update_status()