com.github.signaflo.timeseries.model.regression

Class TimeSeriesLinearRegressionBuilder

java.lang.Object

com.github.signaflo.timeseries.model.regression.TimeSeriesLinearRegressionBuilder

public final class TimeSeriesLinearRegressionBuilder
extends Object

A builder for a time series linear regression model.

Constructor Summary

Constructors

Constructor and Description
TimeSeriesLinearRegressionBuilder()

Method Summary

Modifier and Type	Method and Description
TimeSeriesLinearRegression	build()
TimeSeriesLinearRegressionBuilder	externalRegressors(double[]... regressors) Specify prediction variable data for the linear regression model.
TimeSeriesLinearRegressionBuilder	externalRegressors(Matrix regressors) Specify prediction variable data for the linear regression model.
TimeSeriesLinearRegressionBuilder	from(TimeSeriesLinearRegression regression) Copy the attributes of the given regression object to this builder and return this builder.
TimeSeriesLinearRegressionBuilder	hasIntercept(TimeSeriesLinearRegression.Intercept intercept) Specify whether to include an intercept in the regression model.
TimeSeriesLinearRegressionBuilder	response(TimeSeries response) Specify the response, or dependent variable, in the form of a time series.
TimeSeriesLinearRegressionBuilder	seasonal(TimeSeriesLinearRegression.Seasonal seasonal) Specify whether to include a seasonal component in the regression model.
TimeSeriesLinearRegressionBuilder	seasonalCycle(TimePeriod seasonalCycle) Specify the length of time it takes for the seasonal pattern to complete one cycle.
TimeSeriesLinearRegressionBuilder	timeTrend(TimeSeriesLinearRegression.TimeTrend timeTrend) Specify whether to include a time trend in the regression model.

Methods inherited from class java.lang.Object

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

Constructor Detail

TimeSeriesLinearRegressionBuilder

public TimeSeriesLinearRegressionBuilder()

Method Detail

from

public final TimeSeriesLinearRegressionBuilder from(TimeSeriesLinearRegression regression)

Copy the attributes of the given regression object to this builder and return this builder.

Parameters:

regression - the object to copy the attributes from.

Returns:

this builder.

externalRegressors

public TimeSeriesLinearRegressionBuilder externalRegressors(@NonNull
                                                            double[]... regressors)

Specify prediction variable data for the linear regression model. Note that if this method has already been called on this object, then the array of prediction variables will be appended to rather than overwritten. Each element of the two dimensional external regressors outer array is interpreted as a column vector of data for a single prediction variable.

Parameters:

regressors - the external regressors to add to the regression model specification.

Returns:

this builder.

externalRegressors

public TimeSeriesLinearRegressionBuilder externalRegressors(@NonNull
                                                            Matrix regressors)

Specify prediction variable data for the linear regression model. Note that if this method has already been called on this object, then the matrix of prediction variables will be appended to rather than overwritten.

Parameters:

regressors - the external regressors to add to the regression model specification.

Returns:

this builder.

response

public TimeSeriesLinearRegressionBuilder response(@NonNull
                                                  TimeSeries response)

Specify the response, or dependent variable, in the form of a time series.

Parameters:

response - the response, or dependent variable, in the form of a time series.

Returns:

this builder.

hasIntercept

public TimeSeriesLinearRegressionBuilder hasIntercept(@NonNull
                                                      TimeSeriesLinearRegression.Intercept intercept)

Specify whether to include an intercept in the regression model. The default is for an intercept to be included.

Parameters:

intercept - whether or not to include an intercept in the model.

Returns:

this builder.

timeTrend

public TimeSeriesLinearRegressionBuilder timeTrend(@NonNull
                                                   TimeSeriesLinearRegression.TimeTrend timeTrend)

Specify whether to include a time trend in the regression model. The default is for a time trend to be included.

Parameters:

timeTrend - whether or not to include a time trend in the model.

Returns:

this builder.

seasonal

public TimeSeriesLinearRegressionBuilder seasonal(@NonNull
                                                  TimeSeriesLinearRegression.Seasonal seasonal)

Specify whether to include a seasonal component in the regression model. The default is for the seasonal component to be excluded.

Parameters:

seasonal - whether or not to include a seasonal component in the model.

Returns:

this builder.

seasonalCycle

public TimeSeriesLinearRegressionBuilder seasonalCycle(@NonNull
                                                       TimePeriod seasonalCycle)

Specify the length of time it takes for the seasonal pattern to complete one cycle.

Parameters:

seasonalCycle - the length of time it takes for the seasonal pattern to complete one cycle. The default value for this property is one year.

Returns:

this builder.

build

public TimeSeriesLinearRegression build()