Update inference.py

inference.py

@@ -1,13 +1,12 @@
 import joblib
 import numpy as np
 
-# تحميل الموديل والـ encoder مرة واحدة
+# تحميل الموديل مرة واحدة
 def load_model():
     model = joblib.load("RandomForestRegressor.joblib")
     le_airports = joblib.load("le_airports.joblib")
     return model, le_airports
 
-
 model, le_airports = load_model()
 
 
@@ -15,61 +14,51 @@ def predict(inputs):
     try:
         data = inputs["data"]
 
-        # استخراج القيم
-        Year = data["Year"][0]
-        Quarter = data["Quarter"][0]
-        Month = data["Month"][0]
-        DayofMonth = data["DayofMonth"][0]
-
-        Origin = data["Origin"][0].upper().strip()
-        Dest = data["Dest"][0].upper().strip()
-
-        CRSDepTime = data["CRSDepTime"][0]
-        DepTime = data["DepTime"][0]
-        DepDelayMinutes = data["DepDelayMinutes"][0]
-        DepDel15 = data["DepDel15"][0]
-
-        time = data["time"][0]
-        tempF = data["tempF"][0]
-        WindChillF = data["WindChillF"][0]
-        humidity = data["humidity"][0]
-        windspeedKmph = data["windspeedKmph"][0]
-        WindGustKmph = data["WindGustKmph"][0]
-        winddirDegree = data["winddirDegree"][0]
-        weatherCode = data["weatherCode"][0]
-        precipMM = data["precipMM"][0]
-        visibility = data["visibility"][0]
-        pressure = data["pressure"][0]
-        cloudcover = data["cloudcover"][0]
-        DewPointF = data["DewPointF"][0]
+        # تنظيف البيانات
+        origin = data["Origin"][0].upper().strip()
+        dest = data["Dest"][0].upper().strip()
 
         # Encoding
-        origin_encoded = le_airports.transform([Origin])[0]
-        dest_encoded = le_airports.transform([Dest])[0]
+        origin_encoded = le_airports.transform([origin])[0]
+        dest_encoded = le_airports.transform([dest])[0]
 
-        # تجهيز البيانات
-        input_features = np.array([[
-            Year, Quarter, Month, DayofMonth,
-            origin_encoded, dest_encoded,
-            CRSDepTime, DepTime, DepDelayMinutes, DepDel15,
-            time, tempF, WindChillF, humidity,
-            windspeedKmph, WindGustKmph, winddirDegree,
-            weatherCode, precipMM, visibility,
-            pressure, cloudcover, DewPointF
+        # تجهيز الفيتشرز
+        features = np.array([[
+            data["Year"][0],
+            data["Quarter"][0],
+            data["Month"][0],
+            data["DayofMonth"][0],
+            origin_encoded,
+            dest_encoded,
+            data["CRSDepTime"][0],
+            data["DepTime"][0],
+            data["DepDelayMinutes"][0],
+            data["DepDel15"][0],
+            data["time"][0],
+            data["tempF"][0],
+            data["WindChillF"][0],
+            data["humidity"][0],
+            data["windspeedKmph"][0],
+            data["WindGustKmph"][0],
+            data["winddirDegree"][0],
+            data["weatherCode"][0],
+            data["precipMM"][0],
+            data["visibility"][0],
+            data["pressure"][0],
+            data["cloudcover"][0],
+            data["DewPointF"][0]
         ]])
 
-        # Prediction
-        predicted_minutes = float(model.predict(input_features)[0])
-        predicted_minutes = max(0.0, round(predicted_minutes, 1))
+        # التنبؤ
+        pred = float(model.predict(features)[0])
+        pred = max(0.0, round(pred, 1))
 
-        status = "Delayed" if predicted_minutes >= 15 else "On Time"
+        status = "Delayed" if pred >= 15 else "On Time"
 
         return {
-            "predicted_delay_minutes": predicted_minutes,
+            "predicted_delay_minutes": pred,
             "status": status
         }
 
     except Exception as e:
-        return {
-            "error": str(e)
-        }
+        return {"error": str(e)}