[Date Prev][Date Next][Date Index]
Interchange Programs for treated SAXS and SANS Data
- Subject: Interchange Programs for treated SAXS and SANS Data
- From: mooney@aps.anl.gov (Tim Mooney)
- Date: Wed, 3 Sep 1997 11:34:07 -0500
Dear folks,
Please redistribute this message to any small-angle scatterers you
know of who are actively involved in data analysis.
Tim Mooney
----- Begin Included Message -----
From sas@basil.ill.fr Wed Sep 3 04:44 CDT 1997
X-Sender: sas@mail.ill.fr
Mime-Version: 1.0
To: boesecke@MPGHDB.DESY.DE, stribeck@vxdesy.desy.de, r.denny@dl.ac.uk,
g.r.mant@dl.ac.uk, bhsiao@sunysb.edu, glatter@balu.kfunigraz.ac.at,
svergun@embl-hamburg.desy.de, lesieur@lure.u-psud.fr,
koch@embl-hamburg.de, gehrke@vxdesy.desy.de, capel@bnl.gov,
mooney@aps.anl.gov, r.heenan@rl.ac.uk, smk@rl.ac.uk, barnes@nist.gov,
schwahn@ell02.iff.kfa-juelich.de, jan.skov.pedersen@risoe.dk,
arr1@phy.cam.ac.uk, David.Maden@psi.ch, Kohlbrecher@psi.ch,
schneider@bio.bnl.gov, hjelm@lanl.gov, thiyaga@anlpns.pns.anl.gov,
lampert@hmi.de, buteau@bali.saclay.cea.fr, beckmann@gkss.de,
yagi@sp8sun.spring8.or.jp, kimura@dokkyomed.ac.jp, TischlerJZ@ornl.gov,
przemek@rrdjazz.nist.gov, Mark.Koennecke@psi.ch, bras@esrf.fr,
ferrero@esrf.fr, ghosh@ill.fr, may@ill.fr, timmins@ill.fr
Subject: Interchange Programs for treated SAXS and SANS Data
Content-Type: text/plain; charset='us-ascii'
Content-Length: 9544
X-Lines: 248
Interchange Programs for treated SAXS and SANS Data
>From Claudio Ferrero, Wim Bras (ESRF)
Peter Timmins, Roland May, Ron Ghosh (ILL)
Replies please to: sas@ill.fr
canSAS Grenoble 4th - 6th February 1997
Collective Aid for Nomadic Small-Angle Scatterers
Summary
We are proposing a hands-on workshop for a small group of active
programmers for SAXS and SANS data treatment to fabricate a set
of data interchange transcription programs for general use by our
shared scientific community. The workshop will take place from
February 4-6th at Grenoble, followed, we hope by the
possibility of a relaxing weekend of skiing.
Dear Colleague,
As an expert in SAS data treatment we would welcome your participation
at this workshop. If you are not able to come yourself, or you can
propose a more active programmer, we would be happy to accept your
designated replacement.
We would be grateful to receive replies by 30th September 1997.
Small-angle scattering, by its nature, tends to require
further data analysis long after the measurement, and can often require
merging of data from different instruments at different centres. Our
recent e-mail trawl elicited a general comment concerning the
frustrations and time-waste spent on reformatting data, part of this
being due to lack of written descriptions of file formats.
Looking at the distribution list for this you will see that we
in Grenoble are not alone in having to face this problem. We are
proposing to hold a limited workshop to try and construct modular
data exchange programs to simplify the problem of importing and
exporting treated data between various packages in use. We hope
to use the opportunity too to provide a forum for informal presentations
on new data treatment techniques if time allows.
Treated data, as far as a visiting scientist is concerned, is simple
to describe, and for the majority of projects requires little more
than a list of corrected intensities and errors as a function of the
scattering vector. Increasingly spreadsheets are being used to manipulate
data reflecting the modern trend away from programming; the principal
problem raised here is error treatment propagation.
We leave it to each centre to design their best initial data reduction
suites, which are closely allied to instruments and calibration
procedures.
We wish to complement the aims of the NeXus group who are facilitating
interchange of standard raw data (see http://www/neutron.anl.gov/NeXus).
One goal of this workshop can perhaps be to offer direct feedback to the
NeXus group to define a minimum subset of necessary measurement
descriptors required for SAXS and SANS experiments. We hope to attract
active proponents for the NeXus standard to join us in open debate.
By limiting numbers of participants we hope that much will take place
in the form of round-table discussions and practical sessions, rather
than formal presentations. Previous workshops have shown the highest
productivity occurred during 'hands-on' sessions.
Before the workshop takes place participants will be invited to
contribute documentation and source routines to an ftp server which
will be set up shortly. We hope too that this e-mail might stimulate
an interchange of ideas before the meeting. If this is wished then
we will setup a mailing list to conserve and forward messages.
Provisional timetable
Wednesday 4th February
12-14:00 Registration
14-15:30 Review of aims
tea
16-18:30 practical sessions
Thursday 5th February
9-10:30 short presentations (20 mins/person)+review of
previous day's activities
coffee
11-12:30 short presentations
designation of a committee to oversee WWW publication
of results
14-18:00 practical session
19:30 dinner
Friday 6th February
9-10:30 review of achievements
comments from/to committee
11-12:30 additional presentations
summing-up and close
For the practical sessions we will have a number of PCs and workstations,
which will be available throughout the meeting.
We will provide buffet lunches on Wednesday and Thursday. The
cost of the dinner will be approx. 300F per person.
A number of rooms have been reserved in the ESRF guesthouse on site
which costs about 150F per night, and we hope the rooms may be available
over the weekend. If there is enough interest we will try to arrange a
ski outing together at the weekend.
We do not have a budget at this stage to help pay for your travel and
accommodation. We are hoping the unrecorded thanks from the SAS
community will be at least a partial reward for your participation.
Initial invitees:
P. Boesecke boesecke@MPGHDB.DESY.DE
N. Stribeck stribeck@vxdesy.desy.de
R. Denny r.denny@dl.ac.uk
G. Mant g.r.mant@dl.ac.uk
B. Hsiao bhsiao@sunysb.edu
O. Glatter glatter@balu.kfunigraz.ac.at
D. Svergun svergun@embl-hamburg.desy.de
P. Lesieur lesieur@lure.u-psud.fr
M.H.J. Koch koch@embl-hamburg.de
R. Gehrke gehrke@vxdesy.desy.de
M. Capel capel@bnl.gov
T. Mooney mooney@aps.anl.gov
R. Heenan r.heenan@rl.ac.uk
S. King smk@rl.ac.uk
J. Barnes barnes@nist.gov
D. Schwahn schwahn@ell02.iff.kfa-juelich.de
J. Skov Pedersen jan.skov.pedersen@risoe.dk
A. Rennie arr1@phy.cam.ac.uk
D. Maden David.Maden@psi.ch
M. Kohlbrecher Michael.Kohlbrecher@psi.ch
D. Schneider schneider@bio.bnl.gov
R. Hjelm hjelm@lanl.gov
P. Thiyagarajan thiyaga@anlpns.pns.anl.gov
G. Lampert lampert@hmi.de
A. Buteau buteau@bali.saclay.cea.fr
J. Beckmann beckmann@gkss.de
Dr. Yagi yagi@sp8sun.spring8.or.jp
Dr. Kimura kimura@dokkyomed.ac.jp
NeXus specialists
J. Tischler TischlerJZ@ornl.gov
P. Klosowski przemek@rrdjazz.nist.gov
M. Koennecke Mark.Koennecke@PSI.ch
and members from the ESRF and ILL
----------------------------------------------------------------------
APPENDIX
We present below some ideas on improving data interchange.
This is a simplistic view intended to serve as a basis your
comments. If flawed it may be deliberate to incite your more
astute reactions...
----------------------------------------------------------------------
Traditionally the results have been treated after regrouping as
linear arrays of Intensity versus the wavevector transfer Q, or as
2-D corrected images. Kinetic or stroboscopic measurements add
additional dimensions, though as in the case of 2-D analyses the
number of distinct samples is far fewer as more time is spent on
setting-up, and in gaining reasonable statistical accuracy per
detector unit.
A global solution is envisageable, but perhaps at the expense of
simplicity; many users are happy to import regrouped data into spreadsheet
applications, though we think we have some duty to urge them to think about
propagating errors through their treatment.
As a consequence we propose that we first deal with the regrouped
data in this proposal, and open the discussion on 2-D and multi-
frame data to you, with your present and proposed facilities for
complex sequences and analysis schemes.
COMMON DATA FORMAT
Dealing first with the simplest case - 1-D (regrouped data)
The solution should allow the minimal interchange of:
Sample Name
Scan results a. Intensities (or cross-sections...)
b. statistical error in the Intensities
c. wave-vector transfer Q (A-1 or nm-1...)
d. error in Q due to finite resolution
One could imagine this: a title
: four columns of space separated data (a-d)
as the basis for interchange data
For a time of flight neutron spectrometer (d) would be calculated by
the program which groups data from different wavelengths, with
approprate weighting. On most other instruments calculation of resolution
effects is left to the modeller, with specific instrument parameters:
e.g.
1. Wavelength (A or nm)
2. Wavelength spread (%,A,nm)
3. Sample-detector distance
4. e.g. pinhole camera source/sample distance & dimensions
5. detector pixel size (mm...)
etc.
Although most users would accept an approximation for the resolution
being included with the data we know that we often use 2. for example
as a sensitive refinable variable in fitting programs to fit marked cusps
in SAS data. Specific instrument details can be partially hidden as
a text block which precedes the columnar data and can easily be
either skipped or edited out of treatments.
2-D Data and Sequences
These are much more difficult to present in a simple fashion appropriate
to a wide number of packages unless one of the CIF, TIFF, HDF, NeXus etc.
file formats is chosen. This should be discussed in more detail during
the workshop proceedings.
DATA ACCESS ROUTINES
This meeting offers us all the opportunity to exchange basic file access
routines, and to assemble a document/web page with file formatting
and access procedures.
Initially we were going to propose a stylised subroutine/procedure call
for which each attendee would write a transposition of data to local
standards, or, if missing certain necessary parameters would prompt
the user. One can also imagine a resource file describing mapping of data
between different formats etc. Enough time remains between now and the
meeting to advance on these routes; as a small demonstration we intend
a trial of data interchange between ILL and ESRF; we hope to complete
this exercise in the first half of September.
----- End Included Message -----