set_prior

Synopsis

Define a prior function for a particular Sherpa model parameter.

Syntax

set_prior(par, prior)

The Sherpa prior functions constitute a subset of the pyBLoCXS suite of functions; pyBLoCXS is a sophisticated Markov chain Monte Carlo (MCMC) based algorithm designed to carry out Bayesian Low-Count X-ray Spectral (BLoCXS) analysis in the Sherpa environment. The algorithm explores parameter space at a suspected minimum using a predefined Sherpa model to high-energy X-ray spectral data.

By default, pyBLoCXS uses a flat prior defined between the hardcoded parameter minima and maxima. The set_prior() function is used to associate a function or model ('prior') with a thawed fit parameter ('par'). The function signature for 'prior' is of the form lambda x, and can be a Sherpa model or an arbitrary user-defined function.

prior - prior function, of the form lambda x, associated with a Sherpa model parameter
par - Sherpa model parameter name

The list of currently set prior-parameter pairs is returned by the list_priors command (). The prior function associated with a particular Sherpa model parameter may be accessed with get_prior ("ahelp get_prior").

Refer to the pyBLoCXSdocumentation for additional information about the algorithm.

Example 1

set_prior(therm.kT,g1)

Define Gaussian model 'g1' using the Sherpa 'normgauss' model, and set it as the prior function for the temperature parameter of the Sherpa model 'xsapec.therm'.

sherpa> normgauss1d.g1
sherpa> g1.pos=2.5; g1.fwhm=0.5
sherpa> set_prior(therm.kT,g1)

sherpa> set_sampler_opt('defaultprior', False)
sherpa> set_sampler_opt('priorshape', [True, False, False])
sherpa> set_sampler_opt('originalscale', [True, True, True])

Example 2

set_prior(abs1.NH,lognorm)

Define custom function 'lognorm' and set it as the prior for the 'nH' parameter of a Sherpa absorption model.


def lognorm(x, sigma=0.5, norm=1.0, x0=20.):
   xl=numpy.log10(x)+22.
   return
   (norm/numpy.sqrt(2*numpy.pi)/sigma)*numpy.exp(-0.5*(xl-x0)*(xl-x0)/sigma/sigma)


set_prior(abs1.NH,lognorm)
set_sampler_opt('defaultprior', False)
set_sampler_opt('priorshape', [True, False, False])
set_sampler_opt('originalscale', [True, True, True])

Bugs

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