canSAS-3 Grenoble, 17-19th May, 2001

[Ron Ghosh]

canSAS-3 - Collective Action for Nomadic Small-angle Scatterers
Grenoble, 17-19 May, 2001, http://wwww.ill.fr/lss/canSAS/

1.  For some time I have tried to wave the flage for helping the
very large and ill-organised community of SAS users, and I was
greatly heartened at the PSI meeting when discussion moved on
to treated data.  Given corrected intensities, and estimates of 
resolution, these simple data from different centres can be merged for
analysis.  This would satisfy a large part of the SAS community,
especially if there is agreement on formats for data exchange.
For simple 1D patterns ascii files are most attractive.  There 
is an increasing tendency to perform more complex analysis of
2D detector data, plus temperature or time variation, and
other multi-parameter experiments.  This second class is well
suited to NeXus for use with graphical browsers, since it
can accommodate such a wide range of data, attributes and
commentaries, and for development as an interchange format.

2.  After the meeting at PSI I proposed that the SAS community
should draw every benefit possible from the NeXus activities, and 
especially the software expertise to advance to use HDF for
these complex treated data.  Because considerable work has 
advanced dictionaries for SAS work, primarily for archiving
data systematically, part can be re-used in establishing
matches with more standard NeXus terms and classes.  As long
as individual instruments cannot, in general, be mapped into
the same description and data reduction is centre dependent,
it would seem much more profitable to concentrate on HDF
storage of treated data as a common format, rather than
raw, and ignore the other (descriptive) classes, but extend 
use of simple commentary blocks.  

This would allow a period to gain acquaintance of
the potential of the full NeXus form, which could also
serve those for whom a complete archive of experiments
is considered an end result.  Most SAS users are often only
seeking a single complementary metric from their measurements, 
and are baffled by the apparent complexity of all the possible
NeXus classes.

3.  There will be a canSAS-3 meeting in Grenoble on 17-19th May, 2001,
to promote sharing of data and treatment programs.  I have
made a biased summary of the PSI-NeXus meeting which has
a link from the canSAS web page.  At canSAS-3 we hope
to advance matching NeXus and sasCIF structures, which
we consider to be complementary in nature.

With great trepidation I invite your comments for or against
our proposals. 

At the canSAS-1 meeting (Grenoble, 1998) NeXus aroused considerable
interest.  We would be happy to welcome members of the development
team to the meeting to demonstrate the present maturity.

Ron Ghosh

-- 
Ron Ghosh, Institut Laue Langevin
6 rue Jules Horowitz,                  telephone: +33-476207178
B.P. 156  cedex 9,                     secretary: +33-476207226
Grenoble, 38042 FRANCE                       fax: +33-476483906