-- DirkPetry - 2012-10-10

Reports from the CASA 4.0 beta testing carried out at the EU ARC

Search for the string "ISSUE" to find potential issues. Will still be updated.

casapy-test-40.0.21143-001-2-64b, 2012-09-28, Roberto Galvan, 64bit RHEL

CLEAN: ISSUE 1 - STATUS: obtained data and script from reporter, posted on CAS-4471, got response from Urvashi: not a blocker, see CAS-4471

Trying to CLEAN an SMA uvdataset that consists of a 22-point mosaic x 4 epocs (at different array configurations).

The desired cube should have 1536x1536 pix (of 0.15" each) and 2 channels (more really, but this is a test as

I have noticed that CASA 3.4 can clean a few channels but crashes or stales when >~20 to 30 channels are set)

CASA 4.0 crashes with a message that is different from the problems in 3.3 and 3.4. This time is:

CASA <7>: execfile('w49concat_imaging_H30cube.py')

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%
Processing channel 1 of 2

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%
2012-09-28 13:12:09 WARN image::setboxregion (file /var/rpmbuild/BUILD/casapy-test/casapy-test-40.0.21143/gcwrap/tools/images/image_cmpt.cc, line 3265) THIS METHOD IS DEPRECATED AND WILL BE REMOVED. USE rg.box() INSTEAD.
*** Error * The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()
2012-09-28 13:12:09 SEVERE clean::::casa An error occurred running task clean.
0:13:58.005713

I don't use any box or region command in my script so I don't understand the error message.

Whereas the same dataset and imaging parameters work just fine (in 21 mins) in casa 3.4. Or at least the 2-channel test,

when increasing the number of channels to >30, casa 3.4 also crashes.

2012-10-02 15:13:24 INFO casa CASA Version 3.4.0 (release r19988)

%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%
Processing channel 1 of 2

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%
Processing channel 2 of 2

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%

0%....10....20....30....40....50....60....70....80....90....100%
0:21:46.385763

PLOTMS: Problem ISSUE 2 - STATUS: potential bug, use same data as posted for issue 1 on CAS-4471

After loading for the first time the data with:

CASA <2>: plotms(vis='smasher.concat.1_24.usb.lin.all_spw.ms',xaxis='freq',yaxis='amp',selectdata=True,field='0~10',spw='',averagedata=True,avgtime='86000')
NB: Iteration-capable plotting is ENABLED, but relatively new and somewhat experimental.
Out[2]: True

when I try to resize the window, it never updates again.

No error message, just empty plotms window.


casapy-stable-40.0.21234-001-64b, 01 Oct 2012, Anita Richards, platform Scientific Linux

Problem with applycal on MMS - ISSUE 3 - STATUS: this very probably resolved as in the test version 21234 an important fix to wvrgcal was still missing (is is in 21246)

I started working on BR1202_Band7, following the SV data script for calibration with changes as seemed required. I added
partition(vis='uid___A002_X36ffba_X3ef_orig.ms', outputvis='uid___A002_X36ffba_X3ef.ms', numsubms=8) 
at the start. This produced a warning (several times)
WARN    partition::Partition::selectSpw()       
Selected input spw(s) 9 10 11 12 13 14 15 16 17 18 19 20 21 22 were not found in DATA_DESCRIPTION and are being excluded.

The first few steps (listobs, fixplanets, flagdata, plotting Tsys) in the script then ran OK. I changed flagdata to tflagdata. I notice that the 'diameter' parameter has vanished, it might be an idea to reinstate this for general usability (or to include a tip/trick somewhere else for the easiest way to add this to the antenna table).

applycal (applying Tsys and wvr tables) was very unhappy (both sets of warnings repeated several times) *This also happens in 3.4 so probably it is an issue with these old data, not a 4.0 bug *
NewCalTable::confomant tabDesc is not superset of requiredCalMain

SEVERE  applycal::Calibrater::setapply(type, applypar)  Table column SPECTRAL_WINDOW_ID is unknown Check inputs and try again.

SEVERE  applycal::ms::open      Exception Reported: Table::lock cannot sync table 
/raid1/scratch/amsr/BR1202_Band7_UnCalibratedMSAndTablesForReduction/uid___A002_X36ffba_X3ef.ms/SUBMSS/uid___A002_X36ffba_X3ef.0000.ms; 
another process changed the number of columns

applycal::::casa        An error occurred running task applycal.
and indeed, inspection in plotms suggests that no calibration has been applied.


Problem with setjy - ISSUE 4 - STATUS: This is caused by the non-availablility of the Butler 2012 models in this test version, should be fixed now

Try again on IRAS16293

Convert to multi-ms

Listobs is OK

plotms(vis='X1ed.ms', xaxis='time', yaxis='elevation', coloraxis='field', antenna='DV03&DV04')
all the points come out in all the colours. If I plot one field at a time they come out in different colours.

setjy does not work on multi-ms
setjy(vis=vis, field='Neptune', standard='Butler-JPL-Horizons 2012',scalebychan=F)
MSSelectionNullSelection : The selected table has zero rows. 

but this works OK on pre-multi ms



casapy-test-40.0.21143-001-2-64b, 2012-10-01, Rosita Paladino, 64bit Linux

I used one of SV M100 dataset to run a simple calibration script:
os.system('mv X146-line-ave.ms X146-line-ave_ORI.ms')

partition(vis='X146-line-ave_ORI.ms', outputvis='X146-line-ave.ms',numsubms=8) 

setjy

Running setjy gives some warnings and flux measurements different from those obtained with CASA3.4 and 3.3:
 
setjy(vis="X146-line-ave.ms",field="1",standard="Butler-JPL-Horizons 2010") 

INFO setjy::Imager::open() Opening MeasurementSet /arcfs0/homesarc/paladino/DATI_ALMA/CASA_TEST/CASA_40/X146-line-ave.ms/SUBMSS/X146-line-ave.0000.ms
INFO setjy::imager::setjy() Using channel dependent flux densities
WARN setjy::imager::setjy() No data was selected for field 1.
setjy::imager::setjy() Titan (fld ind 1) spw 0 [I=0.26328, Q=0, U=0, V=0] Jy, (JPL-Butler Solar System Object)
setjy::imager::data selection Selected 1980 out of 7920 rows.
setjy::imager::setjy() Titan (fld ind 1) spw 1 [I=0.25482, Q=0, U=0, V=0] Jy, (JPL-Butler Solar Sy
SETJY::IMAGER::DATA SELECTION SELECTED 1980 OUT OF 7920 ROWS.
SETJY::IMAGER::SETJY() TITAN (FLD IND 1) SPW 2 [I=0.21944, Q=0, U=0, V=0] JY, (JPL-BUTLER SOLAR SYSTEM OBJECT)
SETJY::IMAGER::DATA SELECTION SELECTED 1980 OUT OF 7920 ROWS.
SETJY::IMAGER::SETJY() TITAN (FLD IND 1) SPW 3 [I=0.21195, Q=0, U=0, V=0] JY, (JPL-BUTLER SOLAR SYSTEM OBJECT)

The reported fluxes are different from what casa 3.3 and 3.4 found ISSUE 5 - STATUS: This is caused by the recent changes in fluxscale, not a bug

SETJY() TITAN (FLD IND 1) SPW 0 [I=0.26754, Q=0, U=0, V=0] JY, (JPL-BUTLER SOLAR SY$
SETJY::IMAGER::DATA SELECTION SELECTED 2730 OUT OF 123760 ROWS.
SETJY::IMAGER::FT() FOURIER TRANSFORMING: REPLACING MODEL_DATA COLUMN
SETJY() TITAN (FLD IND 1) SPW 1 [I=0.25902, Q=0, U=0, V=0] JY, (JPL-BUTLER SOLAR SY$
SETJY::IMAGER::DATA SELECTION SELECTED 2730 OUT OF 123760 ROWS.
SETJY::IMAGER::FT() FOURIER TRANSFORMING: REPLACING MODEL_DATA COLUMN
SETJY::IMAGER::CREATESKYEQUATION() PROCESSING AFTER SUBTRACTING COMPONENTLIST _SETJY_SPW1_TITAN_112.78$
SETJY::IMAGER::CREATEFTMACHINE() PERFORMING INTERFEROMETRIC GRIDDING...
SETJY::IMAGER::SETJY() TITAN (FLD IND 1) SPW 2 [I=0.21555, Q=0, U=0, V=0] JY, (JPL-BUTLER SOLAR SY$
SETJY::IMAGER::DATA SELECTION SELECTED 2730 OUT OF 123760 ROWS.
SETJY::IMAGER::FT() FOURIER TRANSFORMING: REPLACING MODEL_DATA COLUMN
SETJY::IMAGER::CREATESKYEQUATION() PROCESSING AFTER SUBTRACTING COMPONENTLIST _SETJY_SPW2_TITAN_102.72$
SETJY::IMAGER::CREATEFTMACHINE() PERFORMING INTERFEROMETRIC GRIDDING...
SETJY::IMAGER::SETJY() TITAN (FLD IND 1) SPW 3 [I=0.20813, Q=0, U=0, V=0] JY, (JPL-BUTLER SOLAR SY$
SETJY::IMAGER::DATA SELECTION SELECTED 2730 OUT OF 123760 ROWS.
gaincal and bandpass

gaincal and bandpass seemed to work properly.
Fluxscale

ignores the parameter 'transfer'. Running:
fluxscale(vis="X146-line-ave.ms",caltable="X146-line-ave.infPhase.AP.G",fluxtable="X146-line-ave.flux",reference=["Titan"],transfer=["*Phase"],append=False,refspwmap=[2, 2, 2, 3]) 

the result is always the transfer to all calibrators, and the fluxes for the two phase calibrators, are again different from previous CASA3.3 results:

Results in CASA4.0

FLUXSCALE:::: FLUX DENSITY FOR 1224+213 PHASE IN SPW=0 (REF SPW=2) IS: 1.22694 +/- 0.0192806 (SNR = 63.6$
FLUXSCALE:::: FLUX DENSITY FOR 1224+213 PHASE IN SPW=1 (REF SPW=2) IS: 1.25215 +/- 0.0178779 (SNR = 70.0$
FLUXSCALE:::: FLUX DENSITY FOR 1224+213 PHASE IN SPW=2 IS: 1.28571 +/- 0.0170686 (SNR = 75.3259, N = 26)
FLUXSCALE:::: FLUX DENSITY FOR 1224+213 PHASE IN SPW=3 IS: 1.30962 +/- 0.0181049 (SNR = 72.3348, N = 24)
FLUXSCALE:::: FLUX DENSITY FOR 3C273 - PHASE IN SPW=0 (REF SPW=2) IS: 7.84148 +/- 0.0281546 (SNR = 278.5$
FLUXSCALE:::: FLUX DENSITY FOR 3C273 - PHASE IN SPW=1 (REF SPW=2) IS: 7.87668 +/- 0.0246188 (SNR = 319.9$
FLUXSCALE:::: FLUX DENSITY FOR 3C273 - PHASE IN SPW=2 IS: 8.41605 +/- 0.0252413 (SNR = 333.424, N = 26)
FLUXSCALE:::: FLUX DENSITY FOR 3C273 - PHASE IN SPW=3 IS: 8.58803 +/- 0.0298713 (SNR = 287.502, N = 24)

and results in CASA3.3
FLUX DENSITY FOR 1224+213 PHASE IN SPW=0 (REF SPW=2) IS: 0.566424 +/- 0.0146724 (SNR = 38.$
FLUX DENSITY FOR 1224+213 PHASE IN SPW=1 (REF SPW=2) IS: 0.815983 +/- 0.0153019 (SNR = 53.$
FLUX DENSITY FOR 1224+213 PHASE IN SPW=2 IS: 1.21837 +/- 0.017551 (SNR = 69.419, N= 26)
FLUX DENSITY FOR 1224+213 PHASE IN SPW=3 IS: 1.2407 +/- 0.0235755 (SNR = 52.6267, N= 24)
FLUX DENSITY FOR 3C273 - PHASE IN SPW=0 (REF SPW=2) IS: 4.11298 +/- 0.0270927 (SNR = 151.8$
FLUX DENSITY FOR 3C273 - PHASE IN SPW=1 (REF SPW=2) IS: 5.71944 +/- 0.0264656 (SNR = 216.1$
FLUX DENSITY FOR 3C273 - PHASE IN SPW=2 IS: 8.39618 +/- 0.0287451 (SNR = 292.091, N= 26)
Flux density for 3c273 - Phase in SpW=3 is: 8.57712 +/- 0.0500491 (SNR = 171.374, N= 24)

2012/10/02

I just read in CASA4.0 release notes that fluxscale uses the median instead of the average for flux scaling, so these differences are due to that change maybe.
Applycal

worked properly on this dataset:
applycal(vis='X146-line-ave.ms',
            field='3c273 - Phase',
            gaintable=['X146-line-ave_Bandpass.B','X146-line-ave.ScanPhase.G','X146-line-ave.flux'],
            gainfield=['*Band*','1224*','1224*'],
            interp=['nearest','linear','linear'],
            calwt=F, flagbackup=F) 

Plotms does not allow to change interactively parameters like field, spw, etc.

Only using the command line it is possible to visualize different fields.
plotms(vis='X146-line-ave.ms',field='3c273 - Phase',xaxis='freq', 
          yaxis='amp', ydatacolumn='corrected',spw='0')

Split ISSUE 6 - STATUS: This is the expected output, not a bug

worked properly but gave some warnings I don't understand:
WARN split::::casa * Summary: there were failures in 6 SUBMSs:
WARN split::::casa * (these may be harmless if they are caused by selection):
WARN split::::casa X146-line-ave.0000.ms: False
WARN split::::casa X146-line-ave.0001.ms: False
WARN split::::casa X146-line-ave.0002.ms: False
WARN split::::casa X146-line-ave.0004.ms: False
WARN split::::casa X146-line-ave.0005.ms: False
WARN split::::casa X146-line-ave.0006.ms: False

pclean - Problem: pclean does not work in interactive mode ISSUE 7 - STATUS: potential bug

(task clean works properly with the same inputs, but mode='mfs')
 pclean(vis='X146-line-ave.ms',
              imagename='X146_3c273Phase',
              field='3c273 - Phase',
              mode='continuum',
              cell=['0.5arcsec','0.5arcsec'],
              imsize=[200,200],
              niter=100,
              interactive=True)

it starts and runs until:
doing weight a.getftweight(msname="X146-line-ave.ms", wgtimage="/arcfs0/homesarc/paladino/DATI_ALMA/CASA_TEST/CASA_40/cluster_project/Temp_0_rvcWtEub.wgt")
intmask 0

then it opens the viewer, after selecting the region to clean, it crashes:
SEVERE    simple_cluster::::casa    An error occurred running task pclean. 

parallel resources - Problem ISSUE 8 - STATUS: probably caused by missing ssh setup, testers provided data and script, forwared to JA Gonzalez

We tested some tasks also using the parallel resources, initializing them with:
from simple_cluster import simple_cluster
simple_cluster().init_cluster('casa_cluster_spec')

The task applycal did'nt work:
SEVERE ParallelTaskHelper::::casa An error occurred running task applycal

the task pclean (clusterdef='casa_cluster_spec') crashed at the same point as without using the parallel resources.
 

viewer - issue with line search ISSUE 9 - STATUS: potential improvement request

Very nice new features!

I like the short header in the loading options it helps a lot in selection. The

spectral profile tool legend preferences: does not print the legend on the plot if location=top left is selected.
molecular line search finds some lines but does not overlay them on the graph.
The search of lines by "chemical name" needs one to know exactly the name of the molecule and the way it is written in the catalog (since it is case sensitive)
Would'nt it be better to search by chemical species?

viewer - issue with region shape manager ISSUE 10 - STATUS: potential bug or lack of documentation

I was not able to load regions from the region shape manager, I don't know if I missed something.



CASA-intel64b-10.6-21234.dmg, 2012-10-08, Attila Juhasz, 64bit OS X 10.6.8

viewer

Cool new features!

- If I started the viewer with the file name as input I got a raster image by default, but when I tried to open a file in an already existing viewer window the 'raster image', 'vector map' buttons do not seem to work in the file opening dialog, I could only plot contour lines.

- I opened 101x101x60 cube and tried the collapse tool to calculate 0th, 1st, 2nd moment maps but it was really slow (took ~2 min vs. ~1sec in case of immoments)

- Spectral profile window, Legend preferences, I couldn't see any effect of the color bar checkbox

viewer - issue with spectral viewer resize ISSUE 11 - STATUS: potential request for improvement

- It would be nice if the spectral profiler window had scrollbars when the the window is resized to a small size. When I tested the viewer on a low-res screen (1280x800) I could just about see that there are three buttons/tab headers at the bottom but I could not read their labels (labels were off the screen/window) and I could not do anything about it.

viewer - issue with line profile fitting panel ISSUE 12 - STATUS: potential request for improvement

- In the line profile fitting panel the information text and the min/max velocity input fields fell on top of each other (again probably because of the low-res screen). Resizing the window did not have any effect as the container tab widget did not change its size when the window is resized.

- When lines are found in the line search field I could not overlay them on the spectrum either.
simobserve/simanalyze/simalma

simobserve and simanalyze seemed to work as before

Run simalma with and without ACA and it seemed to work as expected, although I got a few warnings that it tried to close a file which was already closed:

WARN image::close (file /opt/casa/stable-2012-09/gcwrap/tools/images/image_cmpt.cc, line 659) Image is already closed
Topic revision: r6 - 2012-10-19, DirkPetry
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