def forecastSeasonalARIMAWitheXogenousRegressorsModel():
#reading the model whichis trained on the whole dataset
seasonalARIMAWitheXogenousRegressorsModel = readSeasonalARIMAWitheXogenousRegressorsModelForFullDataset(
)
#reading the dataset
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
#reading the dataset with holidays
seasonalARIMAWitheXogenousRegressorsDatasetWithHoliday = importSeasonalARIMAWitheXogenousRegressorsDatasetWithMissingDataForHoliday(
"RestaurantVisitors.csv")
exog_forecast = seasonalARIMAWitheXogenousRegressorsDatasetWithHoliday[
478:][['holiday']]
#forecasting for 38 months
seasonalARIMAWitheXogenousRegressorsForecastedValues = seasonalARIMAWitheXogenousRegressorsModel.predict(
len(seasonalARIMAWitheXogenousRegressorsDataset),
len(seasonalARIMAWitheXogenousRegressorsDataset) + 38,
exog=exog_forecast).rename("SARIMAX(0, 0, 0)x(1, 0, 1, 7) Prediction")
#saving the forecasted values
saveSeasonalARIMAWitheXogenousRegressorsForecastedValues(
seasonalARIMAWitheXogenousRegressorsForecastedValues)
def testIsDatasetStationary():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
#order of p,d,q and P, D, Q is SARIMAX(0,0,0)x(1,0,1,7)
#hence we do not have take diff to check stationarity.
agumentedDickeyFullerTest(
seasonalARIMAWitheXogenousRegressorsDataset["total"])
def preprocess():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
X_train, X_test = splitSeasonalARIMAWitheXogenousRegressorsDataset(
seasonalARIMAWitheXogenousRegressorsDataset)
saveTrainingAndTestingDataset(X_train, X_test)
def determineSARIMAXOrderOfPAndQ():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
# For SARIMA Orders we set seasonal=True and pass in an m value
autoArimaResult = auto_arima(
seasonalARIMAWitheXogenousRegressorsDataset["total"],
seasonal=True,
m=7,
trace=True)
print(autoArimaResult.summary()) #SARIMAX(0,0,0)x(1,0,1,7)
def plotSeasonalARIMAWitheXogenousRegressorsForecastedValues():
#reading the dataset
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
#reading the forecated values
seasonalARIMAWitheXogenousRegressorsForecastedValues = readSeasonalARIMAWitheXogenousRegressorsForecastedValues(
)
#visualizing the forecated values
visualizeSeasonalARIMAWitheXogenousRegressorsForecastedValues(
seasonalARIMAWitheXogenousRegressorsDataset,
seasonalARIMAWitheXogenousRegressorsForecastedValues)
def plotSeasonalARIMAWitheXogenousRegressorsForecastedValuesWithHolidays():
#reading the dataset with holidays
seasonalARIMAWitheXogenousRegressorsDatasetWithHoliday = importSeasonalARIMAWitheXogenousRegressorsDatasetWithMissingDataForHoliday(
"RestaurantVisitors.csv")
#reading the dataset
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
#reading the forecated values
seasonalARIMAWitheXogenousRegressorsForecastedValues = readSeasonalARIMAWitheXogenousRegressorsForecastedValues(
)
#visualizing the forecated values
visualizeSeasonalARIMAWitheXogenousRegressorsForecastedValuesWithHolidays(
seasonalARIMAWitheXogenousRegressorsDataset,
seasonalARIMAWitheXogenousRegressorsForecastedValues,
seasonalARIMAWitheXogenousRegressorsDatasetWithHoliday)
def trainSeasonalARIMAWitheXogenousRegressorsModelOnFullDataset():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
seasonalARIMAWitheXogenousRegressorsDataset[
"total"] = seasonalARIMAWitheXogenousRegressorsDataset["total"].astype(
'float64')
#training model on the whole dataset
seasonalARIMAWitheXogenousRegressorsModel = SARIMAX(
seasonalARIMAWitheXogenousRegressorsDataset['total'],
exog=seasonalARIMAWitheXogenousRegressorsDataset['holiday'],
order=(0, 0, 0),
seasonal_order=(1, 0, 1, 7),
enforce_invertibility=False)
seasonalARIMAWitheXogenousRegressorsModelFitResult = seasonalARIMAWitheXogenousRegressorsModel.fit(
)
saveSeasonalARIMAWitheXogenousRegressorsModelForFullDataset(
seasonalARIMAWitheXogenousRegressorsModelFitResult)
print(seasonalARIMAWitheXogenousRegressorsModelFitResult.summary())
def etsDecomposition():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
visualizeEtsDecomposition(seasonalARIMAWitheXogenousRegressorsDataset)
def plotTheSourceDataWithHoliday():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
visualizeSourceDataPlotWithHolidays(
seasonalARIMAWitheXogenousRegressorsDataset)
def plotPACFPlot():
seasonalARIMAWitheXogenousRegressorsDataset = importSeasonalARIMAWitheXogenousRegressorsDataset(
"RestaurantVisitors.csv")
visualizePACFPlot(seasonalARIMAWitheXogenousRegressorsDataset)