dfilter: Digital filters

In stepR: Multiscale Change-Point Inference

Digital filters

Description

Create digital filters.

Deprecation warning: This function is mainly used for patchlamp recordings and may be transferred to a specialised package.

Usage

dfilter(type = c("bessel", "gauss", "custom"), param = list(pole = 4, cutoff = 1 / 10),
  len = ceiling(3/param$cutoff))
## S3 method for class 'dfilter'
print(x, ...)

Arguments

type	allows to choose Bessel, Gauss or custom filters
param	for a "bessel" filter a list with entries pole and cutoff giving the filter's number of poles (order) and cut-off frequency, resp.; for a "gauss" filter the filter's bandwidth (standard deviation) as a single numeric; for a custom filter either a numeric vector specifying the filter's kernel or a list with items kern and step of the same length giving the filter's kernel and step-response, resp.
len	filter length (unnecessary for "custom" filters
x	the object
...	for generic methods only

Value

Returns a list of class dfilter that contains elements kern and step, the (digitised) filter kernel and step-response, resp., as well as an element param containing the argument param, for a "bessel" filter alongside the corresponding analogue kernel, step response, power spectrum, and autocorrelation function depending on time or frequency as elements kernfun, stepfun, spectrum, and acfun, resp.

See Also

filter, convolve, BesselPolynomial, Normal, family

Examples

# 6-pole Bessel filter with cut-off frequency 1 / 100, with length 100 (too short!)
dfilter("bessel", list(pole = 6, cutoff = 1 / 100), 100)
# custom filter: running mean of length 3
dfilter("custom", rep(1, 3))
dfilter("custom", rep(1, 3))$kern # normalised!
dfilter("custom", rep(1, 3))$step
# Gaussian filter with bandwidth 3 and length 11 (from -5 to 5)
dfilter("gauss", 3, 11)

stepR documentation built on Oct. 18, 2024, 5:10 p.m.