lag_prediction.py

import pandas_datareader as pdr
import matplotlib.pyplot as plt 
from statsmodels.tsa.stattools import adfuller
from statsmodels.graphics.tsaplots import plot_acf
import seaborn as sns
import warnings
from statsmodels.tsa.arima_model import ARIMA
from sklearn.metrics import mean_squared_error
    
warnings.filterwarnings("ignore")


# Download the information of closing prices from Yahoo Finance
start_date = '01-Jan-08'
end_date   = '14-Feb-20'
ticker     = 'V' # Set the ticker (Visa in this example)

data = pdr.get_data_yahoo(ticker, start_date, end_date)['Adj Close']

# Plot the Adjusted Close Price
plt.rcParams["figure.figsize"] = (10, 7)
plt.style.use('seaborn-darkgrid')


# Splitting data into train and test set
split = int(len(data)*0.90)
train_set, test_set = data[:split], data[split:]

# Convert test set into dataframe and compute the lag 
tst = test_set.to_frame()
tst['Lag'] = test_set.shift(1, fill_value=0)


error = mean_squared_error(tst[1:]['Adj Close'],tst[1:]['Lag'])
print('Test MSE: {mse}'.format(mse=error))

plt.plot(tst[1:]['Adj Close'])
plt.plot(tst[1:]['Lag'], color='red')
plt.title('Visa Stock Price')
plt.xlabel('Date')
plt.ylabel('Price in dollars')
plt.legend(['test_set','test_set (lag1)'])
plt.show()