From gmoellen@cv3.cv.nrao.edu Fri Mar  9 17:07:42 2001
Date: Fri, 9 Mar 2001 16:52:13 -0700 (MST)
From: George Moellenbrock <gmoellen@cv3.cv.nrao.edu>
To: aips2-naug@cv3.cv.nrao.edu
Subject: msplot ideas



Hi All, 

Here is the list of enhancements that I recorded at Tim's msplot
demonstration on March 9.  I've organized them into four categories and
added a few which occurred to me in the process.  Please reply if I
omitted anything.

The majority amount to enhancement requests for what msplot and its c++
replacement should be able to do.  There are only four (probably only
three) distinct bugs which can frustrate use in fundamental ways.

Cheers,
George


Genuine bugs (these can prevent effective use): 
---------------------------------------------- 
o fix antenna/interferometer numbering/translation in display 
o fix "Locate" that didn't find anything
o fix reuse of image name in internals
o reconsider random spirograph option in viewer (a real bug?)


General (applies to both forms of data display):
-----------------------------------------------
o speed on large datasets?
o "Done" button next to "Dismiss" button, like every other tool.
o enable edit=T after startup
o Enable choice of plotting data relative to running mean for
  each baseline.
o Enable pre-average of data, then plot mean or rms
o Enable edit by clipping level, not just boxes
o Refresh button on plots so editing selection can be erased.
o Enable opportunity to write external flag tables at any time 
o Avoid display of generic TaQL since it uses zero-based counting
   (keep as hidden option to aid in debugging?)


XY-Plotting/Flagging:
--------------------
o Axis choice buttons:
  - button should change to selected value
  - order of axis choice options could be optimized (move
     common choices (data, time, uvdist) to top
  - add choice "channel" to see bandpass (this begs for
     option to average)
o report count of only those points that fall inside plot
o default plot ranges should obey unflagged data only
o more general use of color to segregate data in plot
  - not just polarization IDing
  - use only if informative (e.g., single color uv-plot
     unless, e.g., polarization availability is indicated)
  - colorize selected data and let colors persist in
     subsequent plots
o use same linear scale for u and v in u vs. v plot
o Enable "Locate" point nearest to cursor (in addition to
   within box, to avoid potentially long "Locate" listings


Data-as-Image Display/Flagging:
------------------------------
o zoom is hard to accomplish on edge of display
  (this is a known viewer defect and can be worked around
   zooming near the edge, then panning to it with the 
   'hand' button)
o Interferometer axis:
  - separate antennas sets in image
  - reordering by antenna1, antenna2, uvdist, etc.
o Flagged or absent data should be masked, not set to zero in image
o Handling of multiple spectral windows: 
  - extra image axis or stacked in a single plot? 
  - for stacked option, need to label axis properly
o Edit scope selection/visualization:
  - antenna selection by typed entry
  - baseline selection by vertical line
  - antenna selection by wider vertical line (block)
  - time-range selection by variable height block 
  - clarify SpW/pol/channel editing scope