Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.adass.org/adass/proceedings/adass98/ferroaj/
Дата изменения: Sat Jul 17 00:46:38 1999
Дата индексирования: Tue Oct 2 06:57:48 2012
Кодировка:
Поисковые слова: п п п п п п п п п п п п п п п п п п п п п
|
Next: User Defined Functions for the ISO Post Mission Archive
Up: Archiving and Information Services
Previous: Numerical Observatory: a Web Site for the Retrieval, Analysis, and Visualization of Simulated Cluster Data
Table of Contents -
Subject Index -
Author Index -
Search -
PS reprint -
PDF reprint
Barg, I., Ferro, A. J., & Stobie, E. B. 1999, in ASP Conf. Ser., Vol. 172, Astronomical Data Analysis Software and Systems VIII, eds. D. M. Mehringer, R. L. Plante, & D. A. Roberts (San Francisco: ASP), 245
Spiders and Camels and Sybase! Oh, My!
Irene Barg, Anthony J. Ferro, Elizabeth Stobie
University of Arizona, Steward Observatory, NICMOS, Tucson, AZ
85721
Abstract:
The Hubble Space Telescope NICMOS Guaranteed Time Observers (GTOs)
requested a means of sharing point spread function (PSF) observations. Because
of the specifics of the instrument, these PSFs are very useful in the analysis
of observations and can vary with the conditions on the telescope. The GTOs are
geographically diverse, so a centralized processing solution would not work.
The individual PSF observations were reduced by different people, at different
institutions, using different reduction software. These varied observations had
to be combined into a single database and linked to other information as well.
The NICMOS software group at the University of Arizona developed a
solution based on a World Wide Web (WWW) interface, using Perl/CGI forms to
query the submitter about the PSF data to be entered. After some semiautomated
sanity checks, using the FTOOLS package, the metadata are then entered into a
Sybase relational database system. A user of the system can then query the
database, again through a WWW interface, to locate and retrieve PSFs which may
match their observations, as well as determine other information regarding the
telescope conditions at the time of the observations (e.g., the breathing
parameter).
This presentation discusses some of the driving forces in the design,
problems encountered, and the choices made. The tools used, including Sybase,
Perl, FTOOLS, and WWW elements are also discussed.
During a meeting involving many of the NICMOS GTOs,
considerable interest was
expressed in establishing a database containing the PSF star observations
obtained as part of GTO programs. A centralized database was seen as a good
way of making the existing PSF observations available to everyone on the team.
This paper discusses how this PSF database and the user interface evolved.
Submissions to the database of processed PSF images are the responsibility of
the individual team members. These team members are geographically dispersed,
may have used different data reduction software, and may have named their
images differently.
A network interface using CGI forms addressed all these issues. Submissions
could be made remotely. Users would only need a Web browser to access the
submission form. To enforce commonality among submissions, the submission
form was structured in a way to ask specific questions regarding reduction
procedures and details about the PSF star. The submission
form needed to be straight forward and easy to use, it had to impose standards
(file naming conventions, minimum descriptive information), and it had to
be as short as possible
without omitting key information about the PSF star. Our overall design
goal was to automate the maintenance of the database as much as possible.
Figure 1 illustrates the submission and retrieval process in which a
PSF star observation travels from the submitter, is checked and ingested at
University of Arizona (UofA), and is later retrieved and sent from the UofA to
the requester.
Figure 1:
PSF Automated Archive Process
![\begin{figure}
\epsscale{.85}
\plotone{ferroaj1.eps}
\end{figure}](img1.gif) |
The PSF Star Submittal form is fairly detailed and requires some effort on
the part of the submitter. They are asked to enter information about the PSF
star and provide details of their data reductions. Some of the data entered
are used as search parameters on the PSF Data Retrieval form, so it is
important that the submission form be completed as completely as possible. A
small text window in which to enter comments is included. Comments
can be used to add
any information that is important, but was not included in the form. Information
provided on this form is transferred to the FITS headers of the submitted data
file as HISTORY keywords. File naming standards were adopted and other standards
were enforced through the form itself by requiring entries on selected fields.
Upon completion of the submission form, the data are automatically
FTPed from the
submitter's site to the database server at the UofA.
An automated message is sent via e-mail to
the user indicating the status of the transfer (received, failed, etc.).
A copy
of this notification goes to the NICMOS Database Administrator who then runs a
set of Perl scripts that perform the following functions.
- Verification
- Verify the PSF submission (authorize user, files complete, etc.).
- Verify that submitted PSF files conform to the FITS standard.
- Check for minimum keywords.
- Build Metadata
- Extract selected keywords from the submission form and FITS headers to
build the metadata that will populate the PSF database.
- Compress and move the PSF data file to the local UofA archive.
This Web interface allows users to search the PSF archive for data that match
criteria such as camera, filter, spectral type of the PSF star, breathing
ephemeris (relative focus), and x- and y-position on the array. A successful
search will present the user with a table of PSF star images from which they
may select the PSF's they wish to retrieve. Figure 2 shows the
results of such a search.
Figure 2:
Results from a PSF Retrieval search.
![\begin{figure}
\epsscale{.90}
\plotone{ferroaj2.eps}
\end{figure}](img2.gif) |
With the exception of the breathing parameter, all of the data
displayed from the results of the search represent data from the PSF database.
The breathing parameter is a measurement of the change in the
focus due to the temperature changes (breathing) of the telescope.
The amount of this change varies and is highly time dependent.
However, this value is not known at the time of the
observation and therefore is not found in the FITS headers. There are
two steps
for determining relative focus (breathing) at a given time:
- Look up the nominal value for the observation date, either in MJD, day
of year, or days since 1 Jan 1994.
- Look up the modeled variation from this nominal value for the exact
time of the observation and add this value to the nominal.
The resulting value is in secondary mirror microns and is relative to best
focus at WFPC2 PC (zero microns).
Calculating this value presents a problem because the nominal value
and modeled variation are embedded in two separate Space Telescope Science
Institute (STScI) HTML documents that are updated occasionally as the model
is improved. A means to extract the necessary information from the most recent
versions of these documents became a programming challenge that was met by
using a Perl module called libwww-perl (LWP). LWP allows us to
locate the
required two documents by extracting the links from the parent HTML page, fetch
the required documents, and parse the necessary information on the fly. A
description of how LWP accomplishes this would make a paper of its own.
This entire project was written using Perl V5.004 utilizing the
following modules: CGI.pm, Sybperl, libnet, and
libwww-perl (LWP). The
verification scripts, also written in Perl, utilized a set of FITS utility
programs from the FTOOLS package created by the High Energy Astrophysics
Science Archive Research Center (HEASARC). The SQL server is
Sybase1 SQL
Server V11.02.
The goal to automate the entire PSF submittal process drove our decision to use
Perl as our programming language. As shown in Fig. 1, the submittal
process involved communicating with the WWW, transferring data across a network
using FTP, verifying the data using FTOOLS, and then adding the new
PSF entries to
our our SQL Server. The Perl programming language gave us the tools to provide
the glue between all of these separate systems to make automation possible. To
view some of the PSF forms, go to the (NICMOS)
Web page and follow the links to the Database Page.
Footnotes
- ...
Sybase1
- SYBASE is a registered trademark of Sybase, Inc.
© Copyright 1999 Astronomical Society of the Pacific, 390 Ashton Avenue, San Francisco, California 94112, USA
Next: User Defined Functions for the ISO Post Mission Archive
Up: Archiving and Information Services
Previous: Numerical Observatory: a Web Site for the Retrieval, Analysis, and Visualization of Simulated Cluster Data
Table of Contents -
Subject Index -
Author Index -
Search -
PS reprint -
PDF reprint
adass@ncsa.uiuc.edu