print "Resetting"
long_press(0)
long_press(0)
# Main loop
for epoch in range(MAX_EPOCHS):
print "New epoch: %d\n" % epoch
reset()
for step in range(MAX_STEPS):
# Keep note of the fact that we don't have the concept of an episode unlike nathan's implementation
image = get_observation()
best_action = choose_action(step)
# get best possible action from the current neural network
images.push(image)
actions.push(best_action)
# If the current state is dead, push 0 reward and mark state as terminal. then reset and continue loop execution
if am_i_dead():
terminals.push(1)
rewards.push(0)
reset()
continue
# Still alive, still alive!
terminals.push(0)
# long press the best action because humans press keys for longer durations
reward = long_press(best_action)
rewards.push(reward)
list1.Find("3")
print("Reseting the RingBuffer to capacity 0")
list1 = RingBuffer(0)
print("Trying to insert_keep_new(1)")
list1.insert_keep_new("1")
print("-----------------------")
print("Test Case of Stack")
print(
"Assumed Stack capacity is 5 for testing. All the test cases below are according to size 5"
)
sizeOfS = int(input("Enter size of Stack"))
list1 = Stack(sizeOfS)
print("Adding 1")
list1.push(1)
print("Adding 2")
list1.push(2)
print("Adding 3")
list1.push(3)
list1.__str__()
print("poping ", list1.peek())
list1.pop()
list1.__str__()
print("Pushing 4")
list1.push(4)
print("Pushing 5")
list1.push(5)
list1.__str__()
print("Pushing 6")
list1.push(6)
#long_press(0)
# Main loop
for epoch in range(MAX_EPOCHS):
print "New epoch: %d\n" % epoch
reset(epoch)
for step in range(MAX_STEPS):
# Keep note of the fact that we don't have the concept of an episode unlike nathan's implementation
image = get_observation()
best_action = choose_action(image, step, epoch)
# get best possible action from the current neural network
images.push(image)
actions.push(best_action)
# If the current state is dead, push 0 reward and mark state as terminal. then reset and continue loop execution
if am_i_dead():
terminals.push(1)
rewards.push(0)
reset(epoch)
continue
# Still alive, still alive!
terminals.push(0)
# long press the best action because humans press keys for longer durations
reward = long_press(best_action)
rewards.push(reward)
