IRDR (Infrared Data Reduction)
Version: 1.0 (May 2003)

Introduction

  IRDR is a C library and set of stand-alone C programs and perl scripts
  for processing IR imaging data.  See the file irdr/Programs, and
  irdr/doc/irdr.ps.gz.

Installation

  download ftp://ftp.ast.cam.ac.uk/pub/carine/irdr/irdr-1.0.tar.gz
  gunzip irdr-1.0.tar.gz
  tar xvf irdr-1.0.tar
  cd irdr
  make all                                        # compiles WCSTools then IRDR
  setenv IRDR_BASEDIR /home/sabbey/irdr/                   # use your location
  set path = ($path $IRDR_BASEDIR/scripts $IRDR_BASEDIR/bin)
  you should have "sex" command (SExtractor 2.1+) in your path
  (ftp://ftp.iap.fr/pub/from_users/bertin/sextractor/)


CIRSI data reduction (example)

  mkdir AXAF; cd AXAF
  cp ~/irdr/scripts/cirsi/cirsi.csh .
  vi cirsi.csh
  ./cirsi.csh >& cirsi.log &

  Note: change mean.o to meangain.o in Makefile OBJ_MATH definition if you
        want clipping during coaddition of fewer than 5 image planes.  This
        option requires knowing the detector gain.

History

  Version 1.0 (29 May 2003) - Carine Babusiaux
  Debug astrometry_refine.pl, update wcstool version
  Final photometry and astrometry checks done. 

  Version 0.95 (28 May 2002) - Carine Babusiaux
  Added linearity correction (nlinmap.c lindarkflat.c)
  Added dark computation (dark.pl)
  Add option "standalone" in combineloops.pl
  Add option scaling="none" in cubemean.c
  Assume new controler in place, loop1 not skipped 

  Version 0.94 (November 2001) - Carine Babusiaux
  Add 2mass catalogue option in astrometry.pl (using 2mass.pl) 

  Version 0.93 (20 Mar 2001) - Chris Sabbey
  Added program darkflat.c that does dark subtraction then flat correction

  Version 0.92 (02 Feb 2001) - Chris Sabbey
  Modified fitsIO.c:
  Write unsigned short FITS images with BZERO 32768.0, BSCALE 1.0 convention
  Assume bitpix=16 input FITS images are signed short int (not unsigned)

  Version 0.91 (26 Jan 2001) - Chris Sabbey
  Require setenv IRDR_BASEDIR /path/irdr
  Arguments to makemask.pl changed, don't need to specify config directory

  Version 0.9 (18 Jan 2001) - Chris Sabbey
  Initial release


Current Limitations

  Input images can be float (bitpix -32) or signed short integer (bitpix 16)
  Ouput images are float or unsigned short integer:
    unsigned short written as bitpix 16 with BZERO 32768.0, BSCALE = 1.0
    BSCALE and BZERO are applied on input but hence dropped in output
  Some FITS images don't end in .fits, to display outside of IRAF, eg:
    linux> ds9 image.fits.skysub
  Fast histogram analysis was designed for u_short data:
    forces float and signed short data to unsigned short range (0 to 65535)
    assumes a noise level > 1 count 
  Need enough memory to hold all image planes (and weight maps) for coaddition
  Assumes input data images have same dimensions
  Expects CDELT or CDELT1 keywords, or SCALE or SECPIX (image arcsec/pixel)
  Dither offset measurement guesses that North is up and East is left