ai-snake-lab 0.4.5__tar.gz → 0.4.7__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/PKG-INFO +1 -1
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ai/ReplayMemory.py +31 -10
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DDef.py +1 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DDir.py +0 -1
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DFile.py +0 -1
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/game/SnakeGame.py +6 -6
- {ai_snake_lab-0.4.5/ai_snake_lab → ai_snake_lab-0.4.7/ai_snake_lab/ui}/AISim.py +5 -1
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/pyproject.toml +2 -2
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/LICENSE +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/README.md +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ai/AIAgent.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ai/AITrainer.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ai/EpsilonAlgo.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ai/models/ModelL.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ai/models/ModelRNN.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DDb4EPlot.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DEpsilon.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DFields.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DLabels.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DLayout.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DModelL.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DModelLRNN.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DReplayMemory.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/DSim.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/constants/__init__.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/game/GameBoard.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/game/GameElements.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/ui/Db4EPlot.py +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/utils/AISim.tcss +0 -0
- {ai_snake_lab-0.4.5 → ai_snake_lab-0.4.7}/ai_snake_lab/utils/ConstGroup.py +0 -0
@@ -14,10 +14,11 @@ import os
|
|
14
14
|
from collections import deque
|
15
15
|
import random
|
16
16
|
import sqlite3, pickle
|
17
|
+
import tempfile
|
18
|
+
import shutil
|
17
19
|
|
18
20
|
from ai_snake_lab.constants.DReplayMemory import MEM_TYPE
|
19
|
-
from ai_snake_lab.constants.
|
20
|
-
from ai_snake_lab.constants.DDir import DDir
|
21
|
+
from ai_snake_lab.constants.DDef import DDef
|
21
22
|
|
22
23
|
|
23
24
|
class ReplayMemory:
|
@@ -31,11 +32,6 @@ class ReplayMemory:
|
|
31
32
|
self.max_states = 15000
|
32
33
|
self.max_shuffle_games = 40
|
33
34
|
self.max_games = 500
|
34
|
-
self.db_file = os.path.join(DDir.AI_SNAKE_LAB, DDir.DB, DFile.REPLAY_DB)
|
35
|
-
|
36
|
-
# Delete the replay memory file, if it exists
|
37
|
-
if os.path.exists(self.db_file):
|
38
|
-
os.remove(self.db_file)
|
39
35
|
|
40
36
|
if self._mem_type == MEM_TYPE.SHUFFLE:
|
41
37
|
# States are stored in a deque and a random sample will be returned
|
@@ -46,13 +42,33 @@ class ReplayMemory:
|
|
46
42
|
# A complete game will be returned
|
47
43
|
self.cur_memory = []
|
48
44
|
|
45
|
+
# Get a temporary directory for the DB file
|
46
|
+
self._tmpfile = tempfile.NamedTemporaryFile(suffix=DDef.DOT_DB, delete=False)
|
47
|
+
self.db_file = self._tmpfile.name
|
48
|
+
|
49
49
|
# Connect to SQLite
|
50
50
|
self.conn = sqlite3.connect(self.db_file, check_same_thread=False)
|
51
|
+
|
52
|
+
# Get a cursor
|
51
53
|
self.cursor = self.conn.cursor()
|
54
|
+
|
55
|
+
# We don't need the file handle anymore
|
56
|
+
self._tmpfile.close()
|
57
|
+
|
58
|
+
# Intialize the schema
|
52
59
|
self.init_db()
|
53
60
|
|
54
|
-
def
|
55
|
-
return
|
61
|
+
def __enter__(self):
|
62
|
+
return self
|
63
|
+
|
64
|
+
def __exit__(self, exc_type, exc_val, exc_tb):
|
65
|
+
self.close()
|
66
|
+
|
67
|
+
def __del__(self):
|
68
|
+
try:
|
69
|
+
self.close()
|
70
|
+
except Exception:
|
71
|
+
pass # avoid errors on interpreter shutdown
|
56
72
|
|
57
73
|
def append(self, transition):
|
58
74
|
"""Add a transition to the current game."""
|
@@ -75,7 +91,12 @@ class ReplayMemory:
|
|
75
91
|
|
76
92
|
def close(self):
|
77
93
|
"""Close the database connection."""
|
78
|
-
self
|
94
|
+
if getattr(self, "conn", None):
|
95
|
+
self.conn.close()
|
96
|
+
self.conn = None
|
97
|
+
if getattr(self, "db_file", None) and os.path.exists(self.db_file):
|
98
|
+
os.remove(self.db_file)
|
99
|
+
self.db_file = None
|
79
100
|
|
80
101
|
def get_random_game(self):
|
81
102
|
"""Return a random full game from the database."""
|
@@ -129,12 +129,6 @@ class SnakeGame:
|
|
129
129
|
game_over = True
|
130
130
|
reward = -10
|
131
131
|
|
132
|
-
# Set a negative reward if the snake head is adjacent to the snake body.
|
133
|
-
# This is to discourage snake collisions.
|
134
|
-
for segment in self.snake[1:]:
|
135
|
-
if abs(self.head.x - segment.x) < 2 and abs(self.head.y - segment.y) < 2:
|
136
|
-
reward -= -1
|
137
|
-
|
138
132
|
if game_over == True:
|
139
133
|
# Game is over: Snake or wall collision or exceeded max moves
|
140
134
|
self.game_reward += reward
|
@@ -159,6 +153,12 @@ class SnakeGame:
|
|
159
153
|
reward -= 2
|
160
154
|
self.distance_to_food = cur_distance
|
161
155
|
|
156
|
+
## 6. Set a negative reward if the snake head is adjacent to the snake body.
|
157
|
+
# This is to discourage snake collisions.
|
158
|
+
for segment in self.snake[1:]:
|
159
|
+
if abs(self.head.x - segment.x) < 2 and abs(self.head.y - segment.y) < 2:
|
160
|
+
reward -= -2
|
161
|
+
|
162
162
|
self.game_reward += reward
|
163
163
|
self.game_board.update_snake(snake=self.snake, direction=self.direction)
|
164
164
|
self.game_board.update_food(food=self.food)
|
@@ -1,6 +1,6 @@
|
|
1
1
|
[project]
|
2
2
|
name = "ai-snake-lab"
|
3
|
-
version = "0.4.
|
3
|
+
version = "0.4.7"
|
4
4
|
description = "Interactive reinforcement learning sandbox for experimenting with AI agents in a classic Snake Game environment."
|
5
5
|
authors = [{ name = "Nadim-Daniel Ghaznavi", email = "nghaznavi@gmail.com" }]
|
6
6
|
license = { text = "GPL-3.0" }
|
@@ -38,7 +38,7 @@ dependencies = [
|
|
38
38
|
]
|
39
39
|
|
40
40
|
[project.scripts]
|
41
|
-
ai-snake-lab = "ai_snake_lab.AISim:main"
|
41
|
+
ai-snake-lab = "ai_snake_lab.ui.AISim:main"
|
42
42
|
|
43
43
|
[project.urls]
|
44
44
|
"Bug Tracker" = "https://github.com/NadimGhaznavi/ai_snake_lab/issues"
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|
File without changes
|