import streamlit as st
import pandas as pd
from huggingface_hub import hf_hub_download
import joblib

# Download and load the model from Hugging Face Hub
model_path = hf_hub_download(
    repo_id="Sudu1976/tourismpkg_prediction_model", # Corrected repo_id
    filename="tourismpkg_prediction_model_v1.joblib"
)
model = joblib.load(model_path)

# Streamlit UI for Tourism Package Prediction
st.title("Tourism Package Prediction App")
st.write("""
This application predicts whether a customer will purchase the **Wellness Tourism Package** based on their details.
Please enter the required information below to get a prediction.
""")

# User input
age = st.number_input("Age", min_value=18, max_value=100, value=30, step=1)
typeofcontact = st.selectbox("TypeofContact", ["Company Invited", "Self Inquiry"])
citytier = st.number_input("CityTier", min_value=1, max_value=3, value=1, step=1)
occupation = st.selectbox("Occupation", ["Salaried", "Free Lancer", "Small Business", "Large Business"])
gender = st.selectbox("Gender", ["male", "female"])
nrofpersonvisiting = st.number_input("NumberOfPersonVisiting", min_value=1, max_value=8, value=2, step=1)
prfpropertystar = st.number_input("PreferredPropertyStar", min_value=3, max_value=5, value=3, step=1)
maritalstatus = st.selectbox("MaritalStatus", ["Single", "Married", "Unmarried", "Divorced"])
nroftrips = st.number_input("NumberOfTrips", min_value=1, max_value=20, value=3, step=1)
passport =  st.number_input("Passport", min_value=0, max_value=1, value=1, step=1)
designation = st.selectbox("Designation", ["Manager", "Senior Manager", "Executive", "AVP", "VP"])
monthlyincome = st.number_input("MonthlyIncome", min_value=1000, max_value=40000, value=5000, step=100)
csi = st.number_input("PitchSatisfactionScore", min_value=1, max_value=5, value=2, step=1)
productpitched = st.selectbox("ProductPitched", ["Basic", "Standard", "Deluxe", "Super Deluxe", "King"])
nroffups = st.number_input("NumberOfFollowups", min_value=1, max_value=6, value=2, step=1)
pitchduration = st.number_input("DurationOfPitch", min_value=5, max_value=40, value=10, step=1)

# Assemble input into DataFrame
input_data = pd.DataFrame([
    {
    'Age': age,
    'TypeofContact': typeofcontact,
    'CityTier': citytier,
    'Occupation': occupation,
    'Gender': gender,
    'NumberOfPersonVisiting': nrofpersonvisiting,
    'PreferredPropertyStar': prfpropertystar,
    'MaritalStatus': maritalstatus,
    'NumberOfTrips': nroftrips,
    'Passport': passport,
    'Designation': designation,
    'MonthlyIncome': monthlyincome, # Fixed typo here
    'PitchSatisfactionScore': csi,
    'ProductPitched' : productpitched,
    'NumberOfFollowups' : nroffups,
    'DurationOfPitch' :pitchduration

}])

# Prediction
if st.button("Predict Purchase"):
    # The model expects raw categorical features, which its internal preprocessor will handle.
    # The model's predict method should handle the transformation.
    prediction_proba = model.predict_proba(input_data)[:, 1]
    # Using a classification threshold, let's say 0.45, to decide on the class.
    classification_threshold = 0.45
    if prediction_proba[0] >= classification_threshold:
        st.success(f"Prediction: Customer is likely to purchase the Wellness Tourism Package (Probability: {prediction_proba[0]:.2f})")
    else:
        st.info(f"Prediction: Customer is unlikely to purchase the Wellness Tourism Package (Probability: {prediction_proba[0]:.2f})")