Changes between Version 11 and Version 12 of WikiStart


Ignore:
Timestamp:
Jul 18, 2007 6:45:33 PM (15 years ago)
Author:
butler
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • WikiStart

    v11 v12  
    77With the advent of the NSF funded DANSE Project, which aims to build a complete suite of neutron scattering data analysis tools, and the opening of new neutron scattering user facilities at  the Oak Ridge National Lab (HIFR cold source), ANSTO in Australia (OPAL reactor) and KAERI in South Korea (HANARO cold source) we are taking the opportunity to evaluate the packages we provide. 
    88 
    9 This has lead to two initiatives. The first of these is collaboration with the above mentioned facilities to produce a shared code base of IGOR procedures for the reduction and analysis of small angle scattering data. This will provide to users a familiar environment whichever facility they use and make co-analysis of data from different facilities simpler.  
     9This has lead to two initiatives. The first of these is to explore the possibility of collaborating with the above mentioned facilities to produce a shared code base of IGOR procedures for the reduction and analysis of small angle scattering data.  This would provide to users a familiar environment whichever facility they use and make co-analysis of data from different facilities much simpler, an often heard refrain from the nomadic scattering community. 
    1010 
    1111The second initiative is to convert as much of the IGOR procedure code – which is written in an IGOR specific language and compiled within the IGOR program – to a generic library written in C. This separation of the graphical components from the analysis routines (scattering form factor and structure factor calculation, resolution smearing of models to match instrument geometry) provides both speed of calculation benefits and collaborative benefits. The former is important for co-fitting of multiple datasets (varying contrast, concentration, temperature etc) with complex models and initial testing suggests we can gain up to a 10-fold increase in computation speed. The latter is important because by abstracting out the actual computation code we can more easily make use of code from other sources – in particular we are working with the DANSE SANS subproject to define common interfaces to our libraries so that we can utilize each others work.