gymcts 1.4.0__tar.gz → 1.4.2__tar.gz

{gymcts-1.4.0/src/gymcts.egg-info → gymcts-1.4.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gymcts
-Version: 1.4.0
+Version: 1.4.2
 Summary: A minimalistic implementation of the Monte Carlo Tree Search algorithm for planning problems fomulated as gymnaisum reinforcement learning environments.
 Author: Alexander Nasuta
 Author-email: Alexander Nasuta <alexander.nasuta@wzl-iqs.rwth-aachen.de>

{gymcts-1.4.0 → gymcts-1.4.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "gymcts"
-version = "1.4.0"
+version = "1.4.2"
 description = "A minimalistic implementation of the Monte Carlo Tree Search algorithm for planning problems fomulated as gymnaisum reinforcement learning environments."
 readme = "README.md"
 authors = [{ name = "Alexander Nasuta", email = "alexander.nasuta@wzl-iqs.rwth-aachen.de" }]

{gymcts-1.4.0 → gymcts-1.4.2}/src/gymcts/gymcts_neural_agent.py

@@ -126,15 +126,15 @@ class GymctsNeuralNode(GymctsNode):
     score_variate: Literal[
         "PUCT_v0",
         "PUCT_v1",
-        "PUTC_v2",
-        "PUTC_v3",
-        "PUTC_v4",
-        "PUTC_v5",
-        "PUTC_v6",
-        "PUTC_v7",
-        "PUTC_v8",
-        "PUTC_v9",
-        "PUTC_v10",
+        "PUCT_v2",
+        "PUCT_v3",
+        "PUCT_v4",
+        "PUCT_v5",
+        "PUCT_v6",
+        "PUCT_v7",
+        "PUCT_v8",
+        "PUCT_v9",
+        "PUCT_v10",
         "MuZero_v0",
         "MuZero_v1",
     ] = "PUCT_v0"
@@ -227,7 +227,7 @@ class GymctsNeuralNode(GymctsNode):
         if not colored:
 
             if not self.is_root():
-                return f"(a={self.action}, N={self.visit_count}, Q_v={self.mean_value:.2f}, best={self.max_value:.2f}, ubc={self.tree_policy_score():.2f})"
+                return f"(a={self.action}, N={self.visit_count}, Q_v={self.mean_value:.2f}, best={self.max_value:.2f}, {GymctsNeuralNode.score_variate}={self.tree_policy_score():.2f})"
             else:
                 return f"(N={self.visit_count}, Q_v={self.mean_value:.2f}, best={self.max_value:.2f}) [root]"
 

{gymcts-1.4.0 → gymcts-1.4.2}/src/gymcts/gymcts_node.py

@@ -61,7 +61,7 @@ class GymctsNode:
         if not colored:
 
             if not self.is_root():
-                return f"(a={self.action}, N={self.visit_count}, Q_v={self.mean_value:.2f}, best={self.max_value:.2f}, ubc={self.tree_policy_score():.2f})"
+                return f"(a={self.action}, N={self.visit_count}, Q_v={self.mean_value:.2f}, best={self.max_value:.2f}, {GymctsNode.score_variate}={self.tree_policy_score():.2f})"
             else:
                 return f"(N={self.visit_count}, Q_v={self.mean_value:.2f}, best={self.max_value:.2f}) [root]"
 
@@ -102,7 +102,7 @@ class GymctsNode:
                  f"{p}N{e}={colorful_value(self.visit_count)}, "
                  f"{p}Q_v{e}={ccu.wrap_with_color_scale(s=mean_val, value=self.mean_value, min_val=root_node.min_value, max_val=root_node.max_value)}, "
                  f"{p}best{e}={colorful_value(self.max_value)}") +
-                (f", {p}ubc{e}={colorful_value(self.tree_policy_score())})" if not self.is_root() else ")"))
+                (f", {p}{GymctsNode.score_variate}{e}={colorful_value(self.tree_policy_score())})" if not self.is_root() else ")"))
 
     def traverse_nodes(self) -> Generator[TGymctsNode, None, None]:
         """

{gymcts-1.4.0 → gymcts-1.4.2/src/gymcts.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gymcts
-Version: 1.4.0
+Version: 1.4.2
 Summary: A minimalistic implementation of the Monte Carlo Tree Search algorithm for planning problems fomulated as gymnaisum reinforcement learning environments.
 Author: Alexander Nasuta
 Author-email: Alexander Nasuta <alexander.nasuta@wzl-iqs.rwth-aachen.de>