oil_predictor_tests.py

import numpy as np
import matplotlib.pyplot as plt
import quandl
from datetime import datetime
from statsmodels.tsa.arima_model import ARIMA
import pandas as pd
import os

from pandas.plotting import register_matplotlib_converters

register_matplotlib_converters()  # <-- Something about you having to explicity register your datatime converters or matplotlib has a fit.


def validateDateString(date_string):
    """
	Validates whether date_string is in the correct format.
	"""

    result = None

    try:
        datetime.strptime(date_string, "%Y%m%d")
        result = True
    except ValueError:
        result = False
    return result


# Set quandl API key
quandl.ApiConfig.api_key = "NszGhwY_Qh8Ubj1BWhVt"

# Get current date and convert to string:
now = datetime.now()
string_date = now.strftime("%Y%m%d")

# Searches working directory for date oil price files and retreives dates:
directory_contents = os.listdir()
dates = [0]
for item in directory_contents:

    split = item.split(".")
    extension = split[-1]
    name = split[0][0:10]
    date = split[0][11:19]

    extension_check = extension == "npy"
    name_check = name == "oil_prices"
    date_check = validateDateString(date)

    if extension_check & name_check & date_check:
        dates.append(int(date))

# If latest oil price file has been created today load it, else create a new file:
data = None
if int(string_date) > np.max(dates):
    data = quandl.get("EIA/PET_RWTC_D", returns="numpy")
    np.save("oil_prices_{}".format(string_date), data)
else:
    data = np.load("oil_prices_{}.npy".format(string_date), data)

plt.plot(data["Date"], data["Value"])
plt.savefig("test_oil_prices.png")

# Fit ARIMA model to the data
model = ARIMA(data["Value"], order=(5, 1, 0))
model_fit = model.fit(disp=0)
print(model_fit.summary())

# plot residual errors
residuals = pd.DataFrame(model_fit.resid)
residuals.plot()
plt.savefig("residuals_1")
residuals.plot(kind="kde")
plt.savefig("residuals_2")
print(residuals.describe())

## Predictions:

X = data["Value"]
size = int(len(X) * 0.66)
train, test = X[0:size], X[size : len(X)]

history = [x for x in train]
predictions = list()
for t in range(len(test)):
    model = ARIMA(history, order=(5,1,0))
    model_fit = model.fit(disp=0)
    output = model_fit.forecast()
    yhat = output[0]
    predictions.append(yhat)
    obs = test[t]
    history.append(obs)
    print('%s, Predicted = %f, Expected = %f, Difference = %f' % (str(data['Date'][t + size])[0:10], yhat, obs, (yhat-obs)) )