WHATSNEW in 31DEC99 AIPS
AIPS for the Ages
last edited 29 January 2000
31DEC99 is a new form of AIPS release. It will not be frozen on some
date and then shipped to any site desiring it. Instead it will remain
under regular development by what remains of the Classic AIPS Group
and will be available via the WWW to anyone who wants it at any time.
Most sites will probably fetch and install it much like previous
releases, but then they will run what is called the "midnight job" on
some time scale to apply the incremental changes that have been made
to the Charlottesville copy of 31DEC99.
Imaging:
- IMAGR
- The output names of all imaging tasks were changed to be
decimal rather than extended hex and to be 1-based rather
than 0 relative. The are now xCLnnn, xIMnnn, xBMnnn where
x is the Stokes and nnn = 001-512. Both the new and the
old names are supported by all CC model using tasks (e.g.
CALIB, UVSUB, ...)
- Added option ALLOKAY to allow IMAGR to restart without
recomputing beam images and without recomputing residual
data, if the user says that all is okay.
- Added the ability to image and Clean with both point and
Gaussian component models. "Simultaneous" Cleans with
multiple resolutions may help with the corrugation and
extended source problems traditional in Hogbom Cleans.
- A bug affecting multi-field gridding when the first field
was done in memory but later fields required disk-based
gridding was corrected.
- A bug affecting FILEBOX operations on boxes outside the
current filed of view was corrected.
- A bug causing IMAGR to use the wrong scratch files for the
gridded subtraction was corrected. This would appear as an
attempt to read beyond the end of file when the beam is the
same size as the image. Otherwise, it was probably
negligable.
- The maximum number of Clean boxes per field was changed so
that up to 2048 may be used so long as the number of fields
is <= 64. The maximum number decreases to 256 for 512
fields.
- SCIMG - a new task which is a multi-field version of SCMAP to do
iterative self-cal and imaging.
- MAPPR is a new procedure which simplifies the inputs to IMAGR for
imaging a single field without calibration, wide-field
options, multiple resolutions, and the SDI method.
Visibility data:
- EDITR, EDITA, SNEDT :
- The new CROWDED adverb controls whether multiple IFs and/or
polarizations may appear on the plots and be edited at the
same time.
- They now accept the REASON adverb for an initial value.
- They use certain fixed plot scales and mode settings when
starting in expert mode.
- They support scaling amplitude from 0 to max.
- They do a new TV option to display/edit the NEXT
CORRELATOR.
- To save time, difference data are computed only when
needed.
- The ANTUSE adverb is used to select which data are
displayed initially.
- SPFLG has a changed and more general meaning to the SWITCH
ALL-BL FLAG control.
- UVCOP did not do the proper time correction when copying a
single subarray from a multi-array data set. It subtracted
5 days too many.
- CALIB, FRING, KRING had problems re-referencing the phases when
delays and rates are present.
- SPLAT was changed to make a single-source output file properly
when only one source is selected.
- CONFI has new options to control/limit the configurations being
optimized.
- FRING and KRING now allow the data modeling routines to use
gridded model division under user direction (CMETHOD).
Display:
- Tick marks are now plotted enough so that some measure of the
scale is always visible.
- VPLOT can now plot various types of data while time averaging.
- FGPLT is a new task to display what baselines are flagged as a
function of time. Try it.
General:
- New adverbs BPRINT, EPRINT, NPRINT, NPLOTS replaced over-used
adverbs BCOUNT, ECOUNT, NCOUNT et al. in a number of
tasks.
- TBIN had problems with flagged table rows corrected.
- Single dish tasks: BSTST has an added Clean option, BSCLN is a
new beam-switched deconvolution task, BSMOD no longer
requires input BS data. OTFBS now writes a TY table.
- All CookBook chapters are being updated. Appendix Z and the
Table of Contents and Index are not yet done.The latest
version of each chapter can be accessed through an html
version of the Table of Contents
which also contains information about the revision history
of each chapter.