PeakFinder — Peak finding with derivatives.

This module defines a peak finding utility using the derivative of the spectral line profile.

class admit.util.peakfinder.PeakFinder.PeakFinder(spec, x=None, **kwarg)

PeakFinder searches for spectral peaks by taking the first derivative of the spectrum and looks for zero crossings. Noise spikes are eliminated by a noise cutoff, minimum separation of points, and minimum width of lines.

Parameters:

spec : List or numpy array

The spectrum to be analyzed.

x : List or numpy array, optional

The x co-ordinates for the spectrum. Default: None.

kwarg : Dict

Any additional arguments, see the Attributes list for a complete listing.

Attributes

spec	(numpy array) The spectrum to be analyzed.
x	(numpy array) The x co-ordinates of the spectrum.
thresh	(float, optional) The cutoff used to determine if a peak is above the noise. The absolute value of the spectrum is compared to this so that absorption lines are also detected. Default: 0.0.
min_sep	(int) The minimum separation between peaks in channels. Default: 5.
min_width	(int) The minimum width of a line to consider, in channels. Default: 5.

Methods

find()	Method to locate peaks in an input spectrum
wideenough(pk[, mult])	Method to determine whether a line is wide enough, based on the given parameters.

find()

Method to locate peaks in an input spectrum

Parameters:None Returns:Numpy array containing the located peaks

min_sep = 5

min_width = 5

thresh = 0.0

wideenough(pk, mult=1)

Method to determine whether a line is wide enough, based on the given parameters.

Parameters:

pk : int

The peak to examine (in channel units)

mult : int

A mulitpiler to use for the noise level cutoff Default: 1

Returns:

True of the line meets the minimum width criteria in self.min_width, False

otherwise