import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from statsmodels.tsa.stattools import adfuller
from statsmodels.graphics.tsaplots import plot_acf, plot_pacf
from statsmodels.tsa.ar_model import AutoReg
from sklearn.metrics import mean_squared_error

data = pd.read_csv("/content/airline.csv")

print("Shape of the dataset:", data.shape)
print("First 50 rows of the dataset:")
print(data.head(50))

plt.plot(data['International '].head(50))
plt.title('First 50 values of the "International" column')
plt.xlabel('Index')
plt.ylabel('International Passengers')
plt.show()

rolling_mean_5 = data['International '].rolling(window=5).mean()

print("First 10 values of the rolling mean with window size 5:")
print(rolling_mean_5.head(10))

rolling_mean_10 = data['International '].rolling(window=10).mean()

plt.plot(data['International '], label='Original Data')
plt.plot(rolling_mean_10, label='Rolling Mean (window=10)')
plt.title('Original Data and Fitted Value (Rolling Mean)')
plt.xlabel('Index')
plt.ylabel('International Passengers')
plt.legend()
plt.show()

lag_order = 13
model = AutoReg(data['International '], lags=lag_order)
model_fit = model.fit()

plot_acf(data['International '])
plt.title('Autocorrelation Function (ACF)')
plt.show()

plot_pacf(data['International '])
plt.title('Partial Autocorrelation Function (PACF)')
plt.show()

predictions = model_fit.predict(start=lag_order, end=len(data)-1)

mse = mean_squared_error(data['International '][lag_order:], predictions)
print('Mean Squared Error (MSE):', mse)

plt.plot(data['International '][lag_order:], label='Original Data')
plt.plot(predictions, label='Predictions')
plt.title('AR Model Predictions vs Original Data')
plt.xlabel('Index')
plt.ylabel('International Passengers')
plt.legend()
plt.show()