set_options: Specification of Smoothing Options

In deseats: Data-Driven Locally Weighted Regression for Trend and Seasonality in TS

Specification of Smoothing Options

Description

Set the smoothing specifications for locally weighted regression for identifying the trend and the seasonality in an equidistant time series.

Usage

set_options(
  order_poly = 1,
  season = NA_real_,
  kernel_fun = "epanechnikov",
  bwidth = NA_real_,
  boundary_method = "extend"
)

Arguments

order_poly	the order of the local polynomials used for estimating the smooth nonparametric trend; the default is 1.
season	the frequency of observations per time unit, for example per year; set to 12 for monthly data and to 4 for quarterly data and so on; the default is NA_real_, which leads to an automated frequency selection for time series objects in smoothing functions; if the argument is set to NA_real_ and the observations used for smoothing are not formatted as time series objects, the frequency 1 will be used.
kernel_fun	the weighting function to consider; supported are four second-order kernel functions with compact support on [-1, 1]; enter "uniform" for the uniform kernel, "epanechnikov" for the Epanechnikov kernel, "bisquare" for the bisquare kernel or "triweight" for the triweight kernel; the default is "epanechnikov".
bwidth	a numeric value that indicates the relative bandwidth to consider in the smoothing process; the default is NA, which then triggers a data-driven selection of an globally optimal bandwidth when the output of this function is passed to a smoothing function.
boundary_method	a single character value; it indicates, what bandwidth method to use at boundary points; for "extend", the default, the smoothing window around boundary points will be extended towards the center of the data; for "shorten", the window width will keep decreasing at boundary points when approaching the very first and the very last observation.

Details

An object of class "smoothing_options" is created that contains all required information to conduct a locally weighted regression for decomposing a seasonal time series. The information include the order of the trend polynomials, the frequency of the observed series, the second-order kernel function to use in the weighting process, the (relative) bandwidth to employ, and the boundary method for the bandwidth.

Value

The function returns an S4 object with the following elements (access via @):

order_poly

identical to the input argument with that name; please see the description of that input argument.

season

identical to the input argument with that name; please see the description of that input argument.

kernel_fun

identical to the input argument with that name; please see the description of that input argument.

bwidth

identical to the input argument with that name; please see the description of that input argument.

boundary_method

identical to the input argument with that name; please see the description of that input argument.

Author(s)

• Dominik Schulz (Research Assistant) (Department of Economics, Paderborn University),
  Author and Package Creator

deseats documentation built on Sept. 11, 2024, 8:24 p.m.