import gradio as gr
import numpy as np
import pickle

# Load trained model
model = pickle.load(open("model.pkl", "rb"))

# Class labels
classes = ["🌸 Setosa", "🌼 Versicolor", "🌺 Virginica"]

# Prediction function
def predict(sepal_length, sepal_width, petal_length, petal_width):
    try:
        features = np.array([[sepal_length, sepal_width, petal_length, petal_width]])
        prediction = model.predict(features)[0]
        probabilities = model.predict_proba(features)[0]

        result = f"Prediction: {classes[prediction]}\n\n"
        result += "Confidence:\n"
        
        for i, prob in enumerate(probabilities):
            result += f"{classes[i]}: {round(prob*100, 2)}%\n"
        
        return result

    except Exception as e:
        return f"Error: {str(e)}"

# Gradio UI
with gr.Blocks(title="Iris Flower Classifier") as demo:
    
    gr.Markdown("## 🌸 Iris Flower Prediction App")
    gr.Markdown("Enter flower measurements to predict the species")

    with gr.Row():
        sepal_length = gr.Number(label="Sepal Length (cm)")
        sepal_width = gr.Number(label="Sepal Width (cm)")

    with gr.Row():
        petal_length = gr.Number(label="Petal Length (cm)")
        petal_width = gr.Number(label="Petal Width (cm)")

    predict_btn = gr.Button("Predict")

    output = gr.Textbox(label="Result")

    predict_btn.click(
        fn=predict,
        inputs=[sepal_length, sepal_width, petal_length, petal_width],
        outputs=output
    )

# Launch (important for Hugging Face)
if __name__ == "__main__":
    demo.launch()