import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import LSTM, Dropout, Dense
from tensorflow.keras.callbacks import EarlyStopping
from sklearn.model_selection import train_test_split
import keras_tuner as kt

def model_builder(hp):
    model = Sequential()
    
    # Define hyperparameters
    hp_units_1 = hp.Int('units_1', min_value=32, max_value=128, step=32)
    hp_units_2 = hp.Int('units_2', min_value=32, max_value=128, step=32)
    hp_learning_rate = hp.Choice('learning_rate', values=[1e-2, 1e-3, 1e-4])
    hp_optimizer = hp.Choice('optimizer', values=['adam', 'sgd', 'rmsprop', 'adagrad', 'adadelta', 'adamax', 'nadam'])
    
    # Build the model
    model.add(LSTM(units=hp_units_1, return_sequences=True, input_shape=(1, len(signals_predic))))
    model.add(Dropout(0.2))
    model.add(LSTM(units=hp_units_2))
    model.add(Dropout(0.2))
    model.add(Dense(1))
    
    # Compile the model
    model.compile(optimizer=hp_optimizer, loss='mean_squared_error')

    return model

def yourModel():
    # Load and preprocess the data
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
    
    # Set up hyperparameter tuning
    tuner = kt.Hyperband(model_builder, objective='val_loss', max_epochs=10, factor=3, directory='my_dir', project_name='intro_to_kt')
    stop_early = EarlyStopping(monitor='val_loss', patience=10)
    
    # Perform hyperparameter search
    tuner.search(X_train, y_train, epochs=50, validation_split=0.2, callbacks=[stop_early])
    
    # Get the best hyperparameters
    best_hps = tuner.get_best_hyperparameters(num_trials=1)[0]
    
    # Print the best hyperparameters
    print(f"""
        The hyperparameter search is complete. 
        The optimal number of units in the first LSTM layer is {best_hps.get('units_1')} 
        and the optimal number of units in the second LSTM layer is {best_hps.get('units_2')}.
        The optimal learning rate for the optimizer is {best_hps.get('learning_rate')}.
        The best optimizer used is {best_hps.get('optimizer')}.
    """)

    # Build and train the model with the best hyperparameters
    model = tuner.hypermodel.build(best_hps)
    model.fit(X_train, y_train, epochs=100, batch_size=32, verbose=0)
    
    # Return the trained model
    return model