-- NicholasElias - 2010-06-01

Submitted by Dirk Petry (dpetry at eso.org)

CASA-SIT-TOGETHER 2 (26 March 2010)
===============================

Using script prepared by Dirk based on AIPS VLA K-band Spectral Line
analysis by Liz.

Used CASA 3.0.1 test release. casapy-test-30.1.10839-001.tar.gz
(but failed and used 3.0.0 eventually)

Notes and Bugs
---------------

1. Dirk filed tickets reporting bugs before the session
on importvla, split and on CASA test version 3.0.1 10839
that we were unable to use for the sit-together session,
found during preparation of the session materials.

CAS-2082 (resolved), CAS-2086 (open as of 28 May 10), CAS-2087 (resolved)

2. Plotms Bug (Mac Users). When plotms is exited, it leaves
behind an additional CASA application symbol in the toolbar.
This is confusing, as it is unclear which icon is due to
the open CASA session and which is due to the exited
plotms.

(Informed Darrell. May be resolved in 3.0.2)

3. REQUEST FOR PLOTMS: overplotting capability e.g. of
calibrated data overplot on rawdata.

4. Plotcal problem (already reported to Dirk by Wouter)

When making a different Bandpass table using (from the script, only
different table name):

          bandpass(
              vis = basename+'-spectral-selected-flagged-calibd.ms',
              caltable = basename+'-spectral2.bcal',
              field = '2', # bandpass cal
              spw = '', # all available, i.e. 0~1
              opacity = 0.0,
              solint = 'inf',
              combine = '', # i.e. only combine scans, the default
              refant = 'VA04',
              minblperant = 4, # the default
              solnorm = True, # should only be True when bandpass is
first step of cal
              bandtype = 'B',
              gaintable = '', # no previous gain table
              gainfield = '', # if there was a previous gain table, use
all fields
              gaincurve = True
              )

Besides the extra channels (which we seem to have solved) when I do

      plotcal(
          caltable = basename+'-spectral2.bcal',
          xaxis = 'chan',
          yaxis = 'amp',
          iteration = 'antenna, spw'
          )

it works normally, plotting the solution vs. channel number

However, when I plot the BPOLY solution

      plotcal(
          caltable = basename+'-spectral.bcal',
          xaxis = 'chan',
          yaxis = 'amp',
          iteration = 'antenna, spw'
          )

(so same setup, only different name)
It plots solutions vs. frequency, even though xaxis gives 'chan'.
I could imagine this being somewhat annoying should you for example
want to identify bad channels.

(no ticket filed, yet, as of 28 May 10, because plotcal is about to be 
replaced by plotms)

5. 'Self scale' issue in the spectral tool (another item from Wouter).
I think this is when we looked at a spectrum, then clicked far away
out of the window, and then it changed to mJy. But maybe we should wait
for the next version in which this will be solved?  (Liz note, so we can
try this in the CASA test version on Friday.)

(not confirmed, no ticket filed, yet, as of 28 May 10)
Topic revision: r3 - 2010-06-18, NicholasElias
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback