The function performs adaptive smoothing by Intersection of Confidence Intervals (ICI) as described in Katkovnik et al (2006)

ICIsmooth(y, hmax, hinc = 1.45, thresh = NULL, kern = "Gaussian", m = 0,
          sigma = NULL, nsector = 1, sector = 1, symmetric = FALSE,
          presmooth = FALSE, unit = c("SD","FWHM"))

Arguments

y

Object of class "array" containing the original (response) data on a grid

hmax

maximum bandwidth

hinc

factor used to increase the bandwidth from scale to scale

thresh

threshold used in tests to determine the best scale

kern

Determines the kernel function. Object of class "character" kernel, can be any of c("Gaussian","Uniform","Triangle","Epanechnicov","Biweight","Triweight"). Defaults to kern="Gaussian".

m

Object of class "integer" vector of length length(dy) determining the order of derivatives specified for the coordinate directios.

sigma

error standard deviation

nsector

number of sectors to use. Positive weights are restricted to the sector selected by sector

sector

Object of class "integer" between 1 and nsector. sector used.

symmetric

Object of class "logical" determines if sectors are symmetric with respect to the origin.

presmooth

Object of class "logical" determines if bandwidths are smoothed for more stable results.

unit

How should the bandwidth be interpreted in case of a Gaussian kernel. For "SD" the bandwidth refers to the standard deviation of the kernel while "FWHM" interprets the banwidth in terms of Full Width Half Maximum of the kernel.

Details

This mainly follows Chapter 6.1 in Katkovnik et al (2006).

Value

An object of class ICIsmooth

References

V. Katkovnik, K. Egiazarian and J. Astola, Local Approximation Techniques in Signal And Image Processing, SPIE Society of Photo-Optical Instrumentation Engin., 2006, PM157

Author

Joerg Polzehl polzehl@wias-berlin.de

See also