Astronomical image processing packages: IRAF basics

The Image Reduction and Analysis Facility (IRAF) is a suite of software developed by NOAO in the 1980's. It provides an environment for the reduction and analysis of astronomical data that is widely used, especially in the US astronomical community. However, there are certainly a number of astronomers who find the IRAF approach somewhat cumbersome or opaque, and who prefer to develop their own tools for data reduction. Nonetheless, some familiarity at least with IRAF tools is probably a very good idea.

IRAF has been incorporated into a more modern interface with the development of PYRAF, which is a Python front-end to the IRAF routines. In this day and age, use of IRAF through this interface, as opposed to the traditional CL interface, is strongly recommended!

IRAF/DS9 basic operation

• One time only for each directory you will run in: run mkiraf, which creates login.cl, and starts a uparm/ subdirectory for parameter files. The login.cl can be customized at a later time if you have settings you want to start with every time.
  • to enable a larger frame buffer for display, uncomment and modify line: stdimage = imt2048

• The preferred method of running IRAF in the modern era is using the PYTHON interface, pyraf
  • You can use pyraf via a normal python interface: after starting Python, just

                from pyraf import iraf
                or
                from pyraf.iraf import *
    

    (for former, you'll need to precede tasknames with iraf.) You will then need to use standard PYTHON syntax, rather than the old IRAF cl syntax
  • You can use pyraf through a front-end interpreter to emulate the original IRAF command-line interface, using the command: pyraf. This is convenient for previous users and for some tasks, but ``hides" the Python interpreter and its power
  • Note that PYRAF still requires that IRAF is installed on the system: it is just a front end, rather than a rewrite

• If you will be displaying images, start an image display tool, i.e. DS9, in the background: ds9 &. Be aware that the stdimage that is set in the login.cl file may limit the maximum size of the image that will be displayed!

• help:
  • there is an internal iraf.help command (.help in CL). (Don't get confused with Python help!)
  • IRAF help,
  • Users guides
  • tutorials.
• basics:
  • IRAF contains many programs for astronomical analysis. These are grouped into packages, and individual commands are tasks within each package. Before a particular task can be run, the package within which it is located must be loaded, which is done by entering the package name. Several packages are loaded by default, and this can be customized in the login.cl file. If you know a task name, and need to find out what package it can be found in, try the iraf.apropos('task') command (apropos task in IRAF interpreter)
    • e.g., to load noao package via Python interpreter: iraf.noao(); to load noao package via PYRAF interpreter: noao
    • common packages: iraf.imred(), iraf.onedspec(), iraf.twodspec()
    • to list tasks in package: iraf.help('package')

  • Most tasks have a set of adjustable values, or parameters, which govern the specifics of how the task operates. IRAF manages the values of these parameters by having an individual parameter file for each task. You can look at and modify the parameter values for a given task using the iraf.epar('taskname')(in CL: epar taskname) command, or iraf.lpar('taskname') to just print a list. This is useful for seeing what all of possible parameters are. If you modify parameters using epar, the modifications are saved and will be used in all future invocations of the task!
    • parameters are saved in files in a uparm/ directory
    • because of this behavior, it is useful, and perhaps even wise, to know how to reset parameters back to their default values: this can be accomplished using the iraf.unlearn('taskname') (unlearn taskname) command

  • You can get detailed information about each task and its parameters using the iraf.help('taskname') (help tasknamera) command. This is also available through the graphical epar.

  • You can override a value from the parameter file by specifying it as a keyword in the commande, e.g. valname=xxxx. In this case, the parameter file is not permanently modified. This is convenient for scripting. On the command line, keywords are separated from the command and each other by spaces.

  • In the Python interface, parameters can also be accessed/modified as attributes of the task

  • IRAF is a disk-based system: commands that work with images require filenames of input images and filenames for output images.

  • You can issue operating system commands from within IRAF command language by starting the command lines with an !

• image display with DS9 through IRAF: the display task
  • The default mode is to autoscale the image (zscale=yes, zrange=yes). You can manually set the display range using z1=low and z2=high, but you must also turn off autoscaling (zs='no' zr='no') for the manual values to take effect

  • Note that the data value display in the display window is derived from the display pixel value, so you won't see actual data values that are below z1 or above z2, the value display will just say < z1 or > z2 – rather annoying.

  • If using IRAF to control the ds9 display, the ds9 scaling option will not be available

  • if image buffer isn't set correctly, you may not see the full frame! you can reset using, e.g., :

               iraf.set(stdimage='imt2048')
    

• image cross sections: the implot task.
  • See plot window commands ('?')
• Image histogram: imhist. Look at the parameter file. Note that you can specify image subsections using filename[x1:x2,y1:y2]
• Image statistics: imstat. Look at the parameter file. Note you can specify image subsections as above.
• Image arithmetic: imarith. You can do arithmetic with images and constants, or with multiple images. For example: imarith file1.fits - 363 will subtract a constant of 363 from the image, imarith file1.fits / file2.fits will divide file1 by file2 (on a pixel-by-pixel basis).
• Inspection of stellar images: imexam. Note 'a', 'r', and 'm' keys, '?' for help (note you have to exit help to get interactive cursor!), 'q' for quit

• For many tasks that require an input file, it is possible to specify a list of input files if the same action is to be taken on each. This is accomplished by creating a file (e.g., files.lis) that has a list of all of the files that you wish to run the task on, each on a separate line. Then, instead of giving an image name to the task, you give the list file name than contains the image names, preceded by an @ sign, e.g. @files.lis. Remember when you are proccessing images that IRAF wants to write the output files; if you don't have write permission in the input directory, you'll need to also supply an @output.lis file with the names for the output images.

• Quit pyraf using the .exit command

IRAF data reduction

IRAF package imred/ccdred: zerocombine, darkcombine, flatcombine, ccdproc (note need for header cards)

lower level: imcombine, image arithmetic

Note that you may have to do a iraf.setinst first, have to pay attention to ccdproc parameters!

IRAF file list specification: comma-separated string

IRAF simple stellar photometry

phot/apphot

Exercises

1. Display some images, get comfortable with setting display parameters

2. Do some of the usual: image histograms, cross section plots, regions statistics, PSF measurements. Use all of the routines mentioned above.

3. Reduce a nights worth of data. Script it.

4. Quantitatively assess the quality of the flat fields by looking at the flatness of the sky background in the reduced frames.