Socorro: Jeff, Susan, Lindsey, Jim, Kumar, Pam, Sanjay, Urvashi, Juergen, Tak, Martin
CV: Ville, Darrell, Akeem, Dave
Garching: Justo, Julian
News / Meetings / Visitors
Build, Release, Testing
Last issue: Getting polarization info all the way to plotMS
Final Cookbook due tomorrow
Working on process for getting changes into github CasaCore. With any luck that should complete this week.
SVN tie-in for changes coming from Github will probably be manual until release is out.
Issue with SOFA in CASA Core
Martin to send out design.
Sanjay: Issue using GDB with MPI, Justo to produce document.
Preparations for CASA 4.4. / Pipeline Cycle3 R2 release
Regressions, bug fixing, release notes
Preparations for ALMA single dish pipeline commissioning
Ongoing EVLA pipeline commissioning, possible gaincal issue
Preparations on pipeline face to face meeting
Worked on refactoring and parallelizing code that generates and fills the convolution function cache. In progress.
CASA 4.4. / Pipeline Cycle 3 R2 release * Fixed some session / full polarization data set error handling, display, and QA bugs * Final regression testing * Verified (with pipeline team) that the plotms averaging problems are fixed * Drafted release notes
Finished reviewing chapter 8 of the technical handbook
Provided Urvashi with web log links to pipeline clean / tclean results
Fixed sub-table/reindexing related issues in mstransform
Discussions with Pam/Jim regarding correlation product selection and mapping
Changed how corrleation types are returned to user. Still some work to be done on that (both 4.4 and trunk).
Jeff Kern * Admin Stuff
CAS-7534 - Spectral profile fit does not appear to be displaying properly (resolved)
CAS-7532 - 2-d source fitter only appears to fit one source (under test)
CAS-7528 - 2-d source fitter cannot save estimate file to disk (resolved)
CAS-7539 - fitting Gaussians to lines from multiple cubes in optical mode fails with poor error message (resolved)
CAS-7533 - estimate selector in radio spectral profile fitting does not appear to work (resolved)
CAS-7536 - add text to optical mode of spectral profile fitter suggesting that users select "radio" (resolved)
CAS-7529 - viewer stuck on 'could not make a histogram from image' (resolved)
CAS-7537 - specifying intensity range in 2-d profile fitter confusing (resolved)
CAS-7535 - clicking multi-fit in spectral profile crashes viewer (resolved)
CAS-7530 - 2-d fitting estimate box is grayed out when you open an image when the fitter is already open (resolved)
CAS-7531 - posson fitter for 2d histogram does not appear to be working (ready to test)
CAS-7527 - new ellipse drawing slow when spectral profile viewer window is already open (resolved)
CAS-7207 - Flux density not working for multiple beam one per channel image (ready to test)
Finished removal of caption bar and adding user control visibility settings to persistent state.
Reworking layout in response to reviewer concerns about space usage.
Thanks to Ville, Darrell, and Josh, was able to build CASA for the first time in over 5 years on my Mac. Building on this machine is excruciatingly slow. However, I was able to fix both msmd and ia.pv() test failures on the release branch. I cannot fix the msmd test failure on trunk because of casacore freeze.
(CAS-7034) Development of new Common Statistics Framework: Fixed the single defect found so far by Peter's testing, committed to release branch, cannot commit to trunk because of casacore freeze
(CAS-7319) MSMD additions: Have completed implementing most of the 20 or so requested methods. Hopefully will complete this week.
Completed NRAO benefits survey.
(CAS-7485) Modify SubImageFactory interface: Responded to Erica's query re: testing.
(CAS-7583) listobs failure on CASA 4.4. prerelease: Confirmed that setting all STATE_IDs to -1 in invalid MS would produce a valid MS on which listobs would successfully complete and recommended this be done for the version of the MS publicly available. Juergen did so and apparently all is well.
(CAS-7532) 2-d source fitter only appears to fit one source: Discussed with user that it is quite unrealistic to expect code to fit 11 (66 free parameters) or even 5 (30 free parameters) Gaussians simultaneously in a low S/N (=6) image, especially when many of the sources being fit do not even appear Gaussian to the human eye.
(CAS-7573) imval should work on pV images: Fixed on release branch and trunk.
Communicated with Masaya re: new stats framework wrt sdstat/visstat.
(CAS-4283) implement ia.continuumsub()/imcontsub using ImageProfileFitter: Communicated with Catarina re: testing.
(CAS-7523) MSMetaData::getNetSidedbands() should return actual values from MS: Implemented, but cannot commit because of casacore freeze.
Significant time was put toward resolving various hardware configuration issues re: bringing my Mac back online.
(CAS-3034,5617) Various cal lib improvements for v4.4, including better bool parsing, comment (#) support, tinterp repair, applycaltocallib function; also only support file spec in tasks (since dict isn't well-supported in task interface)
Cal Library Quick Reference Guide (will be cookbook appendix)
(CAS-7478) uvcontsub repair
ALMA Poln CASAguide support, plus advised Crystal on calibration of M17 Zeeman tests
(CAS-7600) Fixed calwt=F override for XYf+QU solve
Lots of build study and experimentation (RHEL)
v4.5 planning, JIRA-ing
Science (EVLA circ poln obs planning)
Began integration of new code for parallelized iteration control in tclean into my CASA development repository.
Reviewed LatticeStatistics to prepare for migrating visstat to VI/VB2 and new statistics framework.
Wrote test programs for tclean and fixed several things along the way
103 tests so far, classified as onefield,multifield,iterbot,stokes,cube,widefield,mask,modelvis. More will be added.
Added custom functions for result/error checking, to derived versions of unittest python class.
Fixed several issues related to syncing of task/c++ params,return summary record, summaryplot, enabled more iterbot parameters, edits to log messages, exception handling in the task, mtmfs restoration change to allow restart options, removed some compile warnings,skipping minor cycle inits for chans/pols with zero upcoming iters, etc...
Modified MosaicFT (in my build) to connect it with MTMFS to enable wideband mosaicing for ALMA. Discovered that its normalization rules are different from (and incompatible with) those of AProjection, making it impossible for a neat single set of rules in the centralized normalizer class. Tried various numerical things. Also, since no normalization scheme here is strictly 'right' or 'wrong' , discussions about it went in a nice circle.
cvel updates to mstransform (CAS-7270, CAS-7382)
Propagated the fixes for restfreq = 0 handling in tclean to the prerelease
Post-fix testing on plotms correlation selections
Fixed some bugs in automask C++ method and enabled access of the functionality, though still limited, from tclean
Most of the modifications needed for tclean cube data/image partition have been checked in. Still working on a final piece to connect them in Python to be able to run parallel tclean in cube mode.
Friday NAOJ Meeting
A new developer, Renaud Miel, starting June 1st. He is suppose to work on 50% CASA+Sakura
Erik is re-organizing SD chapter of CASA 4.4 cookbook
All tickets of CASA 4.4 are resolved (user tests completed)
Sent release notes and known issue of single dish module in CASA 4.4 to Juergen.
Sent a draft of simulation chapter of cookbook to Juergen.
Investigated the cause of crash of sdimaging in CASA 4.3 on OSX ( CAS-7251): it turned out it was the number of open file issue. The default limit on OSX is 256. Noted as a known issue with instruction of how to avoid the issue.
Reported the issue of nearest neighbor interpolation in casacore found by Erik. George M. quickly fixed it. Thank you George!
MS transition: on-the-fly averaging in sdms tool ( CAS-7450): activated time average feature in underlying class, MSTransform. Need to support user defined weights. This requires slight extension to Averaging TVI. I'll discuss with Justo.
maskmode='list' in tsdbaseline ( CAS-6998): ticket closed
CAS-6999 : investigated the strange result reported by tester. found a bug in libsakura and fixed it. the latest libsakura bundled with CASA works correctly. ticket closed
output baseline table in tsdbaseline ( CAS-7001): done. ticket closed
baseline fitting using Chebyshev polynomials in tsdbaseline ( CAS-7443): done. ticket closed
applying baseline table in tsdbaseline ( CAS-7002): adding apply mode in the Python layer finished. implementation in the C++ layer is underway.
RTT CAS-7446 Provide a way to obtain pointing directions of an antenna corresponding to MS rows with data selection and transformations
issued some bugs/improvements: CAS-7481, CAS-7482 (essential issue in CAS-7488), CAS-7586, CAS-7590