-
Notifications
You must be signed in to change notification settings - Fork 1
/
Environment.py
299 lines (266 loc) · 11.6 KB
/
Environment.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
import numpy as np
import pygame
import math
SIZE = 500 # Size of Visualization (500*500)
HOLE = 1 # Formalization of board parts
FIGURE = 2
BOARD = 3
d = {1: (255, 255, 255), # RGB values for parts
2: (0, 0, 0),
3: (100, 100, 100)}
JUMP_RIGHT = (1, 0) # Movement possibilities
JUMP_LEFT = (-1, 0)
JUMP_UP = (0, -1)
JUMP_DOWN = (0, 1)
DO_NOTHING = ((0, 0), (0, 0))
MOVES = (JUMP_RIGHT, JUMP_LEFT, JUMP_UP, JUMP_DOWN)
class Board: # Board class for solitaire
def __init__(self):
self.board = np.zeros((9, 9, 3), dtype=np.uint8) # Board array for RGB values
self.holes = np.zeros((9, 9)) # Hole positions
self.figures = np.zeros((9, 9)) # Figure positions
self.boundary = np.zeros((33, 2)) # Actual Board positions
self.boundary = ((3, 1), (4, 1), (5, 1), (3, 2), (4, 2),
(5, 2), (1, 3), (2, 3), (3, 3), (4, 3),
(5, 3), (6, 3), (7, 3), (1, 4), (2, 4),
(3, 4), (4, 4), (5, 4), (6, 4), (7, 4),
(1, 5), (2, 5), (3, 5), (4, 5), (5, 5),
(6, 5), (7, 5), (3, 6), (4, 6), (5, 6),
(3, 7), (4, 7), (5, 7))
self.value_board = np.zeros((9, 9))
self.center = np.zeros((9, 9))
i = 0
for pos in self.boundary:
if i in (0, 2, 6, 12, 20, 26, 30, 32):
self.value_board[pos] = 11250
self.center[pos] = 6
elif i in (1, 13, 19, 31):
self.value_board[pos] = 2250
self.center[pos] = 5
elif i in (3, 5, 7, 11, 21, 25, 27, 29):
self.value_board[pos] = 250
self.center[pos] = 4
elif i in (4, 14, 18, 28):
self.value_board[pos] = 50
self.center[pos] = 3
elif i in (8, 10, 22, 24):
self.value_board[pos] = 10
self.center[pos] = 2
elif i in (9, 15, 17, 23):
self.value_board[pos] = 2
self.center[pos] = 1
else:
self.value_board[pos] = 1
i += 1
# Creating dictionary for later actions
self.action_code = {}
self.create_action_code()
self.pins = 32 # Number of pins (figures) at beginning
self.moves = [] # List for possible actions (movements)
self.jump_reward = 1
self.finish_reward = 10000
self.dead_end_penalty = -1000
# Initializing game visualization
self.game_display = pygame.display.set_mode((SIZE, SIZE))
pygame.display.set_caption('Solitaire')
self.clock = pygame.time.Clock()
def set_params(self, jump_reward, finish_reward, dead_end_penalty):
self.jump_reward = jump_reward
self.finish_reward = finish_reward
self.dead_end_penalty = dead_end_penalty
# Resetting board to starting position (must be called before first game and after each game)
def reset(self):
# Initializing figures and holes
for pos in self.boundary:
self.figures[pos] = True
self.holes[pos] = False
self.figures[4][4] = False
self.holes[4][4] = True
self.update() # Updating self.board RGB values
self.pins = 32 # Resetting number of pins
self.calc_moves() # Calculating possible movements (stored in self.moves)
# Returning values
actions = self.moves
value, center = self.get_state_value()
pins = self.pins
state = (value, center) # state consisting of state value and number of pins
return state, pins, actions
# Updating self.board RGB values
def update(self):
for pos in self.boundary:
if self.holes[pos]:
self.board[pos] = d[HOLE]
elif self.figures[pos]:
self.board[pos] = d[FIGURE]
# Displaying Board with current positioning
def show(self):
self.game_display.fill(d[BOARD])
for pos in self.boundary:
pos_list = list(pos)
pygame.draw.circle(self.game_display, self.board[pos],
[(pos_list[0] + 1) * int(SIZE / 9) - int(SIZE / 18),
(pos_list[1] + 1) * int(SIZE / 9) - int(SIZE / 18)], int(SIZE / 30))
pygame.display.update()
# Visualizing game (must be called repeatedly to display)
def play(self):
for event in pygame.event.get():
if event.type == pygame.QUIT:
pygame.quit()
quit()
self.show()
self.clock.tick(1000)
def save_game(self, filename):
pygame.image.save(self.game_display, filename)
# Checking if one movement is possible
def check_moves(self, pos, direction):
pos_list = list(pos)
direction_list = list(direction)
figure_pos = (pos_list[0] + direction_list[0], pos_list[1] + direction_list[1]) # figure to be jumped
new_pos = (pos_list[0] + 2 * direction_list[0], pos_list[1] + 2 * direction_list[1]) # new position
if pos in self.boundary and new_pos in self.boundary and self.figures[pos] \
and self.figures[figure_pos] and self.holes[new_pos]: # positions must be in boundary,
# current position must be figure, position to be jumped must be figure and new position must be hole
return True
else:
return False
# Calculating possible movements (stored in self.moves)
def calc_moves(self):
self.moves = []
for act_key in self.action_code:
if self.check_moves(self.action_code[act_key][0], self.action_code[act_key][1]):
self.moves.append(act_key)
# Making a move
def move(self, movement):
pos1 = self.action_code[movement][0] # current position
pos2 = (self.action_code[movement][0][0] + self.action_code[movement][1][0], # other figure position
self.action_code[movement][0][1] + self.action_code[movement][1][1])
pos3 = (self.action_code[movement][0][0] + 2 * self.action_code[movement][1][0], # hole position
self.action_code[movement][0][1] + 2 * self.action_code[movement][1][1])
if movement in self.moves: # checking if movement is legit
self.figures[pos1] = False
self.holes[pos1] = True
self.figures[pos2] = False
self.holes[pos2] = True
self.holes[pos3] = False
self.figures[pos3] = True
self.update() # Updating self.board RGB values
self.pins -= 1 # Reducing number of pins by one
# action command (returning state as tuple of state value and number of pins,
# reward depending on occasion, next possible actions as list and done if no more move possible)
def action(self, action):
if action is not DO_NOTHING:
self.move(action)
self.calc_moves()
value, center = self.get_state_value()
actions = self.moves
pins = self.pins
state = (value, center)
done = False
if not len(self.moves):
done = True
reward = self.get_reward()
return state, pins, reward, actions, done
def get_state_value(self):
value = 0
center = 0
dist = 1000
for pos in self.boundary:
if self.figures[pos]:
value += self.value_board[pos]
if self.sum_dist(pos) < dist:
center = self.center[pos]
dist = self.sum_dist(pos)
elif self.sum_dist(pos) is dist:
if self.center[pos] < center:
center = self.center[pos]
dist = self.sum_dist(pos)
value = int(value)
center = int(center)
return value, center
@staticmethod
def dist(pos1, pos2):
pos1_list = list(pos1)
pos2_list = list(pos2)
dist_x = abs(pos2_list[0] - pos1_list[0])
dist_y = abs(pos2_list[1] - pos1_list[1])
dist = math.sqrt(dist_x ** 2 + dist_y ** 2)
return dist
def sum_dist(self, pos):
sum_dist = 0
for fig in self.boundary:
if self.figures[fig]:
sum_dist += self.dist(pos, fig)
return sum_dist
def get_reward(self):
value, center = self.get_state_value()
reward = self.jump_reward * 1
if value == 1: # self.value is only 0 if goal is reached
reward = self.finish_reward
elif not len(self.moves): # dead end if no more moves possible
reward = self.dead_end_penalty * value / 1000
return reward
# Creates dictionary of action codes (is being called in __init__)
def create_action_code(self):
code = 0
for pos in self.boundary:
pos_list = list(pos)
pos_right = (pos_list[0] + 2, pos_list[1])
pos_left = (pos_list[0] - 2, pos_list[1])
pos_up = (pos_list[0], pos_list[1] - 2)
pos_down = (pos_list[0], pos_list[1] + 2)
if pos_right in self.boundary:
self.action_code.update({code: (pos, JUMP_RIGHT)})
code += 1
if pos_left in self.boundary:
self.action_code.update({code: (pos, JUMP_LEFT)})
code += 1
if pos_up in self.boundary:
self.action_code.update({code: (pos, JUMP_UP)})
code += 1
if pos_down in self.boundary:
self.action_code.update({code: (pos, JUMP_DOWN)})
code += 1
# Returns number of theoretically possible actions
def action_space(self):
space = len(self.action_code)
return space
@staticmethod
# Returns number of possible states
def observation_space():
space = [101250, 7]
return space
# Loading board randomly with number of pins (2 - 32)
def load_board(self, pins):
self.pins = pins
for pos in self.boundary:
self.figures[pos] = False
self.holes[pos] = True
self.figures[4][4] = True
self.holes[4][4] = False
if pins > 1:
pins -= 1
while pins:
possible_pos = []
for pos in self.boundary:
if self.figures[pos]:
pos_list = list(pos)
for move in MOVES:
move_list = list(move)
new_fig = (pos_list[0] + move_list[0], pos_list[1] + move_list[1])
new_pos = (pos_list[0] + 2 * move_list[0], pos_list[1] + 2 * move_list[1])
if new_pos in self.boundary and self.holes[new_pos] and self.holes[new_fig]:
possible_pos.append([pos, new_fig, new_pos])
if len(possible_pos):
new = possible_pos[np.random.randint(0, len(possible_pos))]
self.figures[new[0]] = False
self.holes[new[0]] = True
self.figures[new[1]] = True
self.holes[new[1]] = False
self.figures[new[2]] = True
self.holes[new[2]] = False
pins -= 1
else:
break
self.update()
state, pins, reward, actions, done = self.action(DO_NOTHING)
return state, pins, actions