group image photfile psfimage groupfile
image
The list of images containing the stars to be grouped.
photfile
The list of input photometry files, or the directory,
containing the photometry for
the stars to be grouped together. If photfile is "default",
"dir$default" or a directory
specification, GROUP looks for a file called image.mag.? where ?
is the highest existing version number. Otherwise photfile must
specify one photometry file for each input image in image.
Photfile is normally
the output of the PHOT task but may also be the output of the PSF,
PEAK, NSTAR and ALLSTAR tasks. Photfile may be a text file or
an STSDAS database file.
psfimage
The list of images, or directory, containing the PSF computed
by the DAOPHOT PSF task. If psfimage is "default",
"dir$default", or a directory specification, then GROUP looks for an
image called image.psf.? where ? is the highest existing version number.
Otherwise psfimage must specify one PSF image for each image in
image.
groupfile =
The list of output photometry files which will
contain the grouped photometry list.
If groupfile is "default", "dir$default", or a directory specification
then GROUP writes a file called image.grp.? where ? is the next available
version number. Otherwise groupfile must specify one output file
for every input file.
If the DAOPHOT package parameter text is "yes"
then a text database is written, otherwise an STSDAS table is written.
datapars =
The name of the text file containing the data dependent parameters.
If datapars is null ("") then the default parameter set in the
user's uparm directory is used.
daopars =
The name of the text file containing the daophot fitting parameters
If daopars is null ("") then the default parameter set in the
user's uparm directory is used.
verbose = )_.verbose
Print messages about the progress of the task?
Verbose may
be set to the value of the daophot package parameter (the default),
"yes", or "no".
verify = )_.verify
Verify the critical task parameters?
Verify may
be set to the value of the daophot package parameter (the default),
"yes", or "no".
update = )_.update
Update the critical task parameters if verify is "yes"?
Update may
be set to the value of the daophot package parameter (the default),
"yes", or "no".
GROUP takes the photometry file photfile file containing the stellar coordinates and photometry and associates the stars into natural groups based upon proximity and the magnitude level at which they overlap. The results are written into groupfile. If the DAOPHOT package parameter text is "yes" then groupfile is a text database, otherwise it is an STSDAS table.
The algorithm works in the following manner. If two stars are within a distance R pixels of one another, where R = psfrad / scale + fitrad / scale + 1, the PSF of the brighter one is evaluated at a distance d pixels, where d = fitrad / scale + 1 away from the fainter. If this value is larger than critoverlap times the expected noise per pixel then the two stars are put into the same group since the brighter star is capable of affecting the photometry of the fainter. Psfrad, fitrad and critoverlap are the psf radius, the fitting radius, and the critical overlap respectively and are located in the DAOPARS task. Scale is the image scale parameter and is located in the DATAPARS task. In order for this algorithm to work correctly it is imperative that the DATAPARS readnoise and gain parameters readnoise and gain be set correctly as these values are used to compute the expected random noise per pixel.
The correct value of critoverlap must be determined by trial and error. For example if a critical overlap of 0.1 divides all the stars in the image into groups smaller than the maxgroup parameter in the DAOPARS task, then unavoidable random errors will dominate over crowding errors. If a critical overlap of 1.0 works, then crowding errors will be no worse than the random errors. If a critical overlap much greater than 1 is required then in most cases crowding will be the dominant source or error.
If verbose is set to "yes", GROUP will write a table on the terminal showing the number of groups as a function of group size. If any group is larger than maxgroup then critoverlap must be increased or the GRPSELECT task used to cut large groups out of the file. When crowding conditions vary across the frame, GROUP and GRPSELECT can be used together to get the best possible photometry for stars in different crowding regimes.
If any stars in photfile have INDEF magnitudes, GROUP will attempt to estimate a magnitude for them based on the weighted sum of the pixels within fitrad of the center of the star. The weight is the product of a radial weighting function and the value of the PSF at that point.
1. Group the PHOT task output results for a field near n4147 using a critical overlap of 1 and printing the output summary on the terminal.
da> group n4147 n4147.mag.1 n4147.psf.1 "default" crit=1.0 verb+
2. Run the same job in batch but save the output summary in the file group.lis. Remember to turn off the verify switch.
da> group n4147 n4147.mag.1 n4147.psf.1 "default" crit=1.0 verb+
verify- > group.lis &