diff --git a/docs/source/history.rst b/docs/source/history.rst index 0c47de74..9a4c7434 100644 --- a/docs/source/history.rst +++ b/docs/source/history.rst @@ -3,7 +3,7 @@ History of quickBayes ===================== -The quickBayes packages started as a replacement for the quasielasticbayes package, which was a Fortran code base that was exposed to Python. +The quickBayes package was originally designed as a successor to the `quasielasticbayes package `_, which was a Fortran code base that was exposed to Python. In the development of quickBayes the code became more modular and flexible. This allows quickBayes to be applied to any problem that requires model selection. diff --git a/docs/source/intro.rst b/docs/source/intro.rst index 6dc3398d..8fe31517 100644 --- a/docs/source/intro.rst +++ b/docs/source/intro.rst @@ -5,9 +5,6 @@ The quickBayes package is an open source library for calculating a fast approxim The package is cross platform, supporting Windows, Mac OS and Linux. This package has been developed by Anthony Lim from STFC’s ISIS Neutron and Muon facility. -The quickBayes package was originally designed as a successor to the `quasielasticbayes package `_. -However, quickBayes has abstracted the key ideas to make a more generic package. - In science hypotheses are tested against data, to determine the underlying behavior of the system. These hypotheses can be in the form of a mathematical expression that originates from first principles (i.e. it has been derived) or is an approximation to other more complex mechanisms (e.g. semi-empirical methods). The quickBayes package is designed to test these hypotheses against the user’s data to identify the most likely. @@ -20,6 +17,11 @@ For example, in Quasi Elastic Neutron Scattering (QENS) the data can be represen :alt: Fits for 1, 2 and 3 Lorentzians to QENs data. The quickBayes packages is designed to make these decisions easier by calculating the probability of one, two or three peaks given the data. +By using quickBayes it is possible to calculate a fast, analytical approximation of the marginal likelihood for a series of models, which describe a specific data set. +For a more detailed discussion see :ref:`key principles and theory behind quickBayes `. +It is important to note that quickBayes does not use traditional Bayesian methods, such as Markov Chain Monte Carlo or nested sampling. +A brief discussion of these alternatives is provided in the :ref:`Bayesian methods ` section. +This includes a direct comparison between the Bayesian method and quickBayes. This documentation is split into the following parts: