Date - 2000.09.05
Tester - S.T. Myers (AOC)
Platform - Linux (kernow)
Version - Stable (1.4 build 241)

Note: bugs are denoted by lines with prefix:
>>>BUG
and queries are denoted by lines with prefix:
>>>QUERY
while comments are prefixed by:
>>>COMMENT:

-------------------------------------------------------------------------------

GOAL: Start working Greg's Cband data testing into a useable script.
Start by following through Greg's script to get to the polarization 
calibration point.

Data taken 24 Feb 2000
AIPS Results in /home/kernow/smyers/VLA/Pol/20000224/
From Greg's Original Directory: /home/pegaso/gbt/aips++/
Working Directory: /home/kernow/smyers/Testing/Feb2000/

First trying out the version Weekly (1.4 build 269)

1. Make ms = cband.ms from CALS00FEB.CEDT  <== readfits.g

2. Try the Imagerwizard GUI.

   include 'imagerwizard.g"
   imagerwizard()  <== Actually, started it up from the Tool Manager

>>>BUG: When I clicked "Go" on the constructor a dialog box came up
   saying the scripter needed saving or dismissal.  I clicked "Save"
   and it bombed glish:

- "glishtk.g", line 118: warning, event glishtk.fail (<task:16>) dropped
LocalExec::SetStatus: abnormal child termination for /home/cluster/aips++/weekly/linux_egcs/bin/glishtk

[fatal error, 'segmentation violation' (signal 11), exiting]
Segmentation fault (core dumped)

   It left an empty "scripter.log_8464" file also -> delete.

   OK - try this again but in stable (1.4 build 241):

   This seemst to start up OK (I will go back later and see if I can
   recreate the crash in weekly) - continue.

3. Imagerwizard continued -- step by step

   (1) set ms by spanner from catalog gui - send&dismiss + Next->
   (2) choose first spectral window (IF 1) only + Next->
   (3) choose polarization calibrator 0927+390 only + Next->
   (4) starting imager (taking a long time) - set initial parameters
          cell size = 0.25 arcsec
          weighting = natural
       let it set the rest + apply + Next->
   (5) note it comes up with User Choice dialog asking to stop or
       continue cleaning, but without display this is meaningless!

       Comes up with a display showing a blank field - something
       happened.  I would have like to see a dirty map first.

       Cancel-> Abort (generated a screenload of red errors - I would hope
       an abort by the user would exit more gracefully!)

       I can't get glish to respond - will have to CNTL-C out!

       CNTL-C will not respond either, just hangs after 
       "exit glish (y/n)? y"

       had to kill -9 it!

>>>BUG: Generate bug report on this ==> Defect AOCso01739

       Note: after filing bug(), I tried to remake the fitstoms but it
       hung also.  But restarting the aips++ it worked.  Odd.

    Summary - this looks promising as a linear way through the
    process.  Will return to trying this later.

4. Make a dirty image of fieldid=1 using makedirty function:
#
# Function to make dirty map (stokes I) of measurement set = prefix.ms 
# (eg. cband.ms)
# Output files named by prefix+qual (eg. cband.run1.*)
# This will create and destroy its own imager object.
# Returns glish record of map stats
#
# Version STM 2000.09.06
#
function makedirty(prefix,qual,fldid,spid,cell_siz,imag_siz)
{	
   # Use imager to make large dirty image
   ms_name := spaste(prefix,".ms")
   imgr := imager(ms_name)
   imgr.setimage(cellx=cell_siz, celly=cell_siz, nx=imag_siz, ny=imag_siz, stokes='I', spwid=spid, fieldid=fldid)
   imgr.setdata(spwid=1, mode="none")
   imgr.weight('robust')
   imgr.uvrange()
   pre_name := spaste(prefix,qual)
   psf_name := spaste(pre_name,".psf")
   bmaj:=F; bmin:=F; bpa:=F; 
   imgr.makeimage(type='psf',image=psf_name)
   imgr.fitpsf(psf=psf_name, bmaj=bmaj, bmin=bmin, bpa=bpa)
   img_name := spaste(pre_name,".dirty")
   imgr.makeimage(type='observed',image=img_name)
   imgr.close()
   imgr.done()
   # Get stats of dirty image
   img_stats := imagestats( img_name )
   peak_flx := img_stats.flux
   peak_pix := img_stats.maxpos
   peak_dir := img_stats.direction
   img_rms := img_stats.rms
   print 'Peak in dirty image ',peak_flx,' Jy at ',dm.time(peak_dir.m0),dm.angle(peak_dir.m1),' at ',peak_pix
   print 'Map rms ',img_rms
   # return glish record of map stats
   return img_stats
}
# End of function makedirty.g

#
# Function to get image stats
# Returns glish records with max, maxpos, maxdirec, rms
#
function imagestats( img_name  )
{
   # Use image constructor to access dirty image and find peak 
   im := image(img_name)
   # Search inner quarter
   regoff := im.shape()/2; regoff[3:4] := [1,1];
   reglen := im.shape()/4; reglen[3:4] := [0,0];
   reg_blc := regoff - reglen
   reg_trc := regoff + reglen;
   r := drm.box(blc=reg_blc,trc=reg_trc)
   im.statistics(statsout=img_stat,list=T,region=r)
   img_max := img_stat.maxpos
   img_dir := im.coordmeasures(pixel=img_max).direction
   img_flx := img_stat.max
   # Get rms noise level from an outer region
   regoff := im.shape()/4; regoff[3:4] := [1,1];
   reglen := im.shape()/6; reglen[3:4] := [0,0];
   reg_blc := regoff - reglen
   reg_trc := regoff + reglen;
   r := drm.box(blc=reg_blc,trc=reg_trc)
   im.statistics(statsout=noise_stat,list=T,region=r)
   img_rms := noise_stat.rms
   im.close()
   im.done()
   # Return glish record, with fluxes as quanta
   return [ flux=dq.quantity(img_flx,'Jy'), maxpos=img_max, direction=img_dir, rms=dq.quantity(img_rms,'Jy') ]
}
# End of function imagestats

   Invoke these 

include 'makedirty.g'
include 'imagestats.g'

stat_rec := makedirty(prefix='cband',qual='.run0',fldid=1,spid=1,cell_siz='0.25arcsec',imag_siz=256)

   Output:

> Peak in dirty image  [value=0.0346042924, unit=Jy]  Jy at  
> 09:27:03.035 +039.02.20.852  at  [128 129 1 1] 
> Map rms  [value=0.00541110057, unit=Jy]

print stat_rec;

> [flux=[value=0.0346042924, unit=Jy], maxpos=[128 129 1 1] , 
> direction=[type=direction, refer=J2000, m1=[value=0.681361278, unit=rad], 
> m0=[unit=rad, value=2.47422495]], rms=[value=0.00541110057, unit=Jy]]

This looks like what it should be for uncalibrated data.

Note: this took about half a day as the name of one input parameter to
a function was changed since I first wrote these functions in May :-(

5. Set flux of 0927+390 (fieldid=1) to 10.65 Jy as determined using AIPS
and posted at http://www.aoc.nrao.edu/~smyers/calibration/C_band.html
Set Q and U (and V) fluxes to 0.0 initially

prefix := "cband"
qual := ".run0"
fldid := 1
spid := 1
cell_siz := '0.25arcsec'
imag_siz := 256

ms_name := spaste(prefix,".ms")

cal_flux := [10.65,0.0,0.0,0.0]

imgr := imager(ms_name);
ok := imgr.setjy(fieldid=fldid, spwid=spid, fluxdensity=cal_flux );
imgr.close()
imgr.done()
 
   Create calibrater object

calr := calibrater(ms_name);

ok := calr.setdata(mode="none" , nchan=1, start=1, step=1, mstart=[value=0.0, unit="km/s" ], mstep=[value=0.0, unit="km/s" ], uvrange=0, msselect='field_id == 1');

sn_name := spaste(pre_name,".sn")

ok := calr.setsolve(type="G" , t=60.0, phaseonly=F, table=sn_name , append=T);
ok := calr.solve();

ok := calr.setapply(type="G" , t=0, table=sn_name , select='field_id == 1');
ok := calr.correct();

   Got error message:

LocalExec::SetStatus: abnormal child termination for /home/sneffels3/aips++/stable/linux_egcs/bin/calibrater

   Last messages on the logger were:

Starting calibrater::correct
Loading G table from cband.run0.sn

   Did CNTL-C to kill, got message:

await terminated (file "servers.g", line 1016):
await theagent->.run_result theagent->.error theagent->.fail 

   See if things are still working.  Try browsing the .sn table...

nt := tablebrowser(sn_name)

   Just get blank table - are the calibrater tables different?
   Logger claims:

successful readonly open of auto-locked table cband.run0.sn: 29 columns, 0 rows
No rows in table!

   Try an oblique attack

ta := table(sn_name)

   Still has no rows ... regroup and try calibration session again.

6. The whole procedure was repeated from scratch (see file
   2000-09-06-myers.g).  It terminated in the same manner as before.

>>>BUG: Generate bug report on this ==> Defect AOCso01755

   Attach script 2000-09-06-myers.g in DDTS

End testing for now.