added txt, data and scripts

Data/Avata_ConvAI_Links.txt

@@ -0,0 +1,58 @@
+
+
+
+Avatare - Conversational AI
+
+Luisa 1 (Kopfhörer) 
+https://x.convai.com/?xpid=65b0cc0b-3f12-478c-94c4-49824bdb2de5&type=unlisted
+
+Luisa 2 (Tastatur) 
+https://x.convai.com/?xpid=752c0f31-8572-4aa9-9b14-c99a90e7c7b2&type=unlisted
+
+Luisa 3 (Maus-Tastatur-Kombi)
+https://x.convai.com/?xpid=759a4548-6e50-4e3d-af0b-43b38128bf13&type=unlisted
+
+Luisa 4 (LED Lampe) 
+https://x.convai.com/?xpid=12478e65-6fea-4472-950e-540b6df8a4a7&type=unlisted
+
+
+
+Robot 1 (Kopfhörer) 
+https://x.convai.com/?xpid=ffca4f69-9e28-41ce-a3c6-67f3c0663205&type=unlisted
+
+Robot 2 (Tastatur) 
+https://x.convai.com/?xpid=f529a1eb-c641-4e0c-863d-225857407011&type=unlisted
+
+Robot 3 (Maus-Tastatur-Kombi)
+https://x.convai.com/?xpid=a465d0c9-451d-4b52-998f-082249eccdcb&type=draft
+
+Robot 4 (LED Lampe) 
+https://x.convai.com/?xpid=a465d0c9-451d-4b52-998f-082249eccdcb&type=draft
+
+
+
+Abstract 1 (Kopfhörer) 
+https://x.convai.com/?xpid=09252c2a-a1ba-46ba-b8a4-82ccee898eb8&type=unlisted
+
+Abstract 2 (Tastatur) 
+https://x.convai.com/?xpid=ed05c1c7-03d5-4d53-b9db-bcf87b62767e&type=unlisted
+
+Abstract 3 (Maus-Tastatur-Kombi)
+https://x.convai.com/?xpid=d51e34b0-2f23-4388-b98f-b75eae220889&type=unlisted
+
+Abstract 4 (LED Lampe) 
+https://x.convai.com/?xpid=21802cc4-1717-4b17-9f29-2d47fc48d4d4&type=unlisted
+
+
+
+Particle 1 (Kopfhörer) 
+https://x.convai.com/?xpid=0249ffab-4ec6-4527-9271-b346297d65f2&type=unlisted
+
+Particle 2 (Tastatur) 
+https://x.convai.com/?xpid=775ef04c-973a-4796-b377-d71465fc8e37&type=unlisted
+
+Particle 3 (Maus-Tastatur-Kombi)
+https://x.convai.com/?xpid=5a107a6c-380a-433e-8002-441d5edfcce1&type=draft
+
+Particle 4 (LED Lampe) 
+https://x.convai.com/?xpid=606ccce7-3edd-4e28-a502-19361e6b1fd9&type=unlisted

Data/scripts/evaluate_distances.py

@@ -0,0 +1,49 @@
+import pandas as pd
+import numpy as np
+from pathlib import Path
+
+def load_csv_with_standard_decimal(file_path):
+    with open(file_path, 'r') as f:
+        scene_line = f.readline().strip()
+        scene_name = scene_line.split(': ')[1].split('_')[0]
+    df = pd.read_csv(file_path, skiprows=1)
+    return df, scene_name
+
+def calculate_distance(row):
+    origin = np.array([row['XR_Origin_X'], row['XR_Origin_Y'], row['XR_Origin_Z']])
+    tracked = np.array([row['TrackedObject_X'], row['TrackedObject_Y'], row['TrackedObject_Z']])
+    return np.linalg.norm(tracked - origin)
+
+def process_position_data(data_dir):
+    data_by_scene = {}
+    all_distances = []
+    csv_files = list(Path(data_dir).glob('**/P*_PositionData*.csv'))
+    for file_path in csv_files:
+        df, scene_name = load_csv_with_standard_decimal(str(file_path))
+        df['Distance'] = df.apply(calculate_distance, axis=1)
+        all_distances.extend(df['Distance'])
+        if scene_name not in data_by_scene:
+            data_by_scene[scene_name] = []
+        data_by_scene[scene_name].extend(df['Distance'])
+    return all_distances, data_by_scene
+
+def print_stats(distances, label):
+    distances = np.array(distances)
+    print(f"\n--- {label} ---")
+    print(f"Count: {len(distances)}")
+    print(f"Mean: {np.mean(distances):.4f}")
+    print(f"Std: {np.std(distances):.4f}")
+    print(f"Min: {np.min(distances):.4f}")
+    print(f"Max: {np.max(distances):.4f}")
+    print(f"25th percentile: {np.percentile(distances, 25):.4f}")
+    print(f"50th percentile (median): {np.percentile(distances, 50):.4f}")
+    print(f"75th percentile: {np.percentile(distances, 75):.4f}")
+
+def main():
+    all_distances, data_by_scene = process_position_data('Recordings')
+    print_stats(all_distances, 'All Scenes Combined')
+    for scene, distances in data_by_scene.items():
+        print_stats(distances, f'Scene: {scene}')
+
+if __name__ == "__main__":
+    main() 

Data/scripts/fix_decimal_csv.py

@@ -0,0 +1,38 @@
+import sys
+import csv
+from pathlib import Path
+
+def fix_decimal_csv(input_path, output_path):
+    with open(input_path, 'r', newline='', encoding='utf-8') as infile, \
+         open(output_path, 'w', newline='', encoding='utf-8') as outfile:
+        reader = csv.reader(infile)
+        writer = csv.writer(outfile)
+        # Write the first line (scene name) as is
+        scene_line = next(reader)
+        writer.writerow(scene_line)
+        # Write the header as is (if present)
+        header = next(reader)
+        writer.writerow(header)
+        for row in reader:
+            # Keep first two columns as is
+            fixed_row = row[:2]
+            # Join every pair of columns from the third onward
+            for i in range(2, len(row), 2):
+                if i+1 < len(row):
+                    fixed_val = f'{row[i]}.{row[i+1]}'
+                    fixed_row.append(fixed_val)
+            writer.writerow(fixed_row)
+
+if __name__ == '__main__':
+    # Find all matching CSVs recursively
+    root = Path('.')
+    out_root = root / 'Recordings'
+    out_root.mkdir(exist_ok=True)
+    csv_files = list(root.rglob('P*_PositionData*.csv'))
+    for csv_file in csv_files:
+        # Compute output path in Recordings dir, preserving subdirs
+        rel_path = csv_file.relative_to(root)
+        out_path = out_root / rel_path
+        out_path.parent.mkdir(parents=True, exist_ok=True)
+        fix_decimal_csv(csv_file, out_path)
+        print(f'Corrected file written to: {out_path}') 

Data/scripts/plot_boxplots.py

@@ -0,0 +1,75 @@
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+from pathlib import Path
+
+plt.rcParams.update({
+    'font.size': 16,
+    'font.family': 'serif',
+    'axes.labelsize': 18,
+    'axes.titlesize': 20,
+    'xtick.labelsize': 14,
+    'ytick.labelsize': 14,
+    'legend.fontsize': 14,
+    'figure.dpi': 300,
+    'axes.linewidth': 1.5,
+    'xtick.direction': 'in',
+    'ytick.direction': 'in',
+    'xtick.major.size': 6,
+    'ytick.major.size': 6,
+    'xtick.minor.size': 3,
+    'ytick.minor.size': 3,
+})
+
+
+def load_csv_with_standard_decimal(file_path):
+    with open(file_path, 'r') as f:
+        scene_line = f.readline().strip()
+        scene_name = scene_line.split(': ')[1].split('_')[0]
+    df = pd.read_csv(file_path, skiprows=1)
+    return df, scene_name
+
+def calculate_distance(row):
+    origin = np.array([row['XR_Origin_X'], row['XR_Origin_Y'], row['XR_Origin_Z']])
+    tracked = np.array([row['TrackedObject_X'], row['TrackedObject_Y'], row['TrackedObject_Z']])
+    return np.linalg.norm(tracked - origin)
+
+def process_position_data(data_dir):
+    data_by_scene = {}
+    all_distances = []
+    csv_files = list(Path(data_dir).glob('**/P*_PositionData*.csv'))
+    for file_path in csv_files:
+        df, scene_name = load_csv_with_standard_decimal(str(file_path))
+        df['Distance'] = df.apply(calculate_distance, axis=1)
+        all_distances.extend(df['Distance'])
+        if scene_name not in data_by_scene:
+            data_by_scene[scene_name] = []
+        data_by_scene[scene_name].extend(df['Distance'])
+    return all_distances, data_by_scene
+
+def plot_boxplots(all_distances, data_by_scene):
+    plt.figure(figsize=(10, 7))
+    data = [distances for _, distances in sorted(data_by_scene.items())]
+    labels = [scene for scene in sorted(data_by_scene.keys())]
+    data = [all_distances] + data
+    labels = ['All Scenes'] + labels
+    box = plt.boxplot(
+        data, labels=labels, showmeans=True, meanline=True, showfliers=False,
+        boxprops=dict(linewidth=2), whiskerprops=dict(linewidth=2), capprops=dict(linewidth=2),
+        medianprops=dict(linewidth=2, color='black'), meanprops=dict(linewidth=2, color='C1')
+    )
+    plt.ylabel('Distance to Tracked Object (units)')
+    plt.xlabel('Scene')
+    plt.title('Distribution of Distances to Tracked Objects by Scene')
+    plt.xticks(rotation=20)
+    plt.grid(axis='y', linestyle='--', linewidth=1, alpha=0.7)
+    plt.tight_layout()
+    plt.savefig('boxplot_distances.png', dpi=600)
+    plt.close()
+
+def main():
+    all_distances, data_by_scene = process_position_data('Recordings')
+    plot_boxplots(all_distances, data_by_scene)
+
+if __name__ == "__main__":
+    main() 

Data/scripts/plot_distances.py

@@ -0,0 +1,76 @@
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+from pathlib import Path
+
+def load_csv_with_standard_decimal(file_path):
+    with open(file_path, 'r') as f:
+        scene_line = f.readline().strip()
+        scene_name = scene_line.split(': ')[1].split('_')[0]
+    df = pd.read_csv(file_path, skiprows=1)
+    return df, scene_name
+
+def process_position_data(data_dir):
+    # Group by scene
+    scene_data = {}
+    csv_files = list(Path(data_dir).glob('**/P*_PositionData*.csv'))
+    for file_path in csv_files:
+        df, scene_name = load_csv_with_standard_decimal(str(file_path))
+        if scene_name not in scene_data:
+            scene_data[scene_name] = []
+        scene_data[scene_name].append(df)
+    return scene_data
+
+def plot_relative_positions(scene_data):
+    scenes = list(scene_data.keys())
+    # Create one figure with 4 subplots
+    fig, axes = plt.subplots(2, 2, figsize=(16, 14))
+    axes = axes.flatten()
+    for idx, (scene, recordings) in enumerate(scene_data.items()):
+        all_rel_x = []
+        all_rel_z = []
+        avg_obj_xs = []
+        avg_obj_zs = []
+        for df in recordings:
+            obj_x = df['TrackedObject_X'].values
+            obj_z = df['TrackedObject_Z'].values
+            avg_obj_x = np.mean(obj_x)
+            avg_obj_z = np.mean(obj_z)
+            avg_obj_xs.append(avg_obj_x)
+            avg_obj_zs.append(avg_obj_z)
+            rel_x = df['XR_Origin_X'].values - avg_obj_x
+            rel_z = df['XR_Origin_Z'].values - avg_obj_z
+            all_rel_x.extend(rel_x)
+            all_rel_z.extend(rel_z)
+        ax = axes[idx]
+        ax.scatter(all_rel_x, all_rel_z, alpha=0.2, c='blue', label='Player Positions (relative)')
+        ax.scatter(0, 0, c='red', s=800, marker='*', label='Agent', edgecolor='black', linewidths=2, zorder=10)
+        ax.set_title(f'Scene: {scene}')
+        ax.set_xlabel('X Position (relative to Agent)')
+        ax.set_ylabel('Z Position (relative to Agent)')
+        ax.legend()
+        ax.axis('equal')
+        ax.grid(True, linestyle='--', alpha=0.7)
+        # Also save individual images
+        plt.figure(figsize=(8, 8))
+        plt.scatter(all_rel_x, all_rel_z, alpha=0.2, c='blue', label='Player Positions (relative)')
+        plt.scatter(0, 0, c='red', s=800, marker='*', label='Agent', edgecolor='black', linewidths=2, zorder=10)
+        plt.title(f'Player Positions Relative to Agent\nScene: {scene}')
+        plt.xlabel('X Position (relative to Agent)')
+        plt.ylabel('Z Position (relative to Agent)')
+        plt.legend()
+        plt.axis('equal')
+        plt.grid(True, linestyle='--', alpha=0.7)
+        plt.tight_layout()
+        plt.savefig(f'room_layout_{scene}.png', dpi=300)
+        plt.close()
+    plt.tight_layout()
+    plt.savefig('room_layout_all_scenes.png', dpi=300)
+    plt.close()
+
+def main():
+    scene_data = process_position_data('Recordings')
+    plot_relative_positions(scene_data)
+
+if __name__ == "__main__":
+    main()