Package gmisclib :: Module mcmc_big

Module mcmc_big

An extension of mcmc that includes new stepping algorithms.

Classes
	BootStepper

Functions

N_maximum(a)

source code

test()

source code

find_closest_p(v, vp)
Searches a list of (v,p) pairs and finds the one whose v is closest to the first argument.

source code

bootstepper(logp, x, v, c=None, strategy='intermediate', fixer=None, repeatable=True)
This is (essentially) another interface to the class constructor. source code

test2d(stepper)

source code

Variables
	SIGFAC = `3.0`
	__package__ = `'gmisclib'`

Imports: math, random, numpy, die, mcmc, Debug, gpk_lsq, NotGoodPosition, diag_variance, stepper, problem_definition_F, problem_definition, position_repeatable, position_nonrepeatable

Function Details

find_closest_p(v, vp)

source code

Searches a list of (v,p) pairs and finds the one whose v is closest to the first argument. Returns (v,p) of the closest pair.

bootstepper(logp, x, v, c=None, strategy=`'intermediate'`, fixer=None, repeatable=True)

source code

This is (essentially) another interface to the class constructor. It's really there for backwards compatibility.

Module mcmc_big

find_closest_p(v, vp)

bootstepper(logp, x, v, c=None, strategy='intermediate', fixer=None, repeatable=True)

bootstepper(logp, x, v, c=None, strategy=`'intermediate'`, fixer=None, repeatable=True)