pkgsrc-wip/py-corrfitter/DESCR

21 lines
1.0 KiB
Plaintext

This module contains tools that facilitate least-squares fits, as functions
of time ``t``, of simulation (or other statistical) data for 2-point and
3-point correlators of the form::
Gab(t) = <b(t) a(0)>
Gavb(t,T) = <b(T) V(t) a(0)>
Each correlator is modeled using |Corr2| for 2-point correlators, or
|Corr3| for 3-point correlators in terms of amplitudes for each source
``a``, sink ``b``, and vertex ``V``, and the energies associated with each
intermediate state. The amplitudes and energies are adjusted in the
least-squares fit to reproduce the data; they are specified in a shared prior
(typically a dictionary).
An object of type |CorrFitter| describes a collection of correlators and is
used to fit multiple models to data simultaneously. Any number of
correlators may be described and fit by a single |CorrFitter| object.
|CorrFitter| objects can also be used to to extract the appropriate fit
data from |Dataset| objects.
This module has been used extensively for analyzing results from lattice
QCD simulations