Next: Functionalities
Up: Data Pipeline
Previous: Goals
  Contents
- The pipeline must be able to process all the data coming from the
array; it must not constitute in any case a bottle-neck in the data
flow.
- The pipeline must operate in fully automated mode for processing from
raw data up to deconvolved images. Although this fully automated
processing may not lead to optimum quality reduced data, it
will constitute the basis for an homogeneous database of processed
data.
- The pipeline must also be able to operate in manual mode for
technical development, inspections by expert engineers and
astronomers on duty but also at the level of the PI/CoI astronomers,
for projects which request interactive
observations. Requirements (input pipeline parameters) may be given
at the stage of the elaboration of the proposals by the astronomers.
- The pipeline must process in quasi real time the observing blocks but,
in the imaging stage, it must also be able to include all the
observing blocks previously obtained since the project has started. This
may mean calibrated data obtained in different
antenna configurations.
- The pipeline must interact with various packages, including the
Dynamic Scheduler, the actual observing program, various high levels
tools for visualization, and the data transmission
(compression/decompression, formatting).
- The pipeline must operate through a readable and comprehensive data
reduction script. A high level tool may be available to
interactively build such a script, but for automated processing it
must be automatically generated from a template, on the basis of the
scheduler status and of the collected actual observing
procedures. For debugging it should also be able to operate at the
command line level. Scripts will be composed of basic components
such as, pointing, focus, phase calibration, etc... Some of these
components will be activated sequentially but other in parallel (in
particular at the mapping stage). It must be able to process in
parallel sub-arrays.
- The pipeline should be run either at or near the telescope, in quasi
real time, or later, at places where the official archives are kept
(see Section 7).
Next: Functionalities
Up: Data Pipeline
Previous: Goals
  Contents
Kate Weatherall
2000-03-08