import gradio as gr
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
from sklearn.datasets import fetch_california_housing
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score

# --- Step 1: Load dataset ---
data = fetch_california_housing(as_frame=True)
df = data.frame

X = df.drop(columns=['MedHouseVal'])
y = df['MedHouseVal']

feature_names = X.columns

# --- Step 2: Split dataset ---
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# --- Step 3: Train model ---
model = LinearRegression()
model.fit(X_train, y_train)

# --- Step 4: Evaluate model ---
y_pred = model.predict(X_test)
mae = mean_absolute_error(y_test, y_pred)
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)

# --- Step 5: Graph - Actual vs Predicted ---
def plot_actual_vs_predicted():
    fig, ax = plt.subplots(figsize=(5, 4))
    ax.scatter(y_test, y_pred, alpha=0.5)
    ax.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'r--')
    ax.set_xlabel("Actual Prices")
    ax.set_ylabel("Predicted Prices")
    ax.set_title("Actual vs Predicted Prices")
    fig.canvas.draw()
    img = np.array(fig.canvas.renderer.buffer_rgba())
    plt.close(fig)
    return img

# --- Step 6: Graph - Feature Importance (coefficients) ---
def plot_feature_importance():
    importance = np.abs(model.coef_)
    fig, ax = plt.subplots(figsize=(6, 4))
    sns.barplot(x=importance, y=feature_names, palette="coolwarm")
    plt.title("Feature Importance (Linear Regression Coefficients)")
    fig.canvas.draw()
    img = np.array(fig.canvas.renderer.buffer_rgba())
    plt.close(fig)
    return img

# --- Step 7: Prediction Function ---
def predict_price(MedInc, HouseAge, AveRooms, AveBedrms, Population, AveOccup, Latitude, Longitude):
    features = np.array([[MedInc, HouseAge, AveRooms, AveBedrms, Population, AveOccup, Latitude, Longitude]])
    predicted_price = model.predict(features)[0]
    return f"🏠 Predicted Median House Value: ${predicted_price * 100000:.2f}"

# --- Step 8: Gradio Interface ---
with gr.Blocks() as demo:
    gr.Markdown("## 🏡 House Price Prediction with Graphs")
    gr.Markdown(f"**Model Performance:**  R² = {r2:.2f}, MAE = {mae:.2f}, MSE = {mse:.2f}")

    with gr.Row():
        MedInc = gr.Number(label="Median Income (10k USD)")
        HouseAge = gr.Number(label="House Age (years)")
        AveRooms = gr.Number(label="Average Rooms per House")
        AveBedrms = gr.Number(label="Average Bedrooms per House")
        Population = gr.Number(label="Population in Block")
        AveOccup = gr.Number(label="Average Occupancy")
        Latitude = gr.Number(label="Latitude")
        Longitude = gr.Number(label="Longitude")

    output_text = gr.Textbox(label="Prediction Result")

    predict_btn = gr.Button("🔍 Predict House Price")
    predict_btn.click(
        predict_price,
        inputs=[MedInc, HouseAge, AveRooms, AveBedrms, Population, AveOccup, Latitude, Longitude],
        outputs=output_text
    )

    gr.Markdown("### 📊 Model Performance Graphs")
    gr.Image(value=plot_actual_vs_predicted(), label="Actual vs Predicted")
    gr.Image(value=plot_feature_importance(), label="Feature Importance")

# --- Step 9: Launch the app ---
demo.launch()