#from keras.layers import Dense
#from keras.optimizers import Adam
import random
from rl_dqn import DeepQNetwork, learning_rate, discount_rate
#Create Gym
#from gym import wrappers
env = gym.make('CartPole-v1')
#env.seed(50) #Set the seed to keep the environment consistent across runs
weights_file = 'rldqn_cartpole.h5'
state_dim = env.observation_space.shape[0] #This is only 4
action_dim = env.action_space.n #Actions
print("Deifning dqn")
dqn1 = DeepQNetwork(state_dim, action_dim, 0.001, 0.95, 1, 0.001, 0.995)
print("loadin dqn")
dqn1.load_weights(weights_file)
TEST_Episodes = 10
rewards = [] #Store rewards for graphing
#Test the agent that was trained
# In this section we ALWAYS use exploit don't train any more
for e_test in range(TEST_Episodes):
state, done = env.reset(), False
state = np.reshape(state, [1, state_dim])
tot_rewards = 0
while not done:
env.render()
action = dqn1.test_action(state)
from collections import deque
import random
from rl_dqn import DeepQNetwork
from kinematic_model import Kinematic_Model
ENV = "CartPole-v1"
env = gym.make(ENV)
state_dim = env.observation_space.shape[0]
action_dim = env.action_space.n
########################################################
# RL Agent for Oracle
weights_file = 'rldqn_cartpole.h5'
oracle = DeepQNetwork(state_dim, action_dim, 0.001, 0.95, 1, 0.001, 0.995 )
oracle.load_weights(weights_file)
'''
TEST_Episodes=10
rewards = [] #Store rewards for graphing
#Test the agent that was trained
# In this section we ALWAYS use exploit don't train any more
for e_test in range(TEST_Episodes):
state,done = env.reset(), False
state = np.reshape(state, [1, state_dim])
tot_rewards = 0
while not done:
env.render()
action = oracle.test_action(state)
KM_nstate = Kinematic_Model(state[0],action)
nstate, reward, done, _ = env.step(action)
import time
from collections import deque
import random
from rl_dqn import DeepQNetwork
from kinematic_model import Kinematic_Model
ENV = "CartPole-v1"
env = gym.make(ENV)
state_dim = env.observation_space.shape[0]
action_dim = env.action_space.n
########################################################
# RL Agent for Oracle
weights_file = 'rldqn_cartpole.h5'
oracle = DeepQNetwork(state_dim, action_dim, 0.001, 0.95, 1, 0.001, 0.995)
oracle.load_weights(weights_file)
'''
TEST_Episodes=10
rewards = [] #Store rewards for graphing
#Test the agent that was trained
# In this section we ALWAYS use exploit don't train any more
for e_test in range(TEST_Episodes):
state,done = env.reset(), False
state = np.reshape(state, [1, state_dim])
tot_rewards = 0
while not done:
env.render()
action = oracle.test_action(state)
KM_nstate = Kinematic_Model(state[0],action)
nstate, reward, done, _ = env.step(action)