app.py

import gradio as gr
import numpy as np
from tensorflow.keras.models import load_model
import pickle

# Load pre-trained models
model_paths = {
    "Simple RNN": "model_SimpleRNN.keras",
    "GRU": "model_GRU.keras",
    "LSTM": "model_LSTM.keras"
}

models = {name: load_model(path) for name, path in model_paths.items()}

# Load ensemble weights from pickle file
with open("ensemble_weights.pkl", "rb") as f:
    ensemble_weights = pickle.load(f)

# Preprocessing function
def preprocess_input(user_input_text):
    try:
        values = list(map(float, user_input_text.replace('\n', '').split(',')))
        if len(values) != 35:
            return None, "❌ Input must be exactly 35 floats (5 time steps × 7 features)."
        input_array = np.array(values).reshape(1, 5, 7)
        return input_array, None
    except ValueError:
        return None, "❌ Invalid input: Ensure all values are valid floats."

# Prediction function
def predict(input_text, selected_model):
    input_array, error = preprocess_input(input_text)
    if error:
        return error

    if selected_model == "Ensemble":
        preds = []
        for name, model in models.items():
            pred = model.predict(input_array)[0][0]
            weight = ensemble_weights.get(name, 0)
            preds.append(weight * pred)
        prediction = sum(preds)
    else:
        model = models[selected_model]
        prediction = model.predict(input_array)[0][0]

    return f"✅ Predicted Probability of Positive Class: {prediction:.4f}"

# Gradio interface
with gr.Blocks() as demo:
    gr.Markdown("# Live RNN Model Ensemble Tester")
    gr.Markdown("Input 35 comma-separated float values (5 time steps × 7 features)")

    input_text = gr.Textbox(label="Input Features (comma-separated, 35 values)", lines=5)
    selected_model = gr.Dropdown(choices=list(model_paths.keys()) + ["Ensemble"], value="Simple RNN", label="Select Model")
    output = gr.Textbox(label="Output")

    submit_btn = gr.Button("Submit")

    submit_btn.click(fn=predict, inputs=[input_text, selected_model], outputs=output)

    gr.Examples([
        ["0.1,0.2,0.3,0.4,0.5,0.6,0.7,"*5, "Simple RNN"]
    ], inputs=[input_text, selected_model])

# Launch
if __name__ == "__main__":
    demo.launch()