import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import mean_squared_error
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, LSTM

# Fix random seed for reproducibility
tf.random.set_seed(7)

# Function to create the dataset matrix
def create_dataset(dataset, look_back=1):
    dataX, dataY = [], []
    for i in range(len(dataset)-look_back-1):
        a = dataset[i:(i+look_back), 0]
        dataX.append(a)
        dataY.append(dataset[i + look_back, 0])
    return np.array(dataX), np.array(dataY)

def lstm_prediction(file, epochs):
    # Load the dataset
    dataframe = pd.read_csv(file, usecols=[1], engine='python', encoding="big5")
    dataset = dataframe.values
    dataset = dataset.astype('float32')

    # Normalize the dataset
    scaler = MinMaxScaler(feature_range=(0, 1))
    dataset = scaler.fit_transform(dataset)

    # Split into train and test sets
    train_size = int(len(dataset) * 0.8)
    test_size = len(dataset) - train_size
    train, test = dataset[0:train_size,:], dataset[train_size:len(dataset),:]

    # Create the dataset matrix
    look_back = 1
    trainX, trainY = create_dataset(train, look_back)
    testX, testY = create_dataset(test, look_back)

    # Reshape input to be [samples, time steps, features]
    trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
    testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))

    # Create and fit the LSTM network
    model = Sequential()
    model.add(LSTM(4, input_shape=(1, look_back)))
    model.add(Dense(1))
    model.compile(loss='mean_squared_error', optimizer='adam')

    # Set up a callback to update Streamlit during training
    class StreamlitCallback(tf.keras.callbacks.Callback):
        def __init__(self):
            super().__init__()
            self.epoch_bar = st.progress(0)
            self.loss_placeholder = st.empty()

        def on_epoch_end(self, epoch, logs=None):
            self.epoch_bar.progress((epoch + 1) / epochs)
            self.loss_placeholder.text(f'Epoch {epoch + 1}/{epochs}, Loss: {logs["loss"]:.4f}')

    # Fit the model
    model.fit(trainX, trainY, epochs=epochs, batch_size=1, verbose=0, callbacks=[StreamlitCallback()])

    # Make predictions
    trainPredict = model.predict(trainX)
    testPredict = model.predict(testX)

    # Invert predictions
    trainPredict = scaler.inverse_transform(trainPredict)
    trainY = scaler.inverse_transform([trainY])
    testPredict = scaler.inverse_transform(testPredict)
    testY = scaler.inverse_transform([testY])

    # Calculate root mean squared error
    trainScore = np.sqrt(mean_squared_error(trainY[0], trainPredict[:,0]))
    testScore = np.sqrt(mean_squared_error(testY[0], testPredict[:,0]))

    # Prepare the plot
    trainPredictPlot = np.empty_like(dataset)
    trainPredictPlot[:, :] = np.nan
    trainPredictPlot[look_back:len(trainPredict)+look_back, :] = trainPredict

    testPredictPlot = np.empty_like(dataset)
    testPredictPlot[:, :] = np.nan
    testPredictPlot[len(trainPredict)+(look_back*2)+1:len(dataset)-1, :] = testPredict

    plt.figure(figsize=(12, 8))
    plt.plot(scaler.inverse_transform(dataset), label='Original Data', color='blue')
    plt.plot(trainPredictPlot, label='Training Predictions', linestyle='--', color='green')
    plt.plot(testPredictPlot, label='Test Predictions', linestyle='--', color='red')
    plt.xlabel('Time')
    plt.ylabel('Scaled Values')
    plt.title('Original Data and Predictions')
    plt.legend()
    plt.grid(True, linestyle='--', alpha=0.7)

    # Show the plot
    st.pyplot(plt)

    return trainScore, testScore

st.title("LSTM Time Series Prediction")
st.write("Upload a CSV file with time series data for prediction.")

# File uploader
uploaded_file = st.file_uploader("Choose a CSV file", type="csv")

# Number input for epochs
epochs = st.number_input("Enter number of epochs", min_value=1, max_value=1000, value=50, step=1)

if uploaded_file is not None:
    train_score, test_score = lstm_prediction(uploaded_file, epochs)
    st.write(f'Train Score: {train_score:.2f} RMSE')
    st.write(f'Test Score: {test_score:.2f} RMSE')