Changed reward scheme for Pyraminx LBL Method and Skewb Sarah's Method

DoubleGremlin181 · Jul 13, 2020 · d95fd04 · d95fd04
1 parent 8975f0d
commit d95fd04
Show file tree

Hide file tree

Showing 4 changed files with 9 additions and 10 deletions.
diff --git a/rubiks_cube_gym/__init__.py b/rubiks_cube_gym/__init__.py
@@ -25,7 +25,7 @@
 )
 
 register(
-    id='pyraminx-wo-tips-lbl-v0',
+    id='pyraminx-wo-tips-lbl-v1',
     entry_point='rubiks_cube_gym.envs:PyraminxWoTipsEnvLBL',
     max_episode_steps=250,
 )
@@ -37,7 +37,7 @@
 )
 
 register(
-    id='skewb-sarah-v0',
+    id='skewb-sarah-v1',
     entry_point='rubiks_cube_gym.envs:SkewbEnvSarah',
     max_episode_steps=250,
 )
diff --git a/rubiks_cube_gym/envs/pyraminx_wo_tips_lbl.py b/rubiks_cube_gym/envs/pyraminx_wo_tips_lbl.py
@@ -7,7 +7,6 @@ class PyraminxWoTipsEnvLBL(PyraminxWoTipsEnv):
     def __init__(self):
         super(PyraminxWoTipsEnvLBL, self).__init__()
         self.FL = None
-        self.OLL = None
 
     def check_FL(self):
         for pos in FL_POS:
@@ -20,17 +19,17 @@ def check_solved(self):
             return True
 
     def reward(self):
-        reward = -40 * self.FL
+        reward = -25 * self.FL
         done = False
 
         if self.check_FL():
-            reward += 40
+            reward += 25
             self.FL = True
         else:
             self.FL = False
 
         if self.check_solved():
-            reward += 60
+            reward += 75
             done = True
 
         if reward <= 0:

diff --git a/rubiks_cube_gym/envs/skewb_sarah.py b/rubiks_cube_gym/envs/skewb_sarah.py
@@ -19,17 +19,17 @@ def check_solved(self):
             return True
 
     def reward(self):
-        reward = -40 * self.FL
+        reward = -25 * self.FL
         done = False
 
         if self.check_FL():
-            reward += 40
+            reward += 25
             self.FL = True
         else:
             self.FL = False
 
         if self.check_solved():
-            reward += 60
+            reward += 75
             done = True
 
         if reward <= 0:

diff --git a/setup.py b/setup.py
@@ -4,7 +4,7 @@
     long_description = f.read()
 
 setup(name='rubiks_cube_gym',
-      version='0.3.1',
+      version='0.4.0',
       url="https://github.com/DoubleGremlin181/RubiksCubeGym/",
       description="OpenAI Gym environments for various twisty puzzles",
       long_description=long_description,