Update app.py

app.py

@@ -1 +1,67 @@
-dex
+from flask import Flask, request, render_template
+import joblib
+import numpy as np
+from scipy.sparse import hstack
+
+app = Flask(__name__)
+
+# Load saved objects
+model = joblib.load('rf_model.joblib')
+le_target = joblib.load('le_target.joblib')
+encoders = joblib.load('encoders.joblib')
+tfidf = joblib.load('tfidf_vectorizer.joblib')
+
+@app.route('/', methods=['GET', 'POST'])
+def index():
+    prediction = None
+    if request.method == 'POST':
+        # Get form data
+        ingredients = request.form['ingredients'].lower()
+        ingredients = ''.join(c for c in ingredients if c.isalnum() or c.isspace())  # basic cleanup
+
+        diet = request.form['diet']
+        course = request.form['course']
+        region = request.form['region']
+        prep_time = request.form['prep_time']
+        cook_time = request.form['cook_time']
+
+        # Transform ingredients using TF-IDF
+        X_ingredients = tfidf.transform([ingredients])
+
+        # Encode categorical features
+        try:
+            diet_enc = encoders['diet'].transform([diet])[0]
+        except:
+            diet_enc = 0  # fallback or handle unknown categories
+        try:
+            course_enc = encoders['course'].transform([course])[0]
+        except:
+            course_enc = 0
+        try:
+            region_enc = encoders['region'].transform([region])[0]
+        except:
+            region_enc = 0
+
+        # Prepare numeric features, convert to float/int
+        try:
+            prep_time = float(prep_time)
+        except:
+            prep_time = 0.0
+        try:
+            cook_time = float(cook_time)
+        except:
+            cook_time = 0.0
+
+        # Stack all features
+        from scipy.sparse import csr_matrix
+        X_other = csr_matrix([[diet_enc, course_enc, region_enc, prep_time, cook_time]])
+        X_input = hstack([X_ingredients, X_other])
+
+        # Predict
+        pred_encoded = model.predict(X_input)[0]
+        prediction = le_target.inverse_transform([pred_encoded])[0]
+
+    return render_template('index.html', prediction=prediction)
+
+if __name__ == '__main__':
+    app.run(debug=True)