phitter.fit.prior#

Module Contents#

Classes#

uniform_prior

Uniform distribution prior

gaussian_prior

Gaussian / normal distribution prior

multivariate_gaussian_prior

Multivariate Gaussian / normal distribution prior

const_prior

Constant value prior

prior_collection

Collection of prior objects. Transformation from unit cube to parameter space takes place with the prior_transform() function. Contains separate prior transform functions for use with different sampling software.

API#

class phitter.fit.prior.uniform_prior(bound_lo, bound_up)#

Bases: object

Uniform distribution prior

bound_lofloat

Lower bound on the distribution

bound_upfloat

Upper bound on the distribution

Initialization

__call__(cube)#
__repr__()#
class phitter.fit.prior.gaussian_prior(mean, sigma)#

Bases: object

Gaussian / normal distribution prior

meanfloat

Mean of the distribution

sigmafloat

Sigma of the distribution

Initialization

__call__(cube)#
__repr__()#
class phitter.fit.prior.multivariate_gaussian_prior(means, sigmas, covar)#

Bases: object

Multivariate Gaussian / normal distribution prior

meansnp.array(dtype=float)

Means of the distribution for each parameter

sigmasnp.array(dtype=float)

Sigmas of the distribution for each parameter

covarnp.array(dtype=float)

Covariance matrix between the quantities, of shape [parameter x parameter]. Assume this has been calculated for the quantities after normalization. i.e.: calculated for each quantity after: (quant - mean(quant))/sigma(quant)

Initialization

__call__(cube)#
__repr__()#
class phitter.fit.prior.const_prior(value)#

Bases: object

Constant value prior

valuefloat

Constant value to return

Initialization

__call__(cube)#
__repr__()#
class phitter.fit.prior.prior_collection(priors_list)#

Bases: object

Collection of prior objects. Transformation from unit cube to parameter space takes place with the prior_transform() function. Contains separate prior transform functions for use with different sampling software.

priors_listlist[prior]

List of priors that consitute the full set of parameters being modeled.

Initialization

prior_transform_multinest(cube, ndim, nparam)#

Prior transform function for use with PyMultiNest

prior_transform_ultranest(cube)#

Prior transform function for use with Ultranest

prior_transform_dynesty(u)#

Prior transform function for use with Dynesty