def test_acer():
def Sphere(individual):
"""Sphere test objective function.
F(x) = sum_{i=1}^d xi^2
d=1,2,3,...
Range: [-100,100]
Minima: 0
"""
return sum(x**2 for x in individual)
nx=5
bounds={}
for i in range(1,nx+1):
bounds['x'+str(i)]=['int', -100, 100]
#create an enviroment class
env=CreateEnvironment(method='acer', fit=Sphere,
bounds=bounds, mode='min', episode_length=50)
#create a callback function to log data
cb=RLLogger(check_freq=1, mode='min')
#create an acer object based on the env object
acer = ACER(MlpPolicy, env=env, n_steps=25, q_coef=0.55, ent_coef=0.02)
#optimise the enviroment class
acer.learn(total_timesteps=2000, callback=cb)
#print the best results
print('--------------- ACER results ---------------')
print('The best value of x found:', cb.xbest)
print('The best value of y found:', cb.rbest)
return
def test_acer():
#create an object from the class
env = IntegerSphere()
#create a callback function to log data
cb = RLLogger(check_freq=1)
#create an acer object based on the env object
acer = ACER(MlpPolicy, env=env, n_steps=25, q_coef=0.55, ent_coef=0.02)
#optimise the enviroment class
acer.learn(total_timesteps=2000, callback=cb)
#print the best results
print('--------------- ACER results ---------------')
print('The best value of x found:', cb.xbest)
print('The best value of y found:', cb.rbest)
return
def test_ppo():
#create an object from the class
env = Sphere()
#create a callback function to log data
cb = RLLogger(check_freq=1)
#create an a2c object based on the env object
ppo = PPO2(MlpPolicy, env=env, n_steps=12)
#optimise the enviroment class
ppo.learn(total_timesteps=2000, callback=cb)
#print the best results
print('--------------- PPO results ---------------')
print('The best value of x found:', cb.xbest)
print('The best value of y found:', cb.rbest)
return
def test_dqn():
#create an object from the class
env = IntegerSphere()
#create a callback function to log data
cb = RLLogger(check_freq=1)
#create an a2c object based on the env object
dqn = DQN(DQNPolicy, env=env)
#optimise the enviroment class
dqn.learn(total_timesteps=2000, callback=cb)
#print the best results
print('--------------- DQN results ---------------')
print('The best value of x found:', cb.xbest)
print('The best value of y found:', cb.rbest)
return
F(x) = sum_{i=1}^d xi^2
d=1,2,3,...
Range: [-100,100]
Minima: 0
"""
#-1 is used to convert minimization to maximization
return -sum(x**2 for x in individual)
def reset(self):
self.done = False
return self.action_space.sample()
def render(self, mode='human'):
pass
#--------------------------------------------------------
# RL Optimisation
#--------------------------------------------------------
#create an object from the class
env = Sphere()
#create a callback function to log data
cb = RLLogger(check_freq=1)
#create an acktr object based on the env object
acktr = ACKTR(MlpPolicy, env=env, n_steps=12)
#optimise the enviroment class
acktr.learn(total_timesteps=2500, callback=cb)
#print the best results
print('--------------- ACKTR results ---------------')
print('The best value of x found:', cb.xbest)
print('The best value of y found:', cb.rbest)
def Sphere(individual):
"""Sphere test objective function.
F(x) = sum_{i=1}^d xi^2
d=1,2,3,...
Range: [-100,100]
Minima: 0
"""
return sum(x**2 for x in individual)
nx = 5
bounds = {}
for i in range(1, nx + 1):
bounds['x' + str(i)] = ['float', -10, 10]
#create an enviroment class
env = CreateEnvironment(method='acktr',
fit=Sphere,
bounds=bounds,
mode='min',
episode_length=50)
#create a callback function to log data
cb = RLLogger(check_freq=1, mode='min')
#create an acktr object based on the env object
acktr = ACKTR(MlpPolicy, env=env, n_steps=12, seed=1)
#optimise the enviroment class
acktr.learn(total_timesteps=2000, callback=cb)
#print the best results
print('--------------- ACKTR results ---------------')
print('The best value of x found:', cb.xbest)
print('The best value of y found:', cb.rbest)