|Examples of pipeline processing|
In many cases, raw data are archived and a consisten description of the file (FITS header) allows to process them on the fly. This allows to make customized parameter extractions.
Lets consider as an example this file: L93111HP1/00147, a spectrum of NGC7785.
This first command makes a "pre-view"
display of the image.
The following command executes: The
flat-field correction, a removal of skipes (cosmics), the resampling in
wavelength, subtraction of the sky, display of the 2D corrected image,
of the central 10 arcsec and finally display of the final 1D spectrum.
The commands consists in a series of
pairs parameters=values separated
by the sign &. In these example the parameters are o
z respectively corresponding to (o) the address of the data
injected in the pipeline and (z) the suite of commands executed on the
The commands executed on the pipeline are separated with the character | and have in general arguments given within [ ]. These arguments can be omitted when they have a default value.
In this example the address of the data is fa:... (for FITS archive). The next example will show other possible address.
The pipeline is executed on a single machine, but a series of pipelines on differents machine can be attached together. Possibly, arguments of commands on the pipeline may also be pipelines.
The type of output is controlled by the parameter a. The next command will download the result of the pipeline into a FITS file (the most useful case).
One more example: Take the spectrum of
a star observed with ELODIE (R=42000), show it, calibrate in physical
flux, show it, broaden it to a velocity dispersion of 200 km/s and
red-shift it to 5000 km/s, show it.
It is easy to inject this file into the
processing pipeline. For example the following command takes this
spectra, display it as it is (ie. sampled in log(wavelength))
And the next command resample this spectrum in linear wavelength in the range 450 to 600 nm with a step 0f 0.1 nm, and show the result: