A list with RA and Dec in the first two columns followed by user data
is grouped into sublist based on spatial proximity. A separation parameter
defines the grouping by looking for gaps in both RA and Dec that are
bigger than the specified amount. The output sublists may or may not
include the RA and Dec columns. A typical example of user data might be
Input tabular text file containing RA and Dec in the first two whitespace
separated columns and user data in the remaining columns. The RA may
be in hours or degrees while the Dec must be in degrees. The RA values
must lie in the range 0h to 24h or 0d to 360d and the Dec values
must lie in the range -90d to 90d.
Output root filename. The root filename itself will contain a list of
the sublists. The sublists will have _NNN appended to the root name
where NNN is a three digit number. If there are more than 999 sublists
the number of digits will increase. A check is made for any pre-existing
filenames with this root, sequence pattern, and optional extension and
an error will result if any are found.
Keep the input coordinate columns in the output lists? If no then only
the user data will be placed in the output lists. This option allows
taking a list of RA, Dec, and filenames and producing only lists of
filenames to be used as @files.
The format for printing the RA and Dec in the output lists if
keepcoords is yes. See the help for printf for the formats.
Note that the raformat may be given in %H format to convert input RA
in degrees into output hours. The default produces sexagesimal format
keeping the RA in the same units as the input.
This task groups a list of user data with RA and Dec coordinates
into sublists where all points in a group have at least one member with
celestial distance in RA or Dec separately less than or equal to the
specified separation. In other words, groups are defined by gaps in RA
The input format is a text table where each line consists of an RA,
a Dec, and arbitrary user data. Whitespace separates these three parts.
The RA and Dec have certain restrictions on units and ranges as described
in the parameters. However, the RA may be given either in hours or degrees
and may be output in hours if given in degrees.
The output is a set of sublists as well as a file containing the set
of sublist filenames. The sublists contain the input user data with
or without the input coordinates.
The grouping algorithm is summarized as follows. The input list is
sorted by declination. The declination ordered list is traversed
to form groups with consecutive declination intervals less than or
equal to the specified separation. These groups are then
sorted in RA and these are traversed to form the final groups with
consecutive RA intervals less than or equal to the specified separation.
Note that the RA intervals are actually computed by skysep and
make use of both the RA and Dec.
A challenge is dealing with the wrap around in RA at the zero meridian.
This is handled by duplicating points near 0 beyond 24h or 360d. This is
the reason the input is required to only be in a specific range. This
duplication can result in entries appearing in more than one output group.
A merging step handles this situation.
1. A set of images is to be grouped based on their FITS tangent point
coordinates. Note this make most sense when the tangent point pixel
coordinates are the same in the image. The image will then be grouped
to find those that overlap by some amount. If the images have 10 arc
minute fields of view and we want to group those that overlap by at least
50% then the separation parameter should be something like 5 arc minutes.
We want to the output to a list of only the file names which will then
be passed on to an image stacking program.
The CRVAL values are for the RA and Dec world axes respectively. Because
the FITS reference values must be in degrees the input RA unit is specified
as degrees. Because we want only the output file names we use keepcoords=no.
The output lists will be group_001.lis, group_002.lis, etc.