Zaktualizuj 'predict.py'

predict.py

@@ -4,7 +4,18 @@ import numpy as np
 import sklearn
 import sklearn.model_selection
 from tensorflow.keras.models import load_model
-from sklearn.metrics import accuracy_score, precision_score, f1_score
+from sklearn.metrics import mean_absolute_error, mean_squared_error
+
+import mlflow
+
+# Wskazujemy ścieżkę do folderu, gdzie zostaną zapisane wyniki MLflow
+mlflow.set_tracking_uri("file:/mlflow")
+
+# Ustawiamy nazwę eksperymentu
+mlflow.set_experiment("nazwa eksperymentu")
+
+feature_cols = ['year', 'mileage', 'vol_engine']
+
 
 feature_cols = ['year', 'mileage', 'vol_engine']
 
@@ -19,16 +30,13 @@ results = pd.DataFrame({'id': test_data['id'], 'year': test_data['year'], 'milea
 results.to_csv('predictions.csv', index=False)
 
 y_true = test_data['price']
-y_pred = y_pred = [round(p[0]) for p in predictions]
+y_pred = [round(p[0]) for p in predictions]
 
-print(y_pred)
-print(y_true)
-
-accuracy = accuracy_score(y_true, y_pred)
-precision = precision_score(y_true, y_pred, average='micro')
-f1 = f1_score(y_true, y_pred, average='micro')
+mae = mean_absolute_error(y_true, y_pred)
+mse = mean_squared_error(y_true, y_pred)
+rmse = np.sqrt(mse)
 
 with open('metrics.txt', 'w') as f:
-    f.write(f"Accuracy: {accuracy:.4f}\n")
-    f.write(f"Micro-average Precision: {precision:.4f}\n")
-    f.write(f"Micro-average F1-score: {f1:.4f}\n")
+    f.write(f"MAE: {mae:.4f}\n")
+    f.write(f"MSE: {mse:.4f}\n")
+    f.write(f"RMSE: {rmse:.4f}\n")