# app.py

import pandas as pd
import numpy as np
import streamlit as st
from xgboost import XGBRegressor
from sklearn.model_selection import train_test_split
from sklearn.metrics import r2_score
from io import BytesIO

# -------------------------
# 🔃 Load Dataset
# -------------------------
@st.cache_data
def load_data():
    df = pd.read_csv("Nigeria Economy Dataset_1990-2022.csv")
    return df

df = load_data()

# -------------------------
# ⚙️ Model Training
# -------------------------
features = ['Agriculture to GDP', 'Industry to GDP', 'Services to GDP', 'Inflation rate', 'Government debt']
target = 'Real GDP'

X = df[features]
y = df[target]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

model = XGBRegressor()
model.fit(X_train, y_train)
y_pred = model.predict(X_test)
score = r2_score(y_test, y_pred)

# -------------------------
# 🎨 Streamlit UI
# -------------------------
st.set_page_config(page_title="Nigeria GDP Predictor", layout="wide")
st.title("🇳🇬 Nigeria GDP Forecasting App")
st.markdown("Use the controls to simulate policy changes and forecast **GDP** using AI.")

st.sidebar.header("📊 Economic Inputs")
agri = st.sidebar.slider("Agriculture Contribution (%)", 0.0, 100.0, 25.0)
indus = st.sidebar.slider("Industry Contribution (%)", 0.0, 100.0, 30.0)
service = st.sidebar.slider("Services Contribution (%)", 0.0, 100.0, 40.0)
inflation = st.sidebar.slider("Inflation Rate (%)", 0.0, 100.0, 15.0)
debt = st.sidebar.slider("Govt Debt (Billion USD)", 0.0, 200.0, 60.0)

# -------------------------
# 🧮 Make Prediction
# -------------------------
input_data = pd.DataFrame([[agri, indus, service, inflation, debt]], columns=features)
gdp_prediction = model.predict(input_data)[0]

st.subheader("🧠 Predicted GDP Result")
st.metric(label="Predicted GDP", value=f"${gdp_prediction:,.2f} Billion USD")
st.caption(f"Model Accuracy (R²): {score:.4f}")

# -------------------------
# 📈 GDP Trend Line Chart
# -------------------------
st.subheader("📉 Historical GDP Trend (Nigeria)")
gdp_df = df[['Year', 'Real GDP']].dropna()
gdp_df = gdp_df.sort_values("Year")
st.line_chart(gdp_df.set_index("Year"))

# -------------------------
# 📥 Download Prediction
# -------------------------
def convert_df_to_csv(data):
    return data.to_csv(index=False).encode('utf-8')

st.subheader("⬇️ Export Prediction")
output_df = input_data.copy()
output_df['Predicted_GDP'] = gdp_prediction
csv_data = convert_df_to_csv(output_df)

st.download_button(
    label="Download Prediction as CSV",
    data=csv_data,
    file_name="gdp_prediction.csv",
    mime='text/csv'
)

# -------------------------
# 🌐 Gradio Integration (optional)
# -------------------------
st.subheader("🤖 Gradio AI Simulator (Embed)")
st.markdown(
    """
    <iframe src="https://your-huggingface-username.hf.space" width="100%" height="600" frameborder="0"></iframe>
    """,
    unsafe_allow_html=True
)