import os
import joblib
import pandas as pd
import streamlit as st

st.set_page_config(
    page_title="Bank Marketing Prediction",
    page_icon="🏦",
    layout="centered"
)

BASE_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
MODEL_PATH = os.path.join(BASE_DIR, "bank_model.pkl")
COLUMNS_PATH = os.path.join(BASE_DIR, "bank_columns.pkl")


@st.cache_resource
def load_artifacts():
    model = joblib.load(MODEL_PATH)
    columns = joblib.load(COLUMNS_PATH)
    return model, columns


model, model_columns = load_artifacts()

st.title("🏦 Bank Marketing Campaign Prediction")
st.write("Enter customer information to predict whether the customer will subscribe to the campaign.")

age = st.number_input("Age", min_value=18, max_value=100, value=35)
job = st.selectbox("Job", [
    "admin.", "blue-collar", "entrepreneur", "housemaid", "management",
    "retired", "self-employed", "services", "student", "technician",
    "unemployed", "unknown"
])
marital = st.selectbox("Marital Status", ["divorced", "married", "single"])
education = st.selectbox("Education", ["primary", "secondary", "tertiary", "unknown"])
default = st.selectbox("Default", ["no", "yes"])
balance = st.number_input("Balance", value=1000)
housing = st.selectbox("Housing Loan", ["no", "yes"])
loan = st.selectbox("Personal Loan", ["no", "yes"])
contact = st.selectbox("Contact Type", ["cellular", "telephone", "unknown"])
day = st.number_input("Last Contact Day", min_value=1, max_value=31, value=15)
month = st.selectbox("Last Contact Month", [
    "jan", "feb", "mar", "apr", "may", "jun",
    "jul", "aug", "sep", "oct", "nov", "dec"
])
duration = st.number_input("Call Duration (seconds)", min_value=0, value=180)
campaign = st.number_input("Number of Contacts During Campaign", min_value=1, value=1)
pdays = st.number_input("Days Since Last Contact", value=999)
previous = st.number_input("Number of Previous Contacts", min_value=0, value=0)
poutcome = st.selectbox("Previous Campaign Outcome", ["failure", "other", "success", "unknown"])

if st.button("Predict"):
    input_data = pd.DataFrame([{
        "age": age,
        "job": job,
        "marital": marital,
        "education": education,
        "default": default,
        "balance": balance,
        "housing": housing,
        "loan": loan,
        "contact": contact,
        "day": day,
        "month": month,
        "duration": duration,
        "campaign": campaign,
        "pdays": pdays,
        "previous": previous,
        "poutcome": poutcome
    }])

    input_encoded = pd.get_dummies(input_data, drop_first=True)
    input_encoded = input_encoded.reindex(columns=model_columns, fill_value=0)

    prediction = model.predict(input_encoded)[0]
    prediction_proba = model.predict_proba(input_encoded)[0][1] if hasattr(model, "predict_proba") else None

    if prediction == 1:
        st.success("Prediction: Customer is likely to subscribe.")
    else:
        st.error("Prediction: Customer is unlikely to subscribe.")

    if prediction_proba is not None:
        st.info(f"Subscription probability: {prediction_proba:.2%}")