Update app.py

app.py

@@ -9,16 +9,10 @@ from datetime import datetime
 with open("anomaly_detector_rf_model.pkl", "rb") as f:
     model = pickle.load(f)
 
-# Function to preprocess and predict anomalies
 def detect_anomalies(df):
-    required_columns = {"transaction_id", "amount", "merchant", "location", "transaction_date"}
-    if not required_columns.issubset(df.columns):
-        missing = required_columns - set(df.columns)
-        raise ValueError(f"Missing required columns: {', '.join(missing)}")
-
     df["log_amount"] = np.log1p(df["amount"])
     df["amount_zscore"] = (df["amount"] - df["amount"].mean()) / df["amount"].std()
-    df["transaction_date"] = pd.to_datetime(df["transaction_date"], errors='coerce')
+    df["transaction_date"] = pd.to_datetime(df["transaction_date"])
     df["day_of_week"] = df["transaction_date"].dt.dayofweek
     df["hour"] = df["transaction_date"].dt.hour
     df["is_weekend"] = df["day_of_week"].isin([5, 6]).astype(int)
@@ -27,6 +21,8 @@ def detect_anomalies(df):
     df["is_anomalous"] = model.predict(df[features])
 
     anomalies = df[df["is_anomalous"] == 1][["transaction_id", "amount", "merchant", "location", "transaction_date"]]
+
+    # Save anomalies to CSV
     csv_path = "/tmp/anomalies.csv"
     anomalies.to_csv(csv_path, index=False)