kieker.tools.opad.timeseries.forecast
Class AbstractRForecaster

java.lang.Object
  kieker.tools.opad.timeseries.forecast.AbstractForecaster<java.lang.Double>
      kieker.tools.opad.timeseries.forecast.AbstractRForecaster

All Implemented Interfaces:

IForecaster<java.lang.Double>

Direct Known Subclasses:

ARIMA101Forecaster, ARIMAForecaster, CrostonForecaster, CSForecaster, ETSForecaster, MeanForecaster, NaiveForecaster, SESRForecaster

public abstract class AbstractRForecaster
extends AbstractForecaster<java.lang.Double>

Convenience class to implement an IForecaster with R.

Since:

1.10

Author:

Andre van Hoorn, Nikolas Herbst, Andreas Eberlein, Tobias Rudolph, Thomas Duellmann

Field Summary

static int

MIN_TS_SIZE_DEFAULT

Constructor Summary

AbstractRForecaster(ITimeSeries<java.lang.Double> historyTimeseries, java.lang.String modelFunc, java.lang.String forecastFunc, ForecastMethod strategy)

AbstractRForecaster(ITimeSeries<java.lang.Double> historyTimeseries, java.lang.String modelFunc, java.lang.String forecastFunc, int confidenceLevel, ForecastMethod strategy)

Method Summary

protected IForecastResult	createNaNForecast(ITimeSeries<java.lang.Double> timeseries, int numForecastSteps)
IForecastResult	forecast(int numForecastSteps) Performs a time series forecast for the given number of steps in the future.
protected java.lang.String	forecastOperationOnResult(java.lang.String varNameForecast)
protected abstract java.lang.String[]	getForecastFuncParams() Returns additional parameters to be appended to the call of the R forecast function.
protected abstract java.lang.String[]	getModelFuncParams() Returns additional parameters to be appended to the call of the R forecast model.
protected java.lang.String	lowerOperationOnResult(java.lang.String varNameForecast)
static java.lang.Double[]	removeNullValues(java.util.List<java.lang.Double> allHistory)
protected boolean	satisfiesInputTSRequirements(ITimeSeries<java.lang.Double> timeSeries) Checks whether the requirements for the input TS are met.
protected java.lang.String	upperOperationOnResult(java.lang.String varNameForecast)

Methods inherited from class kieker.tools.opad.timeseries.forecast.AbstractForecaster

getConfidenceLevel, getTsOriginal, prepareForecastTS, supportsConfidence

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

MIN_TS_SIZE_DEFAULT

public static final int MIN_TS_SIZE_DEFAULT

See Also:

Constant Field Values

Constructor Detail

AbstractRForecaster

public AbstractRForecaster(ITimeSeries<java.lang.Double> historyTimeseries,
                           java.lang.String modelFunc,
                           java.lang.String forecastFunc,
                           ForecastMethod strategy)

Parameters:

historyTimeseries - timeseries

modelFunc - modelFunction

forecastFunc - forecastfunction

strategy - FC strategy

AbstractRForecaster

public AbstractRForecaster(ITimeSeries<java.lang.Double> historyTimeseries,
                           java.lang.String modelFunc,
                           java.lang.String forecastFunc,
                           int confidenceLevel,
                           ForecastMethod strategy)

Parameters:

historyTimeseries - timeseries

modelFunc - modelFunction

forecastFunc - forecastfunction

confidenceLevel - value of confedenclevel

strategy - FC strategy

Method Detail

createNaNForecast

protected IForecastResult createNaNForecast(ITimeSeries<java.lang.Double> timeseries,
                                            int numForecastSteps)

forecast

public final IForecastResult forecast(int numForecastSteps)

Description copied from interface: IForecaster

Performs a time series forecast for the given number of steps in the future.

Parameters:

numForecastSteps - amount of to calculate FC steps

Returns:

ForecastResult

satisfiesInputTSRequirements

protected boolean satisfiesInputTSRequirements(ITimeSeries<java.lang.Double> timeSeries)

Checks whether the requirements for the input TS are met. This default implementation checks whether the length of the time series is greater or equal to 5. This method can be overridden by any forecaster if more specific requirements are needed.

lowerOperationOnResult

protected java.lang.String lowerOperationOnResult(java.lang.String varNameForecast)

Parameters:

varNameForecast - Name FC

Returns:

string loweropresult

upperOperationOnResult

protected java.lang.String upperOperationOnResult(java.lang.String varNameForecast)

Parameters:

varNameForecast - name fc

Returns:

string upperopresult

forecastOperationOnResult

protected java.lang.String forecastOperationOnResult(java.lang.String varNameForecast)

Parameters:

varNameForecast - name FC

Returns:

string operation result

getModelFuncParams

protected abstract java.lang.String[] getModelFuncParams()

Returns additional parameters to be appended to the call of the R forecast model.

Returns:

the parameters or null if none

getForecastFuncParams

protected abstract java.lang.String[] getForecastFuncParams()

Returns additional parameters to be appended to the call of the R forecast function.

Returns:

the parameters or null if none

removeNullValues

public static java.lang.Double[] removeNullValues(java.util.List<java.lang.Double> allHistory)

Parameters:

allHistory - List there null values should be deleted in this function

Returns:

List/Array with no NullValues