import pandas as pd
import numpy as np
import tensorflow as tf
import random
import matplotlib.pyplot as plt
import gradio as gr
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import LSTM, Dense, Dropout
from tensorflow.keras.callbacks import EarlyStopping

def create_dataset(data, time_step=1):
    X, y = [], []
    for i in range(len(data) - time_step - 1):
        a = data[i:(i + time_step), :]
        X.append(a)
        y.append(data[i + time_step, :])
    return np.array(X), np.array(y)

def predict_future(model, initial_data, time_step, num_months):
    predictions = []
    current_data = initial_data[-time_step:]

    for _ in range(num_months * 30):
        current_data_reshaped = current_data.reshape(1, time_step, current_data.shape[1])
        next_prediction = model.predict(current_data_reshaped)
        predictions.append(next_prediction[0])
        current_data = np.append(current_data[1:], next_prediction, axis=0)

    return np.array(predictions)

def generate_plots(file):
    seed = 65
    tf.random.set_seed(seed)
    np.random.seed(seed)
    random.seed(seed)

    # Чтение данных из Excel файла
    data = pd.read_excel(file)
    data.dropna(inplace=True)
    data = data[['EUR_RUB', 'GBP_RUB', 'USD_RUB']]

    values = data.values
    scaler = MinMaxScaler()
    scaled_values = scaler.fit_transform(values)

    time_step = 30
    X, y = create_dataset(scaled_values, time_step)
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
    X_train = X_train.reshape(X_train.shape[0], X_train.shape[1], X_train.shape[2])
    X_test = X_test.reshape(X_test.shape[0], X_test.shape[1], X_test.shape[2])

    model = Sequential()
    model.add(LSTM(256, return_sequences=True, input_shape=(X_train.shape[1], X_train.shape[2])))
    model.add(Dropout(0.3))
    model.add(LSTM(128, return_sequences=True))
    model.add(Dropout(0.3))
    model.add(LSTM(64, return_sequences=False))
    model.add(Dropout(0.3))
    model.add(Dense(3))
    model.compile(optimizer='adam', loss='mean_squared_error')

    early_stopping = EarlyStopping(monitor='val_loss', patience=15)
    model.fit(X_train, y_train, epochs=100, batch_size=82, validation_data=(X_test, y_test), callbacks=[early_stopping])

    num_months_to_predict = 6
    future_predictions = predict_future(model, scaled_values[-time_step:], time_step, num_months_to_predict)
    future_predictions_rescaled = scaler.inverse_transform(future_predictions)

    monthly_eur = np.mean(future_predictions_rescaled[:, 0].reshape(-1, 30), axis=1)
    monthly_gbp = np.mean(future_predictions_rescaled[:, 1].reshape(-1, 30), axis=1)
    monthly_usd_rub = np.mean(future_predictions_rescaled[:, 2].reshape(-1, 30), axis=1)

    start_date = pd.to_datetime("30.06.2024", dayfirst=True)  # Начальная дата
    date_range = pd.date_range(start=start_date, periods=len(monthly_eur), freq='ME')

    plt.figure(figsize=(15, 10))

    plt.subplot(3, 1, 1)
    plt.plot(date_range, monthly_usd_rub, label='Курс Доллара к Рублю', color='red', marker='o')
    plt.title('Прогноз курса доллара к Рублю на 6 месяцев')
    plt.xlabel('Месяца')
    plt.ylabel('Цена Доллара в Рублях')
    plt.legend()

    plt.subplot(3, 1, 2)
    plt.plot(date_range, monthly_eur, label='Курс Евро к Рублю', color='blue', marker='o')
    plt.title('Прогноз курса Евро к Рублю на 6 месяцев')
    plt.xlabel('Месяца')
    plt.ylabel('Цена Евро в Рублях')
    plt.legend()

    plt.subplot(3, 1, 3)
    plt.plot(date_range, monthly_gbp, label='Курс Фунтов Стерлингов к Рублю', color='green', marker='o')
    plt.title('Прогноз курс Фунтов Стерлингов к Рублю на 6 месяцев')
    plt.xlabel('Месяца')
    plt.ylabel('Цена Фунтов Стерлингов в Рублях')
    plt.legend()

    plt.tight_layout()

    plot_file_path = "predictions.png"
    plt.savefig(plot_file_path)

    return plot_file_path


def get_currency_values(file, date_input):
    data = pd.read_excel(file)
    data.dropna(inplace=True)

    data['Date'] = pd.to_datetime(data['Date'], dayfirst=True)

    date = pd.to_datetime(date_input, dayfirst=True)
    row = data[data['Date'] == date]

    if not row.empty:
        values = row[['EUR_RUB', 'GBP_RUB', 'USD_RUB']].values.flatten()
        return f"На {date_input}: USD/RUB: {values[2]}, EUR/RUB: {values[0]}, GBP/RUB: {values[1]}"
    else:
        return f"Данные на {date_input} не найдены."


app1 = gr.Interface(
    fn=get_currency_values,
    inputs=[gr.File(label="Загрузите файл Excel"), gr.Text(label="Введите дату (дд.мм.гггг):")],
    outputs=gr.Text(label="Результаты валют")
)

app2 = gr.Interface(
    fn=generate_plots,
    inputs=gr.File(label="Загрузите файл Excel"),
    outputs=gr.Image(type="filepath", label="Графики прогнозов"),
)

demo = gr.TabbedInterface([app1, app2], ["Вывод цены за дату", "Предсказания на 6 месяцев"])
demo.launch()