GGTAU
Testing - GG Tau regression script (continued)
DSS - 25feb07
-----------------------------------------------------------------------
Summary of issues:
- For 1mm continuum, I'm getting transfered flux densities of about
3.9 and 1.5 Jy. Script says flux densities of 4.3 and 1.8 are good
values - different enough that it looks like there may be a problem.
We think this is okay. Calibration/fluxscale are now paying attention
to flags from bad solutions. Should be better.
- Also, script has:
## Record flux values from logger window. Manually insert
## fluxes with imgr.setjy:
setjy('ggtau_07feb97.ms',fieldid=0,spwid=10,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=0,spwid=14,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=0,spwid=18,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=2,spwid=10,fluxdensity=[1.842,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=14,fluxdensity=[1.842,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=18,fluxdensity=[1.842,0.,0.,0.])
### Note, fieldid=2 should be 1 for swpid 10. Script should get
the wrong flux density for 0528 because of this.
- clean of HCO+ line forgot the mode='channel' keyword so a single
image with all channels is created. Given the nchan, start, width
specifications, it looks like this was not your intent. The clean
command in the script is:
clean('ggtau.hco.split.ms', imagename='ggtau.hco', alg='clark',
niter=100, gain=0.1, nchan=14, start=74, width=4,
spwid=0, fieldid=0, stokes='I',
weighting='briggs', rmode='norm', robust=0.5,
cell=[0.2,0.2], imsize=[256,256])
viewer
# only one channel is created. mode = 'mfs' is the default.
Yes - puzzled why it worked.
- viewer crashes when I try to close one of the images brought in as
raster images: Symptoms are:
I started viewing ggtau.1mm.image but then I wanted to bring up the
same image that I had created to compare with it. The image I
created is in another directory called
/home/sola/dss/casa.testing/casa/regression/ggtau/separate.processing
I couldn't tell what was happening so I clicked on "Raster image" 3
times. Then I realized I had 3 copies of the same image in the Data
register. I tried to close the 2nd 2 I had brought in but the
viewer crashed with no error message.
I started the viewer again, brought in the 2 desired images. Then
tried to close one of them after blinking between them for a while.
The viewer crashed with no error message.
David is looking into this.
- When I just ran the regression script using:
execfile 'ggtau_regression.py'
the regression part of the script that gives the timing did not
run. The script stopped at the generation of the 13co image...
I can't see any reason in the script why this would happen. No
error message was created. It just stopped.
Can't reproduce this. It is run several times a day now.
Actual detailed test notes:
alias casadaily 'source /home/ballista/casa/daily/casainit.csh; casapy'
casadaily
Regression scripts:
https://safe.nrao.edu/wiki/bin/view/Software/LegacyCASA
NOTE: refant = NAME now, not ID #
==> previous scripts will get the wrong antenna because they
specified ID #.
Useful commands to work around problems:
tb.clearlocks()
mp.clearplot
restore()
#
source /home/ballista/casa/daily/casainit.csh; casapy
Yesterday I finished the 3mm continuum processing. Today, do the 1mm
continuum and spectral lines.
default('setjy')
setjy('ggtau_07feb97.ms',fieldid=3,spwid=10,fluxdensity=[2.2,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=3,spwid=14,fluxdensity=[2.2,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=3,spwid=18,fluxdensity=[2.2,0.,0.,0.])
# CRL618 spwid= 10 [I=2.2, Q=0, U=0, V=0] Jy, (user-specified)
# CRL618 spwid= 14 [I=2.2, Q=0, U=0, V=0] Jy, (user-specified)
# CRL618 spwid= 18 [I=2.2, Q=0, U=0, V=0] Jy, (user-specified)
## Get first cut phase solutions to improve S/N for BPass determination:
default('gaincal')
gaincal('ggtau_07feb97.ms', 'ggtau.1mm.ph.gcal0', 'channel',
nchan=30, start=14, step=1,
msselect='FIELD_ID==1 && DATA_DESC_ID IN [10,14,18]',
gaintype='GSPLINE', calmode='p', splinetime=10000.,
refant='1', phasewrap=260, preavg=120, gaincurve=False,
opacity=False)
## Derive bandpass calibration for 1mm LSB:
default('bandpass')
bandpass('ggtau_07feb97.ms', 'ggtau.1mm.bpoly', 'channel',
nchan=60, start=2, step=1,
msselect='FIELD_ID==1 && DATA_DESC_ID IN [10,14,18]',
bandtype='BPOLY', degamp=6, degphase=6, visnorm=False,
bpnorm=False, maskcenter=4, maskedge=0, refant='1',
gaintable='ggtau.1mm.ph.gcal0', gaincurve=False, opacity=False)
plotcal(tablein='ggtau.1mm.bpoly',multiplot=false)
# looks good.
## Determine phase solutions for 1mm LSB for all calibrators
## using the 3mm phase solution, transferred
## to the 1mm bands, to reduce the phase rate. The
## net solution (transferred 3mm + 1mm) is written
## to the output table
default('gaincal')
gaincal('ggtau_07feb97.ms', 'ggtau.1mm.ph.gcal', 'channel',
nchan=60, start=2, step=1,
msselect='FIELD_ID IN [0,1,3] && DATA_DESC_ID IN [10,14,18]',
gaintype='GSPLINE', calmode='p', splinetime=10000.,
refant='1', phasewrap=260, gaincurve=False, opacity=False,
bptable='ggtau.1mm.bpoly', gaintable='ggtau.3mm.ph.gcal',
preavg=0.)
## Apply all solutions derived so far, determine
## calibrators' flux densities using a solve for T and
## fluxscale
default('gaincal')
gaincal('ggtau_07feb97.ms', 'ggtau.1mm.temp', 'channel',
nchan=60, start=2, step=1,
msselect='FIELD_ID IN [0,1,3] && DATA_DESC_ID IN [10,14,18]',
solint=600., refant='1', gaintype='T', opacity=False,
gaincurve=False, gaintable='ggtau.1mm.ph.gcal',
bptable='ggtau.1mm.bpoly')
plotcal(tablein='ggtau.1mm.temp',multiplot=false)
default('fluxscale')
fluxscale('ggtau_07feb97.ms', 'ggtau.1mm.temp', 'ggtau.1mm.flux',
reference=['CRL618 '], transfer=['0415+379','0528+134'])
Flux 0415+379 in
SpW=10 is: 3.94822 +/- 0.0690768 (SNR = 57.1569, nAnt= 5)
Flux 0415+379 in
SpW=14 is: 3.6116 +/- 0.0690842 (SNR = 52.2783, nAnt= 5)
Flux 0415+379 in
SpW=18 is: 3.96531 +/- 0.0749953 (SNR = 52.8741, nAnt= 5)
Flux 0528+134 in
SpW=10 is: 1.60516 +/- 0.0263439 (SNR = 60.9309, nAnt= 5)
Flux 0528+134 in
SpW=14 is: 1.48548 +/- 0.0188886 (SNR = 78.6443, nAnt= 5)
Flux 0528+134 in
SpW=18 is: 1.65109 +/- 0.0221837 (SNR = 74.4282, nAnt= 5)
Script says flux densities of 4.3 and 1.8 are good values. I'm
getting 3.9 and 1.5 - different enough that it looks like there may be
a problem.
## Record flux values from logger window. Manually insert
## fluxes with imgr.setjy:
setjy('ggtau_07feb97.ms',fieldid=0,spwid=10,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=0,spwid=14,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=0,spwid=18,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=2,spwid=10,fluxdensity=[1.842,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=14,fluxdensity=[1.842,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=18,fluxdensity=[1.842,0.,0.,0.])
#### Note that fieldid 2 for spwid 10 is incorrect, it should be 1.
Apply this flux scale, just change fieldid2 to 1 for spwid10:
setjy('ggtau_07feb97.ms',fieldid=0,spwid=10,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=0,spwid=14,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=0,spwid=18,fluxdensity=[4.310,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=10,fluxdensity=[1.842,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=14,fluxdensity=[1.842,0.,0.,0.])
setjy('ggtau_07feb97.ms',fieldid=1,spwid=18,fluxdensity=[1.842,0.,0.,0.])
## Amplitude calibration of 1mm LSB:
default('gaincal')
gaincal('ggtau_07feb97.ms', 'ggtau.1mm.amp.gcal', 'channel',
nchan=60, start=2, step=1,
msselect='FIELD_ID IN [0,1,3] && DATA_DESC_ID IN [10,14,18]',
gaintype='GSPLINE', calmode='a', splinetime=20000.,
refant='1', phasewrap=260, gaincurve=False, opacity=False,
bptable='ggtau.1mm.bpoly', preavg=2500.,
gaintable='ggtau.1mm.ph.gcal')
## Correct the target source and all other 1mm LSB continuum data:
default('correct')
correct('ggtau_07feb97.ms', msselect='DATA_DESC_ID IN [10,14,18]',
gaincurve=False, opacity=False, gaintable='ggtau.1mm.ph.gcal',
bptable='ggtau.1mm.bpoly', pointtable='ggtau.1mm.amp.gcal')
## Odd use of pointtable again...
## Split out calibrated target source 1 mm continuum data:
default('split')
split('ggtau_07feb97.ms', 'ggtau.1mm.split.ms',
fieldid=2, spwid=[10,14,18], nchan=60, start=2, step=1,
datacolumn='corrected')
default('clean')
clean('ggtau.1mm.split.ms', imagename='ggtau.1mm',
alg='clark', niter=100, gain=0.1,
nchan=1, start=3, width=58, spwid=[0,1,2], fieldid=0,
stokes='I', weighting='briggs', rmode='norm', robust=0.5,
cell=[0.1,0.1], imsize=[256,256])
# beam: 0.765417 by 0.385166 (arcsec) at pa 17.2348 (deg)
viewer
# image looks poor with stripes. There is a continuum source int he
center (maybe). Can't evaluate this image. Look at the calibrator
instead to see if there is a point source.
default('split')
split('ggtau_07feb97.ms', 'gcal.1mm.split.ms',
fieldid=1, spwid=[10,14,18], nchan=60, start=2, step=1,
datacolumn='corrected')
default('clean')
clean('gcal.1mm.split.ms', imagename='gcal.1mm',
alg='clark', niter=100, gain=0.1,
nchan=1, start=3, width=58, spwid=[0,1,2], fieldid=0,
stokes='I', weighting='briggs', rmode='norm', robust=0.5,
cell=[0.1,0.1], imsize=[256,256])
# beam: 0.75523 by 0.401485 (arcsec) at pa 18.6367 (deg)
viewer
# there is a point source in the center, some striping.
# 3mm HCO+(1-0) line calibration
default('bandpass')
bandpass('ggtau_07feb97.ms', 'ggtau.hco.bpoly', 'channel',
nchan=206, start=25, step=1,
msselect='FIELD_ID==1 && DATA_DESC_ID==6',
bandtype='BPOLY', degamp=1, degphase=1,
visnorm=False, bpnorm=False, maskcenter=40,maskedge=0,
refant='1', gaintable='ggtau.3mm.ph.gcal0',
gaincurve=False, opacity=False)
default('correct')
correct('ggtau_07feb97.ms', msselect='DATA_DESC_ID==6',
gaincurve=False, opacity=False,
gaintable='ggtau.3mm.ph.gcal', bptable='ggtau.hco.bpoly',
pointtable='ggtau.3mm.amp.gcal')
default('split')
split('ggtau_07feb97.ms', 'ggtau.hco.split.ms',
fieldid=2, spwid=6, nchan=206, start=25, step=1, datacolumn='corrected')
default('clean')
clean('ggtau.hco.split.ms', imagename='ggtau.hco', alg='clark',
niter=100, gain=0.1, nchan=14, start=74, width=4,
spwid=0, fieldid=0, stokes='I',
weighting='briggs', rmode='norm', robust=0.5,
cell=[0.2,0.2], imsize=[256,256])
viewer
# only one channel is created. mode = 'mfs' is the default.
default('clean')
clean('ggtau.hco.split.ms', imagename='ggtau.hco2', alg='clark',
mode='channel',
niter=100, gain=0.1, nchan=14, start=74, width=4,
spwid=0, fieldid=0, stokes='I',
weighting='briggs', rmode='norm', robust=0.5,
cell=[0.2,0.2], imsize=[256,256])
# beam: 1.86697 by 0.979325 (arcsec) at pa 15.9725 (deg)
viewer
# OK, now there is a channel map. I don't expect there to be
anything to see - emission is too weak.
# 1mm 13CO(2-1) line calibration
default('bandpass')
bandpass('ggtau_07feb97.ms', 'ggtau.co.bpoly', 'channel',
nchan=240, start=8, step=1,
msselect='FIELD_ID==1 && DATA_DESC_ID==22',
bandtype='BPOLY', degamp=1, degphase=1, visnorm=False,
bpnorm=False, maskcenter=4, maskedge=0, refant='1',
gaintable='ggtau.1mm.ph.gcal0', gaincurve=False, opacity=False)
default('correct')
correct('ggtau_07feb97.ms', msselect='DATA_DESC_ID==22',
gaincurve=False, opacity=False, gaintable='ggtau.1mm.ph.gcal',
bptable='ggtau.co.bpoly', pointtable='ggtau.1mm.amp.gcal')
default('split')
split('ggtau_07feb97.ms', 'ggtau.co.split.ms', fieldid=2,
spwid=22, nchan=240, start=0, step=1, datacolumn='corrected')
default('clean')
clean('ggtau.co.split.ms', imagename='ggtau.co', alg='clark',
niter=100, gain=0.1, nchan=14, start=91, width=4,
spwid=0, fieldid=0, stokes='I',
weighting='briggs', rmode='norm', robust=0.5,
cell=[0.1,0.1], imsize=[256,256])
## Again, no mode='channel' specified in clean, add it
default('clean')
clean('ggtau.co.split.ms', imagename='ggtau.co2', alg='clark',
mode='channel',
niter=100, gain=0.1, nchan=14, start=91, width=4,
spwid=0, fieldid=0, stokes='I',
weighting='briggs', rmode='norm', robust=0.5,
cell=[0.1,0.1], imsize=[256,256])
viewer
# nothing visible in even the 'mfs' image.
some striping, not a great image or calibration.
Now, just run the regression script to get final values and see what
it generates and how long it takes.
cd regression
execfile 'ggtau_regression.py'
# odd, the regression part of the script that gives the timing did not
run. The script stopped at the generation of the 13co image...
viewer
I started viewing ggtau.1mm.image but then I wanted to bring up the
same image that I had created to compare with it. The image I created
is in another directory called
/home/sola/dss/casa.testing/casa/regression/ggtau/separate.processing
I couldn't tell what was happening so I clicked on "Raster image" 3
times. Then I realized I had 3 copies of the same image in the Data
register. I tried to close the 2nd 2 I had brought in but the viewer
crashed with no error message.
I started the viewer again, brought in the 2 desired images. Then
tried to close one of them after blinking between them for a while.
The viewer crashed with no error message.
ggtau.1mm.images look comparable.
viewer
ggtau.3mm images look comparable.
OK, end testing for today.
--
JosephMcMullin - 26 Feb 2007