Last modified: December 2020

AHELP for CIAO 4.13 Sherpa v1


Context: models


One dimensional Voigt profile.




The Voigt profile is a convolution between a Gaussian distribution a Cauchy-Lorentz distribution [1] , [2] . It is often used in analyzing spectroscopy data.


Example 1

>>> create_model_component("voigt1d", "mdl")
>>> print(mdl)

Create a component of the voigt1d model and display its default parameters. The output is:

   Param        Type          Value          Min          Max      Units
   -----        ----          -----          ---          ---      -----
   mdl.fwhm_g   thawed           10  1.17549e-38  3.40282e+38           
   mdl.fwhm_l   thawed           10            0  3.40282e+38           
   mdl.pos      thawed            0 -3.40282e+38  3.40282e+38           
   mdl.ampl     thawed            1 -3.40282e+38  3.40282e+38           

Example 2

Force the widths of the Gaussian and Lorentzian components to be the same:

>>> mdl = Voigt1D()
>>> mdl.fwhm_l = mdl.fwhm_g / np.sqrt(2 * np.log(2))


The attributes for this object are:

Attribute Definition
fwhm_g The full-width half-maximum (FWHM) of the Gaussian distribution.
fwhm_l The full-width half-maximum of the Lorentzian distribution.
pos The center of the profile.
ampl The amplitude of the profile.


Following [2] , the Voigt profile can be written as:

f(x) = ampl * Re[w(z)] / (sqrt(2 * PI) * sigma)

where Re[w(z)] is the real part of the Faddeeva function [3] and sigma and gamma are parameters of the Gaussian and Lorentzian model respectively:

z = (x - pos + i * gamma) / (sqrt(2) * sigma)
sigma = fhwm_g / sqrt(8 * log(2))
gamma = fwhm_l / 2

One common simplification is to tie the sigma and gamma parameters together, which can be achieved by linking the fwhm_l parameter to fwhm_g with the following equation:

fwhm_l = fwhm_g / sqrt(2 * log(2))

An approximation for the FWHM of the profile, taken from [2] , is

0.5346 fwhm_l + sqrt(0.2166 fwhm_l^2 + fwhm_g^2)


Changes in CIAO

Added in CIAO 4.13

The pseudovoigt1d and voigt1d models have been added in CIAO 4.13 and replace the absorptionvoigt and emissionvoigt models.


See the bugs pages on the Sherpa website for an up-to-date listing of known bugs.

See Also

lorentz1d, pseudovoigt1d