APPLICATION SOFTWARE
ATCA DATA ANALYSIS SOFTWARE
Our main packages for reduction of ATCA data are Miriad, AIPS and AIPS++ . We also have a suite of vizualization software for analysis of three dimensional data cubes (e..g spectral-line cubes). Since there exists extensive user documentation for these packages via The Miriad Users Guide, Analysis of ATCA data with AIPS and the user's guide for the ATNF Vizualization Software, we will only give the barest outline here as to how to invoke these packages.
Data can be exchanged via FITS files or via Miriad datasets (between Miriad and the visualisation software).
See accessing RPFITS data on CACCC for details of how to get at your data if it's still on the CACCC archive disk.
Miriad
The standard Narrabri account is setup to give you access to this software. If you have your own .cshrc and .login files, you can find instructions on how to setup Miriad in The Miriad Users Guide. With Miriad, you control the files, their names and locations by interacting with the operating system as usual (unlike AIPS which presents to you its own internal file cataloguing system). The data directory where you should put your data is usually /DATA/KAPUTAR_3. For example, let us say that your user name is len070. First, to assess what the available disk space is, enter the Unix command
% df
Look for the file system called /DATA/KAPUTAR_3. The listings are in Kbytes. Visitors are required to use the main disk (KAPUTAR_3), staff have their own disk (KAPUTAR_2) and so does the local operations staff (KAPUTAR_1). You create a subdirectory for yourself with commands like
% cd /DATA/KAPUTAR_3
% mkdir len070
Then descend into this directory
% cd len070
This then is the main directory from which you will analyze your data with Miriad. You can of course make sub-directories below this one to organize your files however you like. Note that a Miriad file name is actually a directory with a collection of items (files) in it. But you will only access those items via the directory name.
To invoke Miriad, just enter the command
% miriad
and you will be greeted by the Miriad prompt, viz.
miriad%
Aips
If you are going to be working with AIPS, then the file systems are automatically accessed, and you only see the catalogues that AIPS presents to you (you do not have access to the files at the operating system level). Each AIPS user should have their own AIPS user number. These numbers are allocated by the Epping AIPS support staff (Henrietta May and Neil Killeen) and you should contact them to receive one if you have not previously been allocated one at the ATNF. These numbers are used to keep different user's files separate. Start up AIPS by entering the command
% aips
on a terminal logged in to kaputar (you do not need any special commands in your .login file for this). You will then see several messages which offer you useful information about the local AIPS system. You can read them all, or defer reading them as prompted. Once AIPS asks you for your AIPS user number, enter the number you have been allocated. You will then be greeted by the AIPS prompt, viz.
>
You will also see three iconized windows appear on your workstation screen. These are the AIPS TV, the line graphics window, and the message server window. If you would prefer AIPS send its messages to you via the window from which you started it up, just quit the message server window.
(The following info is no longer useful at the ATCA as all computing is done on a single machine now. It describes the Epping setup however)
By default, AIPS will offer you the data reduction file systems local to the computer that you are working on. If for any reason you need to start up AIPS and mount some remote file system from another computer (e.g., because not all machines have an Exabyte drive, so may have loaded the data to the local disk of another computer and now want to copy it across to your machine), you can do this with a command such as
% aips da=host1:host2:hostN
where host1:host2:hostN might be tucana:puppis to see the the file systems from those computers as well. You must remember that writing to remote file systems comes with a large network penalty (for more information see the section on File Systems in the chapter on Unix Computers and also the manual Analysis of ATCA data with AIPS ). For large data files you should always arrange things so that you read from a remote file system, not write to it. With AIPS in particular, which likes to write scratch files wherever it can, you can get into trouble with mounting remote file systems. You can use the AIPS adverb BADDISK to ensure that any scratch files go only to local file systems, not to the remote ones.
For all additional details of how to use AIPS to analyze your ATCA data, see the manual Analysis of ATCA data with AIPS
Vizualization
One of the more interesting aspects of analyzing 3-D data sets is trying to find the best way to absorb and understand the multi-dimensional information. The ATNF has a group that works on finding the best ways to render 3-D data cubes so that you the astronomer can find out what your data are really telling you. This involves converting the 3-D data cube into a 2-D display in such a way as to optimize this process. There is a suite of vizualization software available for this purpose and data is transmitted to the package either via FITS files (from AIPS) or directly by Miriad files.
The standard Narrabri account is setup to give you access to this software. If you have your own .cshrc and .login files, you can find instructions on how to setup the vizualization software in the ATNF Vizualization Software users manual.
Since this software makes heavy use of X-clients, and gives you full control over where any X-windows are displayed, you might find the section on Using X-windows on Remote Computers in the chapter on Unix Computers useful. As this software consists of a set of connected applications, with no front end shell to run, you should see the ATNF Vizualization Software manual for more details on how to use it.
Original: nkilleen@atnf.csiro.au (07-APR-1995)
Modified: Mark.Wieringa@csiro.au (12-DEC-1996)