app.py

import os
import gradio as gr
import tensorflow as tf
import joblib
import numpy as np
import pandas as pd
import yfinance as yf
from huggingface_hub import hf_hub_download

# This forces TensorFlow to only use the CPU.
os.environ['CUDA_VISIBLE_DEVICES'] = '-1'

# --- 1. Download Model and Scalers from Hugging Face Hub ---
MODEL_REPO = "munem420/stock-forecaster-lstm"
MODEL_FILENAME = "model_lstm.h5"
SCALER_FILENAME = "scalers.joblib"

print("--- Downloading model and scalers ---")
try:
    model_path = hf_hub_download(repo_id=MODEL_REPO, filename=MODEL_FILENAME)
    scalers_path = hf_hub_download(repo_id=MODEL_REPO, filename=SCALER_FILENAME)
    print("✅ Files downloaded successfully.")
except Exception as e:
    print(f"❌ Error downloading files: {e}")
    model_path, scalers_path = None, None

# --- 2. Load the Model and Scalers ---
loaded_model_lstm = None
loaded_scalers = None

if model_path and os.path.exists(model_path):
    try:
        # FIX #1: Added custom_objects to handle the 'mse' metric during loading
        loaded_model_lstm = tf.keras.models.load_model(
            model_path,
            custom_objects={"mse": tf.keras.losses.MeanSquaredError()}
        )
        print("✅ Model loaded successfully.")
    except Exception as e:
        print(f"❌ Error loading model: {e}")

if scalers_path and os.path.exists(scalers_path):
    try:
        loaded_scalers = joblib.load(scalers_path)
        print("✅ Scalers loaded successfully.")
    except Exception as e:
        print(f"❌ Error loading scalers: {e}")

# ... (The rest of the file is unchanged) ...

ticker_to_name = {'ZURVY': 'Zurich Insurance Group AG'}

def get_ticker_from_input(input_name):
    return input_name.upper()

def forecast_stock(input_name, model, scalers_dict, input_width=60):
    if not model or not scalers_dict:
        return "Error: Model or scalers not loaded. The backend may have failed to start."
    ticker = get_ticker_from_input(input_name)
    if not ticker:
        return "Error: Invalid stock ticker."
    print(f"\n--- Generating forecast for {ticker} ---")
    try:
        data_df = yf.download(ticker, period="1y", progress=False)
        if data_df.empty:
             return f"Error: No data found for ticker {ticker}. It may be delisted or invalid."
    except Exception as e:
        return f"Error fetching data for {ticker}: {e}"
    if len(data_df) < input_width:
        return f"Error: Not enough historical data. Need {input_width} days, but only have {len(data_df)}."
    recent_data = data_df.tail(input_width)
    close_prices = recent_data['Close'].values.reshape(-1, 1)
    scaler = scalers_dict.get(ticker)
    if not scaler:
        print(f"Warning: No specific scaler found for {ticker}. Using ZURVY's scaler as a fallback.")
        scaler = scalers_dict.get('ZURVY')
        if not scaler:
            return "Error: Default scaler 'ZURVY' not found."
    scaled_data = scaler.transform(close_prices)
    X_pred = scaled_data.reshape(1, input_width, 1)
    prediction_scaled = model.predict(X_pred, verbose=0)[0][0]
    prediction_actual = scaler.inverse_transform(np.array([[prediction_scaled]]))[0][0]
    last_close = recent_data['Close'].iloc[-1]
    result = (
        f"Last known close for {ticker}: ${last_close:.2f}\n"
        f"Predicted next day's close price: ${prediction_actual:.2f}"
    )
    print(result)
    return result

def predict_api(ticker_symbol):
    return forecast_stock(ticker_symbol, loaded_model_lstm, loaded_scalers)

with gr.Blocks() as app:
    gr.Markdown("This is the backend for the React Stock Forecaster App.")
    ticker_input = gr.Textbox(label="Stock Ticker", visible=False)
    output_text = gr.Textbox(label="Forecast", visible=False)
    ticker_input.submit(predict_api, inputs=[ticker_input], outputs=[output_text], api_name="predict")

app = gr.mount_static_directory(app, "build")

if __name__ == "__main__":
    app.launch()