AIPS HELP file for MAXPIXEL in 31DEC24

As of Thu Oct 10 7:55:11 2024

HELP SECTION

MAXPIXEL
Type:  Adverb (scalar)
Use:   Specifies the maximum number of pixels to be searched
       for Clean components during each major cycle of the Clark
       version of Clean.
Null Value: 0 => 20050.
Tasks:
  APCLN.......Deconvolves images with CLEAN algorithm
  IMAGR.......Wide-field and/or wide-frequency Cleaning / imaging
  SCIMG.......Full-featured imaging plus self-calibration loop with
              editing
  SCMAP.......Imaging plus self-calibration loop with editing
  SDCLN.......Deconvolve images, switching to Steer-Dewdney-Ito
              algorithm after a while.
Procedures:
  LINIMAGE....Build image cube from multi-IF data set.  MAXPIXEL is
              passed to IMAGR.
"Verbs":
  OBITIMAG....Access to OBIT task Imager without self-cal or peeling.
              MAXPIXEL is used as in IMAGR,
  OBITPEEL....Access to OBIT task Imager with self-cal and peeling.
              MAXPIXEL is used as in IMAGR,
  OBITSCAL....Access to OBIT task Imager with self-cal, NOT peeling.
              MAXPIXEL is used as in IMAGR,


Discussion:
   AIPS Clean algorithm was designed by Barry Clark and modified by Bill
   Cotton.  In it, a certain number of the highest pixels in the
   residual image is loaded to the "AP" memory to be searched for
   components during each major cycle.  The performance and the results
   of the clean depend a lot on how many pixels are searched.  If too
   few are searched, then each cycle will find only a few components and
   there will have to be many (expensive) major cycles.  If too many are
   searched, then the algorithm wastes a lot of time searching and
   subtracting from pixels it does not ever need.  Also, when a large
   number are searched, a larger beam patch is needed to avoid making
   too much error inside the cycle.  Use this option sparingly, but
   values like 25000 or 30000 are found to speed cleans of large images
   that develop many approximately equal pixels in the residual image.

EXPLAIN SECTION