Running ORAC-DR pipeline

Joint Astronomy Centre

JAC Home
JCMT
UKIRT
Contact info
JAC Divisions
OMP
Outreach
Seminars
Staff-only Wiki
Weather
Web Cameras

____________________

JCMT home

Observing at JCMT
OMP Observation Manager

Telescope

Instrumentation

Schedule
Data Archive

Future Developments
Legacy Surveys
Newsletter & Publications

Running ORAC-DR pipeline JLS logo

Installing, updating and running the ORAC-DR pipeline

These instructions do not require you to have any special permissions.

Installing Starlink

ORAC-DR comes as part of the Starlink Software Collection which can be downloaded from the preceeding link.

If you would like to have the latest versions of both Starlink and ORAC-DR in one fell swoop, please email Tim Jenness for instructions on rsync'ing the latest builds from the JAC. This is only available for relatively new 64-bit Linux systems.

Installing and Updating ORAC-DR

To maintain updates of the ORAC-DR pipeline you will need to execute the following procedure.

First, decide in which directory you will be installing ORAC-DR and go to it. Then, get the most recent version of ORAC-DR by typing
```
  git clone git://starlink.jach.hawaii.edu/oracdr.git
```
This will put ORAC-DR in an oracdr directory in your current working directory (assumed to be /your/directory for the following).
Change to the oracdr/src/lib/perl5 directory.
Download the extra Perl modules to that directory and untar the file.

Now set the following environment variables:

setenv ORAC_DIR /your/directory/oracdr/src
			
setenv ORAC_PERL5LIB /your/directory/oracdr/src/lib/perl5
			
setenv ORAC_CAL_ROOT /your/directory/oracdr/cal

  setenv ORAC_DATA_CAL  /your/directory/oracdr/cal/acsis/

Brad Cavanagh

This should complete the installation of ORAC-DR to /your/directory on your local disk.

To update ORAC-DR, you should enter the /your/directory/oracdr directory and type

  git pull

Running ORAC-DR

When running ORAC-DR, you may see error messages stating "Problem reading engine version information". To eliminate this message download this file and save it to $STARLINK_DIR/manifests. If you do not have write permissions to this directory or cannot otherwise update this file, don't worry. The error messages are harmless and have no effect on data.

You should now be able to run ORAC-DR. To do so, type

  $ORAC_DIR/etc/oracdr_acsis < utdate >
		
  setenv ORAC_DATA_IN < path to directory where data is kept >
		
  setenv ORAC_DATA_OUT < path to directory where reduced data will go >
		
  oracdr -list 20:23,26:29 -batch -log xf -nodisplay -recsuffix=ADV

You need to provide the utdate when running up ORAC-DR.

The next two commands tell ORAC-DR where to find the data to process and where to put the processed data. If you want to work with the raw and processed data in the same directory then you may as well run the pipeline from that directory and provide `pwd` as the path in these two commands.

The last command sets the pipeline running with the following parameters:

  -list  		list of observation numbers
		
  -batch 		runs in batch mode (quicker)
		
  -log xf		writes log to x-term and to a file
		
  -nodisplay		suppresses the display of graphics (quicker)
		
  -recsuffix=ADV	uses the advanced versions of recipes if available

The -recsuffix command will use the advanced version of whatever recipe was stipulated in the MSB and written to the headers. This choice can be overridden in the command line,

  oracdr -list 20:23,26:29 -batch -log xf -nodisplay REDUCE_SCIENCE_NARROWLINE_ADV

The advanced recipes use an iterative technique to improve on the baseline subtraction and can successfully recover faint emission from the data. The more data that go into a cube the better, since the improved baselines are estimated from the coadded group file, and are then applied to the individual cubes. This results in improved single observation products (labelled as _reduced in the output directory). However, the current version of ORAC-DR does not then coadd the resultant improved single observation cubes into an improved coadded group product. This can be done manually using the KAPPA wcsmosaic command. For example,

  wcsmosaic in="a*_01_reduced*" out=final_group

<back to JLS home>

Contact: Antonio Chrysostomou. Updated: Fri Jul 10 20:37:50 HST 2009

Return to top ^