More advanced ways to access the Elodie archive
The user interface of the Elodie archive has been designed to be used
either from a web browser or by a program. This latter feature, a fundamental
principle of the Virtual Observatory, allows web-based
services to transparently access the archive.
Accessing the catalogue of observations
The request http://atlas.obs-hp.fr/elodie/E.cgi?&o=vega
returns an HTML page with all spectra of VEGA available in the archive.
As the HTML format may not be convenient to use in a text editor or
in a program, you can modify the request by asking for ASCII format:
Although this plain-text format is the only machine-oriented format now available
for the catalogue, we may provide a tab-separated format (tsv) or
VO_table format in the future.
To download the content of an URL on a user's computer,
it is very natural to use a
browser, such as Mozilla, Netscape or even Internet Explorer, but this is not
the only solution. In practice, a browser is only one of the clients that
may be used to access the Web. Other common clients are the so-called
"crawlers" which can recursively download files from the web. A common
one is wget, present on most recent Linux distributions (but several
alternatives exist). In the following we will give examples with wget.
From a shell window, you can download the list of observations of Vega
(see above) as:
Accessing CCF results directly (NEW)
Complete CCF results for a given star can be obtained by entering a command with parameters in your browser. For example the command below will list in HTML format all the CCF results for HD190007 contained in the cross-correlation internal table (identified by n=501) :
By replacing a=htab by a=t you obtain a simple ASCII table with columns separated by tab marks. In order to limit the number of metadata appearing in the table you can also type :
To submit a query concerning a specific group of stars, for example all stars in the Gliese catalog, or all Henry Draper stars with names beginning with HD190, use the parameter o=GJ% or o=hd190% :
Please note that some ELODIE identifiers (objname) are still incorrect. We correct these as they appear.
Downloading FITS files directly
Suppose you are interested in all the spectra of Vega and want to
download the FITS files to your disk.
You can naturally go to the web interface with your favorite browser and
download each spectrum one by one.
But you can also issue the commands from the shell. For example:
"http://atlas.obs-hp.fr/elodie/E.cgi?&c=i&o=elodie:19960502/0041&a=mime:application/x-fits" -O 19960502_0041.fits
In the URL above you recognize the unique identifier of a file in the
archive. You may construct similar URLs for any file in the archive.
If you are a familiar with a text editor (such as emacs, pico or nedit)
and with Unix shell commands, you can readily write a procedure to
download all the files you desire.
So, now you know how to easily download archive spectra in the s2d
format. But you may wish instead the more easy-to-use
reconnected, resampled spectra, or continuum normalized
spectra. This is easy to do: just add the pipeline command to the URL.
For example to get the resampled spectrum use:
"http://atlas.obs-hp.fr/elodie/E.cgi?&c=i&o=elodie:19960502/0041&z=wrs&a=mime:application/x-fits" -O 19960502_0041.fits
The pipeline command is the value of the z= parameter in the URL. You
may analyze how the pipeline works by watching the customize menu, and
then produce your own pipeline command. You
can also read the
Useful shell scripts
Two scripts are available to handle ELODIE data (require bash shell):
- Script to retrieve a list of observations for a list of stars
from the ELODIE archive
- Script to retrieve a list of spectra for a list of observations
from the ELODIE archive
Accessing FITS files directly from astronomical packages
We have seen that the archive can be accessed with wget and we said that
other clients may be used. Actually many astronomical packages can
directly read a FITS file over the Internet by giving an address http://...
or ftp://... in place of the name of the file in the local file system.
For example, by using the Pleinpot package, you may
copy a remote file to a local disk file by typing:
or plot a spectrum by typing:
Upcoming feature: how to retrieve
spectra with specific characteristics