updated plots

plot_compared_metrics.py

@@ -0,0 +1,92 @@
+import seaborn as sns
+import matplotlib.pyplot as plt
+import yaml
+import pandas as pd
+import numpy as np
+import os
+
+# Function to read and parse a metric file
+def read_metrics(file_path, label):
+    with open(file_path, 'r') as file:
+        data = yaml.safe_load(file)
+    runs = data['runs']
+    
+    extracted_data = []
+    for run in runs:
+        if 'build' in run:
+            extracted_data.append({'Task': run['n_task'], 'Type': 'Build Time', 'Time': float(run['build']['components_build_time']), 'Source': label})
+
+        if 'code_gen' in run:
+            extracted_data.append({'Task': run['n_task'], 'Type': 'Generation Time', 'Time': float(run['code_gen']['gen_time']), 'Source': label})
+
+        if 'deploy' in run:
+            extracted_data.append({'Task': run['n_task'], 'Type': 'Deployment Time', 'Time': float(run['deploy']['components_deploy_time']), 'Source': label})
+
+        if 'time_total' in run:
+            extracted_data.append({'Task': run['n_task'], 'Type': 'Total Time', 'Time': float(run['time_total']), 'Source': label})
+
+    return extracted_data
+
+# Paths for the two metric files
+file_path_1 = 'project/metrics.yaml'  # First metrics file
+file_path_2 = 'auga/metrics.yaml'  # Second metrics file
+
+# Read and combine the data
+data1 = read_metrics(file_path_1, 'Parallel')
+data2 = read_metrics(file_path_2, 'Sequential')
+
+df = pd.DataFrame(data1 + data2)
+
+# Ensure benchmark directory exists
+os.makedirs('benchmark', exist_ok=True)
+
+# Function to plot boxplot
+def plot_boxplot(metric, filename):
+    subset = df[df['Type'] == metric]
+    plt.figure(figsize=(10, 6))
+    sns.boxplot(x='Task', y='Time', hue='Source', data=subset, showfliers=False)
+
+    plt.ylabel('Time (seconds)')
+    plt.grid(True, linestyle='--', alpha=0.7)
+    plt.legend(title='Source')
+    plt.tight_layout()
+    plt.savefig(f'benchmark/{filename}')
+    plt.close()
+
+# Function to plot line plot with confidence intervals
+def plot_lineplot(metric, filename):
+    subset = df[df['Type'] == metric]
+    plt.figure(figsize=(10, 6))
+    sns.lineplot(x='Task', y='Time', hue='Source', data=subset, errorbar=('ci', 95), linewidth=2, marker='o')
+
+    plt.xlabel('Number of Tasks')
+    plt.ylabel('Time (seconds)')
+    plt.grid(True, linestyle='--', alpha=0.7)
+    plt.legend(title='Source')
+    plt.tight_layout()
+    plt.savefig(f'benchmark/{filename}')
+    plt.close()
+
+# Function to plot bar plot with confidence intervals
+def plot_barplot(metric, filename):
+    subset = df[df['Type'] == metric]
+    plt.figure(figsize=(10, 6))
+    sns.barplot(x='Task', y='Time', hue='Source', data=subset, errorbar=('ci', 95))
+
+    plt.ylabel('Time (seconds)')
+    plt.grid(True, linestyle='--', alpha=0.7)
+    plt.legend(title='Source')
+    plt.tight_layout()
+    plt.savefig(f'benchmark/{filename}')
+    plt.close()
+
+# Generate plots for each metric
+metrics = ['Build Time', 'Generation Time', 'Deployment Time', 'Total Time']
+filenames = ['build_time', 'gen_time', 'deploy_time', 'total_time']
+
+for metric, filename in zip(metrics, filenames):
+    #plot_boxplot(metric, f'{filename}_boxplot.png')
+    #plot_lineplot(metric, f'{filename}_lineplot.png')
+    plot_barplot(metric, f'{filename}_paired_barplot.png')
+
+print('Comparison plots saved successfully in "benchmark" directory!')

plot_metrics.py

@@ -12,7 +12,7 @@ def read_metrics(file_path):
     return data['runs']
 
 # Read metrics from the file
-file_path = 'project/metrics.yaml'  # Replace with your file path
+file_path = 'res/metrics-parallel.yaml'  # Replace with your file path
 runs = read_metrics(file_path)
 
 # Flatten data into a list of dictionaries
@@ -26,6 +26,9 @@ for run in runs:
 
     if 'deploy' in run:
         data.append({'Task': run['n_task'], 'Type': 'Deployment Time', 'Time': float(run['deploy']['components_deploy_time'])})
+    
+    if 'time_total' in run:
+        data.append({'Task': run['n_task'], 'Type': 'Total Time', 'Time': float(run['time_total'])})
 
 # Convert to DataFrame
 df = pd.DataFrame(data)
@@ -91,16 +94,20 @@ def plot_barplot(metric, filename, color):
     plt.close()
 
 # Plot each metric separately
-plot_boxplot('Build Time', 'build_time_boxplot.png', 'skyblue')
-plot_lineplot('Build Time', 'build_time_lineplot.png', 'skyblue')
+#plot_boxplot('Build Time', 'build_time_boxplot.png', 'skyblue')
+#plot_lineplot('Build Time', 'build_time_lineplot.png', 'skyblue')
 plot_barplot('Build Time', 'build_time_barplot.png', 'skyblue')
 
-plot_boxplot('Generation Time', 'gen_time_boxplot.png', 'lightgreen')
-plot_lineplot('Generation Time', 'gen_time_lineplot.png', 'lightgreen')
+#plot_boxplot('Generation Time', 'gen_time_boxplot.png', 'lightgreen')
+#plot_lineplot('Generation Time', 'gen_time_lineplot.png', 'lightgreen')
 plot_barplot('Generation Time', 'gen_time_barplot.png', 'lightgreen')
 
-plot_boxplot('Deployment Time', 'deploy_time_boxplot.png', 'salmon')
-plot_lineplot('Deployment Time', 'deploy_time_lineplot.png', 'salmon')
+#plot_boxplot('Deployment Time', 'deploy_time_boxplot.png', 'salmon')
+#plot_lineplot('Deployment Time', 'deploy_time_lineplot.png', 'salmon')
 plot_barplot('Deployment Time', 'deploy_time_barplot.png', 'salmon')
 
+#plot_boxplot('Total Time', 'total_time_boxplot.png', 'purple')
+#plot_lineplot('Total Time', 'total_time_lineplot.png', 'purple')
+plot_barplot('Total Time', 'total_time_barplot.png', 'orange')
+
 print('Plots saved successfully in "benchmark" directory!')