plot with seaborn

plot_metrics.py

@@ -1,5 +1,7 @@
+import seaborn as sns
 import matplotlib.pyplot as plt
 import yaml
+import pandas as pd
 
 # Function to read and parse the file
 def read_metrics(file_path):
@@ -8,47 +10,47 @@ def read_metrics(file_path):
     return data['runs']
 
 # Read metrics from the file
-file_path = 'project/metrics.yaml'  # Replace with the correct file path
+file_path = 'project/metrics.yaml'  # Replace with your file path
 runs = read_metrics(file_path)
 
-# Extracting values
+#Flatten data into a list of dictionaries
-n_tasks = [run['n_task'] for run in runs]
+data = []
-build_times = [float(run['build']['components_build_time']) for run in runs]
+for run in runs:
-gen_times = [float(run['code_gen']['gen_time']) for run in runs]
+    if 'build' in run:
-deploy_times = [float(run['deploy']['components_deploy_time']) for run in runs]
+        data.append({'Task': run['n_task'], 'Type': 'Build Time', 'Time': float(run['build']['components_build_time'])})
 
-# Plotting Build Time
+    if 'code_gen' in run:
-plt.figure(figsize=(8, 6))
+        data.append({'Task': run['n_task'], 'Type': 'Generation Time', 'Time': float(run['code_gen']['gen_time'])})
-ax = plt.gca()  # Get current axis
+    
-p = ax.bar(n_tasks, build_times, color='b')
+    if 'deploy' in run:
-ax.bar_label(p, label_type='edge')
+        data.append({'Task': run['n_task'], 'Type': 'Deployment Time', 'Time': float(run['deploy']['components_deploy_time'])})
-ax.set_title('Build Time')
-ax.set_xlabel('Number of Tasks')
-ax.set_ylabel('Build Time (seconds)')
-plt.tight_layout()
-plt.savefig('build_time_plot.png')  # Save the plot as an image
-plt.close()
 
-# Plotting Code Generation Time
+# Convert to DataFrame
-plt.figure(figsize=(8, 6))
+df = pd.DataFrame(data)
-ax = plt.gca()  # Get current axis
-p = ax.bar(n_tasks, gen_times, color='g')
-ax.bar_label(p, label_type='edge')
-ax.set_title('Code Generation Time')
-ax.set_xlabel('Number of Tasks')
-ax.set_ylabel('Generation Time (seconds)')
-plt.tight_layout()
-plt.savefig('gen_time_plot.png')  # Save the plot as an image
-plt.close()
 
-# Plotting Deployment Time
+# Function to plot and add median labels
-plt.figure(figsize=(8, 6))
+def plot_metric(metric, filename, color):
-ax = plt.gca()  # Get current axis
+    subset = df[df['Type'] == metric]
-p = ax.bar(n_tasks, deploy_times, color='r')
+    plt.figure(figsize=(8, 6))
-ax.bar_label(p, label_type='edge')
+    ax = sns.boxplot(x='Task', y='Time', data=subset, color=color)
-ax.set_title('Deployment Time')
+
-ax.set_xlabel('Number of Tasks')
+    # Add median labels
-ax.set_ylabel('Deployment Time (seconds)')
+    medians = subset.groupby('Task')['Time'].median()
-plt.tight_layout()
+    for i, task in enumerate(medians.index):
-plt.savefig('deploy_time_plot.png')  # Save the plot as an image
+        median_value = medians[task]
-plt.close()
+        ax.text(i, median_value, f'{median_value:.3f}', ha='center', va='center',
+                fontsize=10, color='white', bbox=dict(facecolor='black', alpha=0.6, boxstyle='round,pad=0.3'))
+
+    plt.title(f'{metric} by Number of Tasks')
+    plt.xlabel('Number of Tasks')
+    plt.ylabel('Time (seconds)')
+    plt.tight_layout()
+    plt.savefig(filename)
+    plt.close()
+
+# Plot each metric separately
+plot_metric('Build Time', 'res/build_time_boxplot.png', 'skyblue')
+plot_metric('Generation Time', 'res/gen_time_boxplot.png', 'lightgreen')
+plot_metric('Deployment Time', 'res/deploy_time_boxplot.png', 'salmon')
+
+print('Plots saved successfully!')