[简介]我是OpenAI的初学者,我已经制作了一个自定义游戏,我想在其中实现一个自学习代理。我跟着this guide在GitHub上设置了一个存储库,但是我不明白如何格式化我的代码以使用gym-foo / gym_foo / envs / foo_env.py的内容
[问题]是否有人可以指导我如何构建我的代码,以便与之兼容:
class FooEnv(gym.Env):
metadata = {'render.modes': ['human']}
def __init__(self):
...
def step(self, action):
...
def reset(self):
...
def render(self, mode='human', close=False):
...
[代码
import pygame
import time
import random
from pygame.locals import *
pygame.init()
display_width = 1002
display_height = 720
black = (0,0,0)
white = (255,255,255)
red = (220,0,0)
blue = (53,155,255)
green = (0,190,0)
bright_red = (255,0,0)
bright_green = (0,255,0)
dark_blue = (0,102,204)
yellow = (255, 255, 0)
gameDisplay = pygame.display.set_mode((display_width,display_height)) #creates surface/ display
pygame.display.set_caption('Blob Arena') #name of project
clock = pygame.time.Clock() #sets a clock
#________________________________________________________________________________________
blobImage = pygame.image.load('blob2.png')
blobIcon = pygame.image.load('blob_img.png')
bulletpicture = pygame.image.load("bullet.png")
pygame.display.set_icon(blobIcon)
pause = True
blob_width = 51
blob_height = 51
bullet_width = 12
bullet_height = 5
bullets=[]
bullets2=[]
def blob(x,y):
gameDisplay.blit(blobImage,(x,y)) #drawing to background
def bullets_hit(count):
font = pygame.font.SysFont(None, 25)
text = font.render("Score: "+str(count), True, black)
gameDisplay.blit(text,(0,20))
def player_lives(count):
font = pygame.font.SysFont(None, 25)
text = font.render("Lives Left: "+str(count), True, bright_red)
gameDisplay.blit(text,(0,0))
def game_intro():
intro = True
while intro:
for event in pygame.event.get():
if event.type == pygame.QUIT:
pygame.quit()
quit()
gameDisplay.fill(dark_blue)
largeText = pygame.font.Font('freesansbold.ttf', 110)
TextSurf, TextRect = text_objects("Blob Arena", largeText) # Returns text surface and rectangle
TextRect.center = ((display_width / 2), (display_height / 2.5))
gameDisplay.blit(TextSurf, TextRect)
button("Training 1", 200, 430, 140, 53, green, bright_green, game_loop)
button("Training 2", 431, 430, 140, 53, green, bright_green, game_loop)
button("Training 3", 662, 430, 140, 53, green, bright_green, game_loop)
button("Human vs AI", 315.5, 550, 140, 53, green, bright_green, game_loop)
button("Quit", 546.5, 550, 140, 53, red, bright_red, quit_game)
pygame.display.update()
clock.tick(15)
def button(msg,x,y,w,h,ic,ac,action=None):
mouse = pygame.mouse.get_pos()
click = pygame.mouse.get_pressed() #collects mouse left, right and middle button
if x + w > mouse[0] > x and y + h > mouse[1] > y:
pygame.draw.rect(gameDisplay, ac, (x, y, w, h))
if click[0] == 1 and action!= None:
action()
else:
pygame.draw.rect(gameDisplay, ic, (x, y, w, h))
smallText = pygame.font.Font("freesansbold.ttf", 20)
textSurf, textRect = text_objects(msg, smallText)
textRect.center = ((x + (w / 2)), (y + (h / 2)))
gameDisplay.blit(textSurf, textRect)
def unpause():
global pause
pause = False
def paused():
largeText = pygame.font.Font('freesansbold.ttf', 110)
TextSurf, TextRect = text_objects("Paused", largeText) # Returns text surface and rectangle
TextRect.center = ((display_width / 2), (display_height / 2.5))
gameDisplay.blit(TextSurf, TextRect)
while pause:
for event in pygame.event.get():
if event.type == pygame.QUIT:
pygame.quit()
quit()
button("Continue", 315.5, 450, 140, 53, green, bright_green, unpause)
button("Quit", 546.5, 450, 140, 53, red, bright_red, quit_game)
pygame.display.update()
clock.tick(15)
def text_objects(text, font):
textSurface = font.render(text, True, black)
return textSurface, textSurface.get_rect()
def quit_game():
pygame.quit()
quit()
def game_over():
largeText = pygame.font.Font('freesansbold.ttf', 110)
TextSurf, TextRect = text_objects("Game Over", largeText) # Returns text surface and rectangle
TextRect.center = ((display_width / 2), (display_height / 2.5))
gameDisplay.blit(TextSurf, TextRect)
while True:
for event in pygame.event.get():
if event.type == pygame.QUIT:
pygame.quit()
quit()
button("Play Again", 280.5, 450, 140, 53, green, bright_green,game_loop)
button("Quit", 581.5, 450, 140, 53, red, bright_red,quit_game)
pygame.display.update()
clock.tick(15)
def game_loop():
global pause
x = (display_width * 0.08)
y = (display_height * 0.2)
x_change = 0
y_change = 0
blob_speed = 2
velocity = [2, 2]
score = 0
lives = 3
pos_x = display_width/1.2
pos_y = display_height/1.2
previous_time = pygame.time.get_ticks()
previous_time2 = pygame.time.get_ticks()
gameExit = False
while not gameExit:
for event in pygame.event.get():#monitors hardware movement/ clicks
if event.type == pygame.QUIT:
pygame.quit()
quit()
pos_x += velocity[0]
pos_y += velocity[1]
if pos_x + blob_width > display_width or pos_x < 601:
velocity[0] = -velocity[0]
if pos_y + blob_height > display_height or pos_y < 0:
velocity[1] = -velocity[1]
for b in range(len(bullets2)):
bullets2[b][0] -= 6
for bullet in bullets2:
if bullet[0] < 0:
bullets2.remove(bullet)
current_time2 = pygame.time.get_ticks()
#ready to fire when 500 ms have passed.
if current_time2 - previous_time2 > 500:
previous_time2 = current_time2
bullets2.append([pos_x+25, pos_y+24])
# Checks to see if any keys are held down and remembers them with the variable keys.
keys = pygame.key.get_pressed()
for b in range(len(bullets)):
bullets[b][0] += 6
for bullet in bullets:
if bullet[0] > 1005:
bullets.remove(bullet)
if keys[pygame.K_SPACE]:
current_time = pygame.time.get_ticks()
#ready to fire when 500 ms have passed.
if current_time - previous_time > 500:
previous_time = current_time
bullets.append([x+25, y+24])
# If the player is holding down one key or the other the blob moves in that direction
if x < 0:
x = 0
if keys[pygame.K_a]:
x_change = -blob_speed
if x > 401 - blob_width:
x = 401 - blob_width
if keys[pygame.K_d]:
x_change = blob_speed
if keys[pygame.K_p]:
pause = True
paused()
# If the player is holding down both or neither of the keys the blob stops
if keys[pygame.K_a] and keys[pygame.K_d]:
x_change = 0
if not keys[pygame.K_a] and not keys[pygame.K_d]:
x_change = 0
if y < 0:
y = 0
if keys[pygame.K_w]:
y_change = -blob_speed
if y > display_height - blob_height:
y = display_height - blob_height
if keys[pygame.K_s]:
y_change = blob_speed
if keys[pygame.K_w] and keys[pygame.K_s]:
y_change = 0
if not keys[pygame.K_w] and not keys[pygame.K_s]:
y_change = 0
#print(event)
# Reset x and y to new position
x += x_change
y += y_change
gameDisplay.fill(blue) #changes background surface
bullets_hit(score)
player_lives(lives)
pygame.draw.line(gameDisplay, black, (601, display_height), (601, 0), 3)
pygame.draw.line(gameDisplay, black, (401, display_height), (401, 0), 3)
blob(pos_x, pos_y)
blob(x, y)
for bullet in bullets:
gameDisplay.blit(bulletpicture, pygame.Rect(bullet[0], bullet[1], 0, 0))
if bullet[0] > pos_x and bullet[0] < pos_x + blob_width:
if bullet[1] > pos_y and bullet[1] < pos_y + blob_height or bullet[1] + bullet_height > pos_y and bullet[1] + bullet_height < pos_y + blob_height:
bullets.remove(bullet)
score+=1
for bullet in bullets2:
gameDisplay.blit(bulletpicture, pygame.Rect(bullet[0], bullet[1], 0, 0))
if bullet[0] + bullet_width < x + blob_width and bullet[0] > x:
if bullet[1] > y and bullet[1] < y + blob_height or bullet[1] + bullet_height > y and bullet[1] + bullet_height < y + blob_height:
bullets2.remove(bullet)
lives-=1
if lives == 0:
game_over()
pygame.display.update() #update screen
clock.tick(120)#moves frame on (fps in parameters)
game_intro()
game_loop()
pygame.quit()
quit()
[附加信息]我将使用来自https://github.com/Hvass-Labs/TensorFlow-Tutorials/blob/master/reinforcement_learning.py的强化学习算法,根据我的理解,需要健身房环境才能使gym.make()工作。非常感谢任何帮助,如果还需要更多帮助,请告诉我。
答案 0 :(得分:1)
我没有使用pygame
库的经验,也不了解其内部工作原理,这可能会对代码需要运行的地方产生一些影响,所以我并不是100%肯定所有这些。但是,从一些直观的理解开始大致应该发生在哪里开始是很好的:
__init__()
应该运行任何一次性设置。我可以想象像pygame.init()
这样的内容可能会在这里进行,但我不能100%肯定,因为我不熟悉pygame
。step()
,然后运行单个框架&#39;&#39;&#39;&#39;对于游戏,在代理人选择的动作的情况下将其向前移动。或者,如果你有一个游戏,其中一个动作需要多个帧,你应该在这里运行多个帧。基本上:保持游戏向前移动,直到你达到代理人应该再次选择新动作的点,然后返回当前游戏状态。reset()
应该......好吧,重置游戏。因此,恢复到(或随机,无论你想要的)初始游戏状态,运行可能需要的任何清理。例如,我可以想象pygame.init()
属于这里。这取决于该功能究竟是做什么的。如果它只需要运行一次,它就属于__init__()
。如果它需要在每个新游戏/&#34;剧集&#34;的开头运行,则ir属于reset()
。render()
应该包含大多数与图形相关的代码。你可以尝试从健身房中的cartpole environment中获取灵感,这里也会绘制一些相当简单的图形。它看起来应该只绘制一帧。现在,看看你开始使用的代码,似乎有相当数量的用户界面代码...各种与按钮相关的代码,暂停/取消暂停,幻想(动画?)介绍在游戏开始时。我不知道你是否有能力摆脱这一切?如果您正在进行纯粹的强化学习,您可能可以。如果您仍然需要用户交互,那么您可能无法做到,然后事情变得更加困难,因为所有这些事情都不适合gym
框架。
我可以尝试对代码的其余部分进行一些有根据的猜测以及它应该去哪里,但是你应该根据上面更一般的指导原则仔细检查所有内容:
def reset(self):
# all the following code seems to define the initial game state
x = (display_width * 0.08)
y = (display_height * 0.2)
x_change = 0
y_change = 0
blob_speed = 2
velocity = [2, 2]
score = 0
lives = 3
pos_x = display_width/1.2
pos_y = display_height/1.2
bullets=[]
bullets2=[]
step()
可能应包含while
中game_loop()
循环中的大部分代码。应该修改密钥输入检查代码,而应该开始使用传递给action
的{{1}}。最后,step()
应返回包含以下内容的元组:
step()
如果游戏结束,否则True
。False
。