示例#1
0
            
            algo.reset_ygradient()
            for point in algo.build_gradient(vars = var):
                
                algo.updateBeta(point)
                
                perf = gameStrat(strat = 6, totalgames=3,Algo=algo, \
                                renderme = False)
                
                y = evalGames(perf)
                
                algo.update_ygradient(y)    
            
            #update algo
            updateB = algo.eval()
            algo.updateBetaFinal(updateB)
            
            #Exit Training Conditions
            if y > 500:
                break
                
        #Print summary of loop outcome
        print str(algo.BetaFinal)
            

        
    else:
        for s in [1,2,3,4]:
            perf = gameStrat(strat = s, totalgames=5)
            eval = evalGames(perf)
            print eval
示例#2
0
    

#import optparse
# p = optparse.OptionParser()
# p.add_option('--foo', '-f', default="yadda")
# p.add_option('--bar', '-b')
# options, arguments = p.parse_args()


if __name__ == "__main__":
    
    if len(sys.argv) > 1:
        
        algo = Algo()
        #cheats, init-Betas
        algo.updateBetaFinal([0,0,0,.1])
        
        env = createEnvNoisy()
        ind = 0
        
        while True:
        
            
            var = [0,1,2,3]  
            ep = [0.2,0.1,.005,0.01]
            
            algo.reset_ygradient()
            algo.reset_ymisc()
            for point in algo.build_gradient(vars = var, eps = ep):
                
                algo.updateBeta(point)
示例#3
0
    return 0


#import optparse
# p = optparse.OptionParser()
# p.add_option('--foo', '-f', default="yadda")
# p.add_option('--bar', '-b')
# options, arguments = p.parse_args()

if __name__ == "__main__":

    if len(sys.argv) > 1:

        algo = Algo()
        #cheats, init-Betas
        algo.updateBetaFinal([0, 0, 0, .1])

        env = createEnvNoisy()
        ind = 0

        while True:

            var = [0, 1, 2, 3]
            ep = [0.2, 0.1, .005, 0.01]

            algo.reset_ygradient()
            algo.reset_ymisc()
            for point in algo.build_gradient(vars=var, eps=ep):

                algo.updateBeta(point)
示例#4
0
            var = [2]

            algo.reset_ygradient()
            for point in algo.build_gradient(vars=var):

                algo.updateBeta(point)

                perf = gameStrat(strat = 6, totalgames=3,Algo=algo, \
                                renderme = False)

                y = evalGames(perf)

                algo.update_ygradient(y)

            #update algo
            updateB = algo.eval()
            algo.updateBetaFinal(updateB)

            #Exit Training Conditions
            if y > 500:
                break

        #Print summary of loop outcome
        print str(algo.BetaFinal)

    else:
        for s in [1, 2, 3, 4]:
            perf = gameStrat(strat=s, totalgames=5)
            eval = evalGames(perf)
            print eval