IUM_06 - update evaluation/Jenkinsfile, update predict.py - metrics plot

evaluation/Jenkinsfile

@@ -34,7 +34,7 @@ pipeline {
             steps {
                 sh "chmod +x ./predict.py"
                 sh "python3 ./predict.py"
-                archiveArtifacts artifacts: 'predictions.csv, metrics.csv', onlyIfSuccessful: true
+                archiveArtifacts artifacts: 'predictions.csv, metrics.csv, metrics.png', onlyIfSuccessful: true
             }
         }
     }

predict.py

@@ -5,6 +5,9 @@ import numpy as np
 
 from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
 
+import matplotlib.pyplot as plt
+import seaborn as sns
+
 from NeuralNetwork import NeuralNetwork
 
 # Load model if it exists
@@ -42,6 +45,18 @@ if os.path.exists('./models/model.pth'):
             pd.DataFrame([[accuracy, precision, recall, f1]], columns=['Accuracy', 'Precision', 'Recall', 'F1']).to_csv('metrics.csv', index=False)
         else:
             # without header
-            pd.DataFrame([[accuracy, precision, recall, f1]], columns=['Accuracy', 'Precision', 'Recall', 'F1']).to_csv('metrics.csv', index=False, mode='a', header=False)
+            metrics = pd.read_csv('metrics.csv')
+            metrics = metrics._append({'Accuracy': accuracy, 'Precision': precision, 'Recall': recall, 'F1': f1}, ignore_index=True)
+            metrics.to_csv('metrics.csv', index=False, mode='a', header=False)
+
+            # Plot metrics line chart
+            sns.set(style='whitegrid')
+            plt.figure(figsize=(8, 6))
+            sns.lineplot(data=metrics)
+            plt.title('Metrics history')
+            plt.xlabel('History number')
+            plt.ylabel('Value')
+            plt.legend()
+            plt.savefig('metrics.png')
 else:
     raise FileNotFoundError('Model not found')