diff --git a/eye_catcher_plot.png b/eye_catcher_plot.png deleted file mode 100644 index f1d5e88..0000000 Binary files a/eye_catcher_plot.png and /dev/null differ diff --git a/eye_catcher_plot.py b/eye_catcher_plot.py deleted file mode 100644 index 21883bb..0000000 --- a/eye_catcher_plot.py +++ /dev/null @@ -1,367 +0,0 @@ -import pandas as pd -import numpy as np -import matplotlib.pyplot as plt -import seaborn as sns -import os -import glob -import argparse -from pathlib import Path - -def parse_arguments(): - parser = argparse.ArgumentParser(description='Cross-experiment analysis of chain performance.') - parser.add_argument('--experiments-dir', '-e', required=True, - help='Path to directory containing experiment subdirectories') - parser.add_argument('--supplementary', '-s', required=True, - help='Path to supplementary.csv file with input delays') - parser.add_argument('--output', '-o', default='cross_experiment_analysis', - help='Output filename prefix for the plots (will add experiment type and .png)') - parser.add_argument('--experiment-duration', '-d', type=int, default=20, - help='Duration of each experiment in seconds (default: 20)') - return parser.parse_args() - -def load_supplementary_data(supplementary_path): - """Load the supplementary data with input delays and theoretical perfect times for each chain.""" - supp_df = pd.read_csv(supplementary_path) - # Create dictionaries for quick lookup - delay_dict = dict(zip(supp_df['chain'], supp_df['input_delay'])) - - # Load theoretical perfect e2e time (assuming the third column exists) - if len(supp_df.columns) >= 3: - perfect_time_dict = dict(zip(supp_df['chain'], supp_df.iloc[:, 2])) # Third column - return delay_dict, perfect_time_dict - else: - print("Warning: No third column found for theoretical perfect times. Using input_delay as fallback.") - perfect_time_dict = delay_dict.copy() # Fallback to input_delay - return delay_dict, perfect_time_dict - -def calculate_theoretical_max_runs(chain, input_delay_ms, experiment_duration_s): - """Calculate the theoretical maximum number of runs for a chain.""" - runs_per_second = 1000 / input_delay_ms # Convert ms to runs per second - max_runs = runs_per_second * experiment_duration_s - return int(max_runs) - -def load_experiment_data(experiments_dir, delay_dict, perfect_time_dict, experiment_duration): - """Load all experiment data and calculate performance metrics.""" - all_data = [] - - # Find all subdirectories containing results.csv - experiment_dirs = [d for d in Path(experiments_dir).iterdir() - if d.is_dir() and (d / 'results.csv').exists()] - - print(f"Found {len(experiment_dirs)} experiment directories") - - for exp_dir in experiment_dirs: - results_path = exp_dir / 'results.csv' - - try: - df = pd.read_csv(results_path) - - # Extract experiment name (remove timestamp if present) - if 'experiment_name' in df.columns: - exp_name = df['experiment_name'].iloc[0] - exp_name = exp_name.split('-')[0] if '-' in exp_name else exp_name - else: - exp_name = exp_dir.name - - # Group by chain and calculate metrics - for chain, chain_data in df.groupby('chain'): - if chain in delay_dict and chain in perfect_time_dict: - # Calculate theoretical maximum runs - input_delay = delay_dict[chain] - perfect_time = perfect_time_dict[chain] - theoretical_max = calculate_theoretical_max_runs( - chain, input_delay, experiment_duration - ) - - # Calculate actual performance metrics - actual_runs = chain_data['count'].mean() - mean_latency = chain_data['mean'].mean() - std_latency = chain_data['std'].mean() - - # Normalize latency by theoretical perfect time - normalized_latency = mean_latency / perfect_time - - # Calculate percentage of theoretical maximum - completion_percentage = (actual_runs / theoretical_max) * 100 - - if completion_percentage > 100: - print(f"Warning: Completion percentage for {chain} in {exp_name} exceeds 100%: {completion_percentage:.2f}%") - # Cap at 105% for visualization purposes - # This is to avoid visual clutter in the plot - # and to handle cases where the actual runs exceed theoretical max. - # This is a safeguard and should be adjusted based on actual data characteristics. - # In practice, this might indicate an issue with the data or the calculation. - completion_percentage = 105 - - all_data.append({ - 'experiment_type': exp_name, - 'experiment_dir': exp_dir.name, - 'chain': chain, - 'mean_latency_ms': mean_latency, - 'normalized_latency': normalized_latency, - 'std_latency_ms': std_latency, - 'actual_runs': actual_runs, - 'theoretical_max_runs': theoretical_max, - 'completion_percentage': completion_percentage, - 'input_delay_ms': input_delay, - 'perfect_time_ms': perfect_time - }) - else: - missing_info = [] - if chain not in delay_dict: - missing_info.append("input delay") - if chain not in perfect_time_dict: - missing_info.append("perfect time") - print(f"Warning: Chain '{chain}' missing {', '.join(missing_info)} in supplementary data") - - except Exception as e: - print(f"Error processing {results_path}: {e}") - - return pd.DataFrame(all_data) - -def create_visualizations(data_df, output_prefix): - """Create separate visualization plots for each experiment type.""" - plt.style.use('seaborn-v0_8-darkgrid') - - # Get unique experiment types - experiment_types = sorted(data_df['experiment_type'].unique()) - - print(f"Creating {len(experiment_types)} separate plots for experiment types: {experiment_types}") - - created_files = [] - - for exp_type in experiment_types: - # Filter data for this experiment type - exp_data = data_df[data_df['experiment_type'] == exp_type] - - # Get unique chains for this experiment - chains = sorted(exp_data['chain'].unique()) - - # Create color palette for chains - chain_colors = sns.color_palette("husl", len(chains)) - chain_color_map = dict(zip(chains, chain_colors)) - - # Set up the figure - fig, ax = plt.subplots(figsize=(14, 10)) - - # Plot data points for each chain - for chain in chains: - chain_data = exp_data[exp_data['chain'] == chain] - - ax.scatter( - chain_data['completion_percentage'], - chain_data['normalized_latency'], - color=chain_color_map[chain], - label=chain, - s=120, - alpha=0.8, - edgecolors='black', - linewidth=0.8 - ) - - # Set labels and title - ax.set_xlabel('Completion Rate (% of Theoretical Maximum)', fontsize=14, fontweight='bold') - ax.set_ylabel('Normalized Latency (Actual / Theoretical Perfect)', fontsize=14, fontweight='bold') - ax.set_title(f'Performance Analysis: {exp_type}\nNormalized Latency vs Chain Completion Rate', - fontsize=16, fontweight='bold', pad=20) - - # Add grid for better readability - ax.grid(True, alpha=0.3) - - # Set axis limits - ax.set_xlim(0, 107) - ax.set_ylim(bottom=1) - - # Create legend for chains - legend = ax.legend(title='Chain Output', - loc='best', - fontsize=10, - title_fontsize=12, - framealpha=0.9, - fancybox=True, - shadow=True, - bbox_to_anchor=(1.05, 1)) - - # Adjust layout to accommodate legend - plt.tight_layout() - - # Save the plot - safe_exp_name = exp_type.replace('/', '_').replace(' ', '_') - output_path = f"{output_prefix}_{safe_exp_name}.png" - plt.savefig(output_path, dpi=300, bbox_inches='tight') - created_files.append(output_path) - - # Show the plot - plt.show() - - # Close the figure to free memory - plt.close() - - return created_files - -def create_combined_summary_plot(data_df, output_prefix): - """Create a combined summary plot showing all experiment types in subplots.""" - experiment_types = sorted(data_df['experiment_type'].unique()) - n_experiments = len(experiment_types) - - # Calculate subplot grid dimensions - n_cols = min(3, n_experiments) # Max 3 columns - n_rows = (n_experiments + n_cols - 1) // n_cols # Ceiling division - - fig, axes = plt.subplots(n_rows, n_cols, figsize=(6*n_cols, 5*n_rows)) - - # Ensure axes is always a 2D array - if n_rows == 1 and n_cols == 1: - axes = np.array([[axes]]) - elif n_rows == 1: - axes = axes.reshape(1, -1) - elif n_cols == 1: - axes = axes.reshape(-1, 1) - - plt.style.use('seaborn-v0_8-darkgrid') - - for i, exp_type in enumerate(experiment_types): - row = i // n_cols - col = i % n_cols - ax = axes[row, col] - - # Filter data for this experiment type - exp_data = data_df[data_df['experiment_type'] == exp_type] - chains = sorted(exp_data['chain'].unique()) - - # Create color palette for chains - chain_colors = sns.color_palette("husl", len(chains)) - chain_color_map = dict(zip(chains, chain_colors)) - - # Plot data points - for chain in chains: - chain_data = exp_data[exp_data['chain'] == chain] - ax.scatter( - chain_data['completion_percentage'], - chain_data['normalized_latency'], - color=chain_color_map[chain], - s=60, - alpha=0.7, - edgecolors='black', - linewidth=0.5 - ) - - ax.set_title(exp_type, fontsize=12, fontweight='bold') - ax.set_xlabel('Completion Rate (%)', fontsize=10) - ax.set_ylabel('Normalized Latency', fontsize=10) - ax.grid(True, alpha=0.3) - - # Set axis limits for consistency - ax.set_xlim(0, 107) - ax.set_ylim(bottom=1) - - # Hide unused subplots - for i in range(n_experiments, n_rows * n_cols): - row = i // n_cols - col = i % n_cols - axes[row, col].set_visible(False) - - plt.suptitle('Performance Analysis Summary - All Experiment Types\n(Normalized Latency vs Completion Rate)', - fontsize=16, fontweight='bold', y=0.98) - plt.tight_layout() - - summary_output = f"{output_prefix}_summary.png" - plt.savefig(summary_output, dpi=300, bbox_inches='tight') - plt.show() - plt.close() - - return summary_output - -def print_summary_statistics(data_df): - """Print summary statistics for the analysis.""" - print("\n" + "="*80) - print("CROSS-EXPERIMENT ANALYSIS SUMMARY") - print("="*80) - - print(f"\nTotal experiments analyzed: {data_df['experiment_type'].nunique()}") - print(f"Total chains analyzed: {data_df['chain'].nunique()}") - print(f"Total data points: {len(data_df)}") - - print("\nPer Experiment Type Summary:") - exp_summary = data_df.groupby('experiment_type').agg({ - 'completion_percentage': ['mean', 'std', 'min', 'max'], - 'normalized_latency': ['mean', 'std', 'min', 'max'], - 'mean_latency_ms': ['mean', 'std', 'min', 'max'], - 'chain': 'count' - }).round(2) - print(exp_summary) - - print("\nPer Chain Summary:") - chain_summary = data_df.groupby('chain').agg({ - 'completion_percentage': ['mean', 'std'], - 'normalized_latency': ['mean', 'std'], - 'mean_latency_ms': ['mean', 'std'], - 'experiment_type': 'count' - }).round(2) - print(chain_summary) - - # Find best and worst performing combinations - print("\nBest Performance (highest completion rate):") - best_completion = data_df.loc[data_df['completion_percentage'].idxmax()] - print(f" {best_completion['experiment_type']} - {best_completion['chain']}") - print(f" Completion: {best_completion['completion_percentage']:.1f}%, Normalized Latency: {best_completion['normalized_latency']:.2f}x, Raw Latency: {best_completion['mean_latency_ms']:.1f}ms") - - print("\nWorst Performance (lowest completion rate):") - worst_completion = data_df.loc[data_df['completion_percentage'].idxmin()] - print(f" {worst_completion['experiment_type']} - {worst_completion['chain']}") - print(f" Completion: {worst_completion['completion_percentage']:.1f}%, Normalized Latency: {worst_completion['normalized_latency']:.2f}x, Raw Latency: {worst_completion['mean_latency_ms']:.1f}ms") - - print("\nBest Normalized Latency (closest to theoretical perfect):") - best_latency = data_df.loc[data_df['normalized_latency'].idxmin()] - print(f" {best_latency['experiment_type']} - {best_latency['chain']}") - print(f" Normalized Latency: {best_latency['normalized_latency']:.2f}x, Completion: {best_latency['completion_percentage']:.1f}%, Raw Latency: {best_latency['mean_latency_ms']:.1f}ms") - - print("\nWorst Normalized Latency (furthest from theoretical perfect):") - worst_latency = data_df.loc[data_df['normalized_latency'].idxmax()] - print(f" {worst_latency['experiment_type']} - {worst_latency['chain']}") - print(f" Normalized Latency: {worst_latency['normalized_latency']:.2f}x, Completion: {worst_latency['completion_percentage']:.1f}%, Raw Latency: {worst_latency['mean_latency_ms']:.1f}ms") - -def main(): - args = parse_arguments() - - print("Starting cross-experiment analysis...") - - # Load supplementary data - print(f"Loading supplementary data from: {args.supplementary}") - delay_dict, perfect_time_dict = load_supplementary_data(args.supplementary) - print(f"Found delay information for {len(delay_dict)} chains") - print(f"Found perfect time information for {len(perfect_time_dict)} chains") - - # Load all experiment data - print(f"Loading experiment data from: {args.experiments_dir}") - data_df = load_experiment_data(args.experiments_dir, delay_dict, perfect_time_dict, args.experiment_duration) - - if data_df.empty: - print("No data found! Please check your paths and file formats.") - return - - print(f"Loaded data for {len(data_df)} experiment-chain combinations") - - # Create individual visualizations for each experiment type - print(f"Creating individual visualizations...") - created_files = create_visualizations(data_df, args.output) - - # Create combined summary plot - print(f"Creating combined summary plot...") - summary_file = create_combined_summary_plot(data_df, args.output) - created_files.append(summary_file) - - # Print summary statistics - print_summary_statistics(data_df) - - # Save detailed data to CSV for further analysis - csv_output = f"{args.output}_detailed_data.csv" - data_df.to_csv(csv_output, index=False) - print(f"\nDetailed data saved to: {csv_output}") - - print(f"\nCreated visualization files:") - for file in created_files: - print(f" - {file}") - -if __name__ == "__main__": - main() \ No newline at end of file diff --git a/eye_catcher_plot_detailed_data.csv b/eye_catcher_plot_detailed_data.csv deleted file mode 100644 index 75cb442..0000000 --- a/eye_catcher_plot_detailed_data.csv +++ /dev/null @@ -1,113 +0,0 @@ -experiment_type,experiment_dir,chain,mean_latency_ms,std_latency_ms,actual_runs,theoretical_max_runs,completion_percentage,input_delay_ms -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/baroA/alt --> /output/flight/cmd,35.42897435897436,6.192051282051281,4.102564102564102,200,2.051282051282051,100 -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/cameraA/raw --> /output/cameraA/mapped,74.52780000000001,6.6764,26.0,100,26.0,200 -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/cameraB/raw --> /output/classifier/classification,98.844,31.873800000000003,32.7,133,24.586466165413537,150 -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/gpsA/fix --> /output/flight/cmd,30.827368421052633,5.925526315789473,4.105263157894737,200,2.0526315789473686,100 -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/imuA/data --> /output/flight/cmd,33.086052631578944,6.046052631578946,4.105263157894737,200,2.0526315789473686,100 -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/lidar/scan --> /output/flight/cmd,23.061794871794874,5.396410256410257,4.102564102564102,200,2.051282051282051,100 -edf_single_timed_20_boosted_10,edf_single_timed_20_boosted_10,/input/operator/commands --> /output/flight/cmd,29.039743589743587,5.063333333333334,4.102564102564102,200,2.051282051282051,100 -edf_single_timed_20,edf_single_timed_20,/input/baroA/alt --> /output/flight/cmd,35.124418604651154,6.056976744186046,3.8372093023255816,200,1.9186046511627908,100 -edf_single_timed_20,edf_single_timed_20,/input/cameraA/raw --> /output/cameraA/mapped,74.8638,6.913800000000001,25.96,100,25.96,200 -edf_single_timed_20,edf_single_timed_20,/input/cameraB/raw --> /output/classifier/classification,99.209,32.5044,32.6,133,24.51127819548872,150 -edf_single_timed_20,edf_single_timed_20,/input/gpsA/fix --> /output/flight/cmd,30.65209302325581,5.489767441860465,3.8372093023255816,200,1.9186046511627908,100 -edf_single_timed_20,edf_single_timed_20,/input/imuA/data --> /output/flight/cmd,32.79976744186047,5.736976744186046,3.8372093023255816,200,1.9186046511627908,100 -edf_single_timed_20,edf_single_timed_20,/input/lidar/scan --> /output/flight/cmd,23.58,4.84953488372093,3.8372093023255816,200,1.9186046511627908,100 -edf_single_timed_20,edf_single_timed_20,/input/operator/commands --> /output/flight/cmd,27.936046511627907,4.471627906976744,3.8372093023255816,200,1.9186046511627908,100 -ros_multi_timed_20,ros_multi_timed_20,/input/baroA/alt --> /output/flight/cmd,77.7224,38.486999999999995,1073.12,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/baroB/alt --> /output/telemetry/radio,67.8986,38.959799999999994,3197.82,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/cameraA/raw --> /output/cameraA/mapped,76.226,8.531600000000001,96.86,100,96.86,200 -ros_multi_timed_20,ros_multi_timed_20,/input/cameraA/raw --> /output/telemetry/radio,183.4386,68.3676,3184.16,100,105.0,200 -ros_multi_timed_20,ros_multi_timed_20,/input/cameraB/raw --> /output/classifier/classification,81.2846,9.4038,128.34,133,96.49624060150376,150 -ros_multi_timed_20,ros_multi_timed_20,/input/gpsA/fix --> /output/flight/cmd,77.76100000000001,37.5182,1073.06,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/gpsB/fix --> /output/telemetry/radio,67.8846,36.48500000000001,3200.64,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/imuA/data --> /output/flight/cmd,77.796,38.1552,1073.18,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/imuB/data --> /output/telemetry/radio,67.9932,37.3536,3199.1,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/lidar/scan --> /output/flight/cmd,68.4622,37.3688,1073.76,200,105.0,100 -ros_multi_timed_20,ros_multi_timed_20,/input/operator/commands --> /output/flight/cmd,68.36500000000001,35.9894,1073.48,200,105.0,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/baroA/alt --> /output/flight/cmd,59.1838775510204,29.148571428571433,769.0204081632653,200,105.0,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/baroB/alt --> /output/telemetry/radio,35.7862,16.4298,174.42,200,87.21,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/cameraA/raw --> /output/cameraA/mapped,74.4708,6.9518,98.46,100,98.46,200 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/cameraA/raw --> /output/telemetry/radio,161.5576,33.4002,174.2,100,105.0,200 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/cameraB/raw --> /output/classifier/classification,80.88102040816325,9.17,130.55102040816325,133,98.158661961025,150 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/gpsA/fix --> /output/flight/cmd,63.50000000000001,31.609387755102045,769.0,200,105.0,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/gpsB/fix --> /output/telemetry/radio,31.027199999999997,17.3174,174.42,200,87.21,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/imuA/data --> /output/flight/cmd,59.050399999999996,28.727199999999996,753.64,200,105.0,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/imuB/data --> /output/telemetry/radio,33.2768,18.6738,174.42,200,87.21,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/lidar/scan --> /output/flight/cmd,53.76862745098039,29.196470588235297,753.8627450980392,200,105.0,100 -edf_multi_timed_20_boosted_500,edf_multi_timed_20_boosted_500,/input/operator/commands --> /output/flight/cmd,58.61078431372549,29.589607843137255,753.8823529411765,200,105.0,100 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/baroA/alt --> /output/flight/cmd,35.738139534883715,5.748372093023256,4.0,200,2.0,100 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/cameraA/raw --> /output/cameraA/mapped,74.54119999999999,6.5826,26.0,100,26.0,200 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/cameraB/raw --> /output/classifier/classification,99.0342,32.3718,32.74,133,24.61654135338346,150 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/gpsA/fix --> /output/flight/cmd,30.709534883720924,5.705581395348838,4.0,200,2.0,100 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/imuA/data --> /output/flight/cmd,33.36255813953488,5.513488372093024,4.0,200,2.0,100 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/lidar/scan --> /output/flight/cmd,23.437441860465114,4.923255813953489,4.0,200,2.0,100 -edf_single_timed_20_boosted_50,edf_single_timed_20_boosted_50,/input/operator/commands --> /output/flight/cmd,28.58860465116279,3.7818604651162793,4.0,200,2.0,100 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/baroA/alt --> /output/flight/cmd,35.773250000000004,5.296,3.9,200,1.95,100 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/cameraA/raw --> /output/cameraA/mapped,74.5806,6.636,26.0,100,26.0,200 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/cameraB/raw --> /output/classifier/classification,99.0724,32.465999999999994,32.84,133,24.691729323308273,150 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/gpsA/fix --> /output/flight/cmd,30.87375,5.2219999999999995,3.9,200,1.95,100 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/imuA/data --> /output/flight/cmd,33.338,5.1255,3.9,200,1.95,100 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/lidar/scan --> /output/flight/cmd,23.27475,4.28075,3.9,200,1.95,100 -edf_single_timed_20_boosted_500,edf_single_timed_20_boosted_500,/input/operator/commands --> /output/flight/cmd,29.66375,4.34175,3.9,200,1.95,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/baroA/alt --> /output/flight/cmd,59.2382,28.2522,774.5,200,105.0,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/baroB/alt --> /output/telemetry/radio,34.864599999999996,14.1428,171.72,200,85.86,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/cameraA/raw --> /output/cameraA/mapped,74.5012,6.878599999999999,98.72,100,98.72,200 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/cameraA/raw --> /output/telemetry/radio,159.89939999999996,31.602999999999998,171.68,100,105.0,200 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/cameraB/raw --> /output/classifier/classification,80.8814,8.9892,130.94,133,98.45112781954887,150 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/gpsA/fix --> /output/flight/cmd,63.6704,30.875800000000005,774.5,200,105.0,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/gpsB/fix --> /output/telemetry/radio,30.210000000000004,15.1086,171.72,200,85.86,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/imuA/data --> /output/flight/cmd,59.562,28.5752,774.5,200,105.0,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/imuB/data --> /output/telemetry/radio,32.63,16.3702,171.72,200,85.86,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/lidar/scan --> /output/flight/cmd,54.1478,28.7682,774.5,200,105.0,100 -edf_multi_timed_20_boosted_10,edf_multi_timed_20_boosted_10,/input/operator/commands --> /output/flight/cmd,58.743599999999994,29.187400000000004,774.5,200,105.0,100 -ros_single_timed_20,ros_single_timed_20,/input/baroA/alt --> /output/flight/cmd,483.0948,63.2838,118.36,200,59.18,100 -ros_single_timed_20,ros_single_timed_20,/input/baroB/alt --> /output/telemetry/radio,467.39779999999996,59.0954,117.08,200,58.540000000000006,100 -ros_single_timed_20,ros_single_timed_20,/input/cameraA/raw --> /output/cameraA/mapped,385.3639999999999,237.4138,94.42,100,94.42,200 -ros_single_timed_20,ros_single_timed_20,/input/cameraA/raw --> /output/telemetry/radio,844.8803999999999,321.32239999999996,115.9,100,105.0,200 -ros_single_timed_20,ros_single_timed_20,/input/cameraB/raw --> /output/classifier/classification,169.2946,66.2818,112.24,133,84.39097744360902,150 -ros_single_timed_20,ros_single_timed_20,/input/gpsA/fix --> /output/flight/cmd,486.62440000000004,64.31439999999999,118.48,200,59.24,100 -ros_single_timed_20,ros_single_timed_20,/input/gpsB/fix --> /output/telemetry/radio,470.49940000000004,61.70079999999999,117.46,200,58.72999999999999,100 -ros_single_timed_20,ros_single_timed_20,/input/imuA/data --> /output/flight/cmd,484.7596,63.57379999999999,118.38,200,59.19,100 -ros_single_timed_20,ros_single_timed_20,/input/imuB/data --> /output/telemetry/radio,468.7872,60.461400000000005,117.16,200,58.58,100 -ros_single_timed_20,ros_single_timed_20,/input/lidar/scan --> /output/flight/cmd,311.30480000000006,43.138400000000004,119.2,200,59.599999999999994,100 -ros_single_timed_20,ros_single_timed_20,/input/operator/commands --> /output/flight/cmd,309.44019999999995,42.80259999999999,119.06,200,59.53000000000001,100 -edf_multi_timed_20,edf_multi_timed_20,/input/baroA/alt --> /output/flight/cmd,59.171400000000006,28.876400000000004,769.32,200,105.0,100 -edf_multi_timed_20,edf_multi_timed_20,/input/baroB/alt --> /output/telemetry/radio,35.613,14.744599999999998,173.64,200,86.82,100 -edf_multi_timed_20,edf_multi_timed_20,/input/cameraA/raw --> /output/cameraA/mapped,74.48700000000001,7.1496,98.68,100,98.68,200 -edf_multi_timed_20,edf_multi_timed_20,/input/cameraA/raw --> /output/telemetry/radio,161.5632,31.481800000000003,173.54,100,105.0,200 -edf_multi_timed_20,edf_multi_timed_20,/input/cameraB/raw --> /output/classifier/classification,80.9862,9.209,130.7,133,98.27067669172932,150 -edf_multi_timed_20,edf_multi_timed_20,/input/gpsA/fix --> /output/flight/cmd,64.0672,32.28,769.32,200,105.0,100 -edf_multi_timed_20,edf_multi_timed_20,/input/gpsB/fix --> /output/telemetry/radio,30.8428,15.6802,173.64,200,86.82,100 -edf_multi_timed_20,edf_multi_timed_20,/input/imuA/data --> /output/flight/cmd,60.100199999999994,29.397199999999994,769.32,200,105.0,100 -edf_multi_timed_20,edf_multi_timed_20,/input/imuB/data --> /output/telemetry/radio,32.967600000000004,16.905,173.64,200,86.82,100 -edf_multi_timed_20,edf_multi_timed_20,/input/lidar/scan --> /output/flight/cmd,53.48519999999999,28.731399999999997,769.28,200,105.0,100 -edf_multi_timed_20,edf_multi_timed_20,/input/operator/commands --> /output/flight/cmd,58.02719999999999,29.204199999999997,769.32,200,105.0,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/baroA/alt --> /output/flight/cmd,59.22893617021277,28.212553191489366,771.1702127659574,200,105.0,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/baroB/alt --> /output/telemetry/radio,34.84583333333333,15.932291666666666,174.875,200,87.4375,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/cameraA/raw --> /output/cameraA/mapped,74.66879999999999,8.4206,98.52,100,98.52,200 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/cameraA/raw --> /output/telemetry/radio,160.73208333333335,32.815000000000005,174.85416666666666,100,105.0,200 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/cameraB/raw --> /output/classifier/classification,81.05958333333332,10.574375,130.64583333333334,133,98.22994987468672,150 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/gpsA/fix --> /output/flight/cmd,62.287659574468094,30.74404255319149,771.1702127659574,200,105.0,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/gpsB/fix --> /output/telemetry/radio,30.258750000000003,16.552916666666665,174.875,200,87.4375,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/imuA/data --> /output/flight/cmd,58.75000000000001,28.383829787234042,771.1702127659574,200,105.0,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/imuB/data --> /output/telemetry/radio,32.87978723404255,17.99340425531915,175.2340425531915,200,87.61702127659575,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/lidar/scan --> /output/flight/cmd,53.87255319148936,29.093617021276597,771.1702127659574,200,105.0,100 -edf_multi_timed_20_boosted_50,edf_multi_timed_20_boosted_50,/input/operator/commands --> /output/flight/cmd,58.9159574468085,29.80680851063829,771.1702127659574,200,105.0,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/baroA/alt --> /output/flight/cmd,59.09428571428571,28.03642857142857,771.4285714285714,200,105.0,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/baroB/alt --> /output/telemetry/radio,34.47714285714286,13.967142857142857,167.28571428571428,200,83.64285714285714,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/cameraA/raw --> /output/cameraA/mapped,74.82000000000001,8.075714285714286,98.71428571428571,100,98.71428571428571,200 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/cameraA/raw --> /output/telemetry/radio,159.91142857142856,31.444999999999997,167.0,100,105.0,200 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/cameraB/raw --> /output/classifier/classification,81.09785714285715,9.535,130.92857142857142,133,98.44253490870031,150 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/gpsA/fix --> /output/flight/cmd,63.925000000000004,30.982857142857142,771.4285714285714,200,105.0,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/gpsB/fix --> /output/telemetry/radio,30.61857142857142,15.20142857142857,167.28571428571428,200,83.64285714285714,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/imuA/data --> /output/flight/cmd,59.98,28.602857142857147,771.4285714285714,200,105.0,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/imuB/data --> /output/telemetry/radio,32.915000000000006,16.017857142857142,167.28571428571428,200,83.64285714285714,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/lidar/scan --> /output/flight/cmd,53.777142857142856,28.602142857142848,771.4285714285714,200,105.0,100 -edf_multi_timed_20_boosted_100,edf_multi_timed_20_boosted_100,/input/operator/commands --> /output/flight/cmd,58.76928571428572,29.2,771.4285714285714,200,105.0,100 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/baroA/alt --> /output/flight/cmd,35.05540540540541,5.996756756756756,4.378378378378378,200,2.189189189189189,100 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/cameraA/raw --> /output/cameraA/mapped,74.62279999999998,6.7632,25.98,100,25.980000000000004,200 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/cameraB/raw --> /output/classifier/classification,99.44580000000002,32.4518,32.7,133,24.586466165413537,150 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/gpsA/fix --> /output/flight/cmd,30.72783783783784,5.941081081081082,4.378378378378378,200,2.189189189189189,100 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/imuA/data --> /output/flight/cmd,32.395135135135135,5.7843243243243245,4.378378378378378,200,2.189189189189189,100 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/lidar/scan --> /output/flight/cmd,23.244864864864862,5.153243243243244,4.378378378378378,200,2.189189189189189,100 -edf_single_timed_20_boosted_100,edf_single_timed_20_boosted_100,/input/operator/commands --> /output/flight/cmd,28.249999999999993,4.603783783783783,4.378378378378378,200,2.189189189189189,100