Minor viz changes

Author: john-b-yang

codeclash/analysis/multiplayer/win_change_rate.py

@@ -125,7 +125,8 @@ def analyze_win_change_rate(log_dir: Path, game_pattern: str = "CoreWar.r15.s100
     ax1.set_xlabel("Number of Lead Changes per Tournament", fontsize=18, fontproperties=FONT_BOLD)
     ax1.set_ylabel("Frequency", fontsize=18, fontproperties=FONT_BOLD)
     ax1.tick_params(axis="both", labelsize=14)
-    ax1.legend(prop=FONT_BOLD, fontsize=12)
+    FONT_BOLD.set_size(16)
+    ax1.legend(prop=FONT_BOLD)
     ax1.grid(alpha=0.3)
 
     # Plot 2: Change rates comparison
@@ -142,7 +143,8 @@ def analyze_win_change_rate(log_dir: Path, game_pattern: str = "CoreWar.r15.s100
     # Add mean markers
     for i, data in enumerate([change_rates_6p, change_rates_2p], 1):
         ax2.plot(i, np.mean(data), "D", color="darkred", markersize=8, label="Mean" if i == 1 else "")
-    ax2.legend(prop=FONT_BOLD, fontsize=12)
+    FONT_BOLD.set_size(16)
+    ax2.legend(prop=FONT_BOLD)
 
     # Set tick labels with custom font
     for label in ax2.get_xticklabels():

codeclash/analysis/multiplayer/win_share.py

@@ -116,7 +116,8 @@ def analyze_winner_share(log_dir: Path, game_pattern: str = "CoreWar.r15.s1000")
     ax1.set_xlabel("Winner's Share of Total Points (%)", fontsize=18, fontproperties=FONT_BOLD)
     ax1.set_ylabel("Frequency", fontsize=18, fontproperties=FONT_BOLD)
     ax1.tick_params(axis="both", labelsize=14)
-    ax1.legend(prop=FONT_BOLD, fontsize=12)
+    FONT_BOLD.set_size(16)
+    ax1.legend(prop=FONT_BOLD)
     ax1.grid(alpha=0.3)
 
     # Plot 2: Box plot comparison
@@ -133,7 +134,8 @@ def analyze_winner_share(log_dir: Path, game_pattern: str = "CoreWar.r15.s1000")
     # Add mean markers
     for i, data in enumerate([winner_shares_6p, winner_shares_2p], 1):
         ax2.plot(i, np.mean(data), "D", color="darkred", markersize=8, label="Mean" if i == 1 else "")
-    ax2.legend(prop=FONT_BOLD, fontsize=12)
+    FONT_BOLD.set_size(16)
+    ax2.legend(prop=FONT_BOLD)
 
     # Set tick labels with custom font
     for label in ax2.get_xticklabels():

codeclash/analysis/viz/scatter_codebase_organization.py

@@ -227,7 +227,10 @@ def analyze_per_player_arena(data: list, N: int = 5) -> pd.DataFrame:
 def analyze_per_player(data: list, N: int = 5) -> pd.DataFrame:
     df = analyze_per_player_arena(data, N=N)
     df = df[df["total_files"] > 0]
-    return df.groupby("player")["active_file_ratio"].agg(["mean", "std"]).reset_index()
+    result = df.groupby("player")["active_file_ratio"].agg(["mean", "std", "count"]).reset_index()
+    # Calculate standard error of the mean (SEM)
+    result["sem"] = result["std"] / (result["count"] ** 0.5)
+    return result
 
 
 def calculate_root_clutter_ratio(file_history: dict) -> dict:
@@ -263,7 +266,10 @@ def analyze_root_clutter_per_player(data: list) -> pd.DataFrame:
 
     df = pd.DataFrame(results)
     df = df[df["total_files"] > 0]
-    return df.groupby("player")["root_clutter_ratio"].agg(["mean", "std"]).reset_index()
+    result = df.groupby("player")["root_clutter_ratio"].agg(["mean", "std", "count"]).reset_index()
+    # Calculate standard error of the mean (SEM)
+    result["sem"] = result["std"] / (result["count"] ** 0.5)
+    return result
 
 
 def calculate_churn_concentration(file_history: dict, use_magnitude: bool = False) -> dict:
@@ -324,7 +330,10 @@ def analyze_churn_concentration_per_player(data: list, use_magnitude: bool = Fal
 
     df = pd.DataFrame(results)
     df = df[df["total_churn"] > 0]
-    return df.groupby("player")["churn_concentration"].agg(["mean", "std"]).reset_index()
+    result = df.groupby("player")["churn_concentration"].agg(["mean", "std", "count"]).reset_index()
+    # Calculate standard error of the mean (SEM)
+    result["sem"] = result["std"] / (result["count"] ** 0.5)
+    return result
 
 
 def plot_organization_metrics(file_reuse_df: pd.DataFrame, root_clutter_df: pd.DataFrame):
@@ -355,8 +364,8 @@ def plot_organization_metrics(file_reuse_df: pd.DataFrame, root_clutter_df: pd.D
         plt.errorbar(
             row["mean_clutter"],
             row["mean_reuse"],
-            xerr=row["std_clutter"],
-            yerr=row["std_reuse"],
+            xerr=row["sem_clutter"],
+            yerr=row["sem_reuse"],
             fmt="none",
             ecolor=color,
             elinewidth=1.5,
@@ -472,7 +481,10 @@ def analyze_file_reuse_per_player(data: list) -> pd.DataFrame:
 
     df = pd.DataFrame(results)
     df = df[df["total_files_created"] > 0]
-    return df.groupby("player")["file_reuse_ratio"].agg(["mean", "std"]).reset_index()
+    result = df.groupby("player")["file_reuse_ratio"].agg(["mean", "std", "count"]).reset_index()
+    # Calculate standard error of the mean (SEM)
+    result["sem"] = result["std"] / (result["count"] ** 0.5)
+    return result
 
 
 def calculate_filename_redundancy(file_history: dict) -> dict:
@@ -542,10 +554,10 @@ def calculate_redundancy_over_rounds(file_history: dict) -> list:
 
 def plot_filename_redundancy_over_rounds(redundancy_df: pd.DataFrame):
     """Line plot showing filename redundancy over rounds per model.
-
-    - X: Round number
-    - Y: Filename redundancy ratio (mean across tournaments)
-    - One line per model, colored consistently
+    sc
+        - X: Round number
+        - Y: Filename redundancy ratio (mean across tournaments)
+        - One line per model, colored consistently
     """
     # Aggregate by player and round (mean across all tournaments)
     agg_redundancy = (