Add app.py for Streamlit app deployment

app.py

@@ -1 +1,126 @@
- 
+import streamlit as st
+import pandas as pd
+import joblib
+from huggingface_hub import hf_hub_download
+
+# ---------------------------------------
+# CONFIG
+# ---------------------------------------
+MODEL_REPO_ID = "SunnyShaurya/tourism-package-model1"
+
+FILES = {
+    "model": "overall_best_model.joblib",
+    "scaler": "scaler.joblib",
+    "cat_cols": "categorical_cols.joblib",
+    "num_cols": "numerical_cols.joblib",
+    "processed_cols": "processed_columns.joblib",
+}
+
+# ---------------------------------------
+# LOAD ARTIFACTS
+# ---------------------------------------
+@st.cache_resource
+def load_artifacts():
+    model = joblib.load(hf_hub_download(MODEL_REPO_ID, FILES["model"]))
+    scaler = joblib.load(hf_hub_download(MODEL_REPO_ID, FILES["scaler"]))
+    categorical_cols = joblib.load(hf_hub_download(MODEL_REPO_ID, FILES["cat_cols"]))
+    numerical_cols = joblib.load(hf_hub_download(MODEL_REPO_ID, FILES["num_cols"]))
+    processed_columns = joblib.load(hf_hub_download(MODEL_REPO_ID, FILES["processed_cols"]))
+    return model, scaler, categorical_cols, numerical_cols, processed_columns
+
+
+model, scaler, categorical_cols, numerical_cols, processed_columns = load_artifacts()
+
+# ---------------------------------------
+# STREAMLIT UI
+# ---------------------------------------
+st.set_page_config(page_title="Tourism Package Prediction", layout="centered")
+st.title("✈️ Tourism Package Prediction")
+st.write("Predict whether a customer will purchase a tourism package.")
+
+with st.form("prediction_form"):
+    col1, col2 = st.columns(2)
+
+    with col1:
+        Age = st.number_input("Age", 18, 100, 30)
+        TypeofContact = st.selectbox("Type of Contact", ["Self Enquiry", "Company Invited"])
+        CityTier = st.selectbox("City Tier", [1, 2, 3])
+        DurationOfPitch = st.number_input("Duration of Pitch", 0, 120, 10)
+        Occupation = st.selectbox(
+            "Occupation",
+            ["Salaried", "Small Business", "Large Business", "Free Lancer", "Government"]
+        )
+        Gender = st.selectbox("Gender", ["Male", "Female", "Fe Male"])
+        NumberOfPersonVisiting = st.number_input("Persons Visiting", 1, 10, 2)
+        NumberOfFollowups = st.number_input("Followups", 0, 10, 3)
+        ProductPitched = st.selectbox(
+            "Product Pitched", ["Basic", "Standard", "Deluxe", "Super Deluxe", "King"]
+        )
+
+    with col2:
+        PreferredPropertyStar = st.number_input("Property Star", 1, 5, 3)
+        MaritalStatus = st.selectbox(
+            "Marital Status", ["Married", "Single", "Unmarried", "Divorced"]
+        )
+        NumberOfTrips = st.number_input("Trips Last Year", 0, 50, 2)
+        Passport = st.selectbox("Passport", [0, 1])
+        PitchSatisfactionScore = st.number_input("Pitch Satisfaction", 1, 5, 3)
+        OwnCar = st.selectbox("Own Car", [0, 1])
+        NumberOfChildrenVisiting = st.number_input("Children Visiting", 0, 10, 1)
+        Designation = st.selectbox(
+            "Designation", ["Executive", "Manager", "Senior Manager", "AVP", "VP"]
+        )
+        MonthlyIncome = st.number_input("Monthly Income", 0, 500000, 25000)
+
+    submitted = st.form_submit_button("Predict")
+
+# ---------------------------------------
+# PREDICTION
+# ---------------------------------------
+if submitted:
+    df = pd.DataFrame([{
+        "Age": Age,
+        "TypeofContact": TypeofContact,
+        "CityTier": CityTier,
+        "DurationOfPitch": DurationOfPitch,
+        "Occupation": Occupation,
+        "Gender": "Female" if Gender == "Fe Male" else Gender,
+        "NumberOfPersonVisiting": NumberOfPersonVisiting,
+        "NumberOfFollowups": NumberOfFollowups,
+        "ProductPitched": ProductPitched,
+        "PreferredPropertyStar": PreferredPropertyStar,
+        "MaritalStatus": MaritalStatus,
+        "NumberOfTrips": NumberOfTrips,
+        "Passport": Passport,
+        "PitchSatisfactionScore": PitchSatisfactionScore,
+        "OwnCar": OwnCar,
+        "NumberOfChildrenVisiting": NumberOfChildrenVisiting,
+        "Designation": Designation,
+        "MonthlyIncome": MonthlyIncome
+    }])
+
+    # 🔑 IMPORTANT: NO drop_first
+    df_encoded = pd.get_dummies(df, columns=categorical_cols)
+
+    # Align with training features
+    df_encoded = df_encoded.reindex(columns=processed_columns, fill_value=0)
+
+    # Scale numerical columns
+    df_encoded[numerical_cols] = scaler.transform(df_encoded[numerical_cols])
+
+    # Debug info
+    st.caption(f"Active features: {(df_encoded != 0).sum().sum()}")
+
+    # Predict
+    pred = model.predict(df_encoded)[0]
+    prob = model.predict_proba(df_encoded)[0][pred]
+
+    st.subheader("Prediction Result")
+    if pred == 1:
+        st.success("✅ Customer is **LIKELY** to purchase the tourism package")
+    else:
+        st.warning("❌ Customer is **UNLIKELY** to purchase the tourism package")
+
+    st.metric("Confidence", f"{prob * 100:.2f}%")
+
+    st.caption("⚠️ ML-based prediction for decision support only.")