build.kernel: Build a Gaussian-Hermite kernel for convolution of a template...

In oczoske/slacR: R-package to measure v and sigma from VIMOS IFU cubes

Description

This function builds a Gauss-Hermite kernel of order 4. A Gaussian kernel is obtained by setting p[3]=0 and p[4]=0.

Usage

  build.kernel(p, nsig = 8)

Arguments

p	vector of parameters v, sigma, h3, h4 of the Gauss-Hermite series to be built.
nsig	One-sided extent of the Gaussian in units of standard deviation

Details

The parameters p[1] and p[2] are expected to be in units of pixels, hence the physical values have to be divided by the velocity step dv before calling build.kernel.

Value

A data.frame with components

x	pixel values at which the Gaussian is evaluated, centred on 0.
y	the values of the Gaussian kernel

Author(s)

Oliver Czoske

See Also

convolve.spectrum

Examples

  sigma <- 300
  dv <- 50
  nsig <- 8
  kern <- build.kernel(c(0, sigma/dv, 0, 0), nsig)
  plot(kern, type='l')

oczoske/slacR documentation built on May 20, 2019, 8:23 p.m.