Monday Morning Meeting March 9

  • DIAL-IN NUMBERS & PASSCODES:
  • USA Toll Free Number: 866-901-8266
  • USA Toll Number: +1-203-566-3863
  • PARTICIPANT PASSCODE: 725760

Attendance

  • Socorro: Jeff, Lindsey, George, Tak, Pam, Martin, Susan, Kumar, Jim, Sanjay and Urvashi
  • CV: Darrell, Dave, Akeem, Andy
  • Garching: Julian, Dirk, Sandra
  • SCO:

News / Meetings / Visitors

  • VLASS
  • Urvashi Return
  • Newsletter and Survey

Build, Release, Testing

  • 4.3.1 Issue with python path
    • If user has python path set, does not pick up the path.
  • Darrell: 7 errors left in the trunk. * Issue with table on trunk * Next up build on 10.10 * Not able to build on 10.8 * Plot-MS still in sandbox, doesn't quite work yet. * May still be issues with averaging performance * Jim working on it * Calibration stuff is in, but not exposed anywhere yet * Sensitivity calculation: should make it in to 4.4 * Linear mosaic checked in but being debugged

Architecture

  • Technical Forums:

HPC

  • Running CASA with MPI. See a quick-guide here
  • Run CASA's functional tests with the MPI framework. See a quick-guide here
  • Documentation about the Multi-MS structure
  • We have successfully run the ALMA pipeline with MMS. Next in the list is to run it with a combination of lazy import and MMS.
  • First scalability tests of running partition with MPI in 6,32 and 64 servers looked good and do not show any strange behaviour.
  • Sanjay has been testing tclean with the MPIInterface and has just produced scaling curves for the parallel execution of the time consuming part of imaging (gridding for PSF and residual image).

Development

  • Issue with definition of OpenMP Kumar is fixing in Code, Kana to fix in ASAP.

Pipeline

  • EVLA pipeline release commissioning - against CASA 4.3.1
  • ALMA pipeline release testing - against CASA 4.3.0
    • calibration / gainfield= nearest issue
  • pipeline CASA 4.4 testing
    • Python binding issue / tb.getcolkeywords,
    • HPC testing - plotms / partition efficiency issue
    • sensitivity calculation - im.sensitivity / units for ALMA data
  • pipeline face to face meeting will be June 15-18

AOB

  • Urvashi: Status of libsakura

Developer Reports

Monday Meeting
  • Sanjay Bhatnagar
    • MPI testing w/ tCLEAN
  • Sandra Castro
    • Consolidated the MPI log messages from Python and C++ layers so that no duplicate log origins are printed.
    • Created MMS data for local Jenkins jobs, which run functional tests with MMS for bandpass, gaincal, fluxscale, uvcontsub and wvrgcal. Will add more soon so that we have many tasks monitored daily.
    • Waiting for user input of problem raised on split when running on MS with 12 and 7 meters antennas. The problem seems to come from an earlier concat/split on the MS, which messed up the WVR spw information in the MS. (CAS-7297)
  • Lindsey Davis
    • refactored ALMA SNR heuristics code
    • pipeline release testing
    • provided input for ALMA ICT 2020 meeting
    • reviewed pipeline infrastructure code for security issues
    • pipeline CASA 4.4 integration testing
  • Joe Deep
  • Pam Ford
    • Further implementation and testing with George of plotms callib parameter (CalibratingVI2) and Calibration tab
      • Discovered issue that Record->XML when Record contains an array is unsupported in the casaqt code
    • Ran Python and C++ plotms tests - most worked
      • Discovered issue with plotting model data for MS with no MODEL_DATA column
      • Discovered re-indexing of indexed data in Main table -- will discuss with Jeff
  • Kumar Golap
    • Linear Mosaic
    • General Debugging
    • Function exists in toolkit but not yet in tClean
  • Justo González Villalba
    • Fixed tile miss-shape problem in the Multi-MeasurementSet structure (noticed when running plotms with MMS in the pipeline)
    • Implemented changes in MSTransformDataHandler to support virtual corrected data column operation
    • Added default implementation of phase shift in the plain VI/VB2 framework.
    • Fixed bug in phase shifting code reported by Pam
    • Restored casacore changes to support balance partition mode
  • Jim Jacobs
    • Finishing writing MS document
    • Issue of deploying developer CASA onto cluster
  • Jeff Kern
    • VLASS, Newsletter, Survey
  • Susan Loveland
    • CARTA Histogram
    • Viewer CAS-7370
  • David Mehringer
    • 3 days in office, 1 day EOS, 1 day vacation
    • Fixed impv bug introduced during casacore merge etal.
    • Andy stopped by to briefly discuss status of (CAS-6354) Rewrite ImageAnalysis::getFreqProfile() testing.
    • Some email exchange regarding if bug fix for handling non-float FITS images in ia.open() was going to be in 4.3.1, to which I didn't know the answer
    • Converted CAS-7268 from a listobs issue to a importuvfits issue, since it appears importuvfits is writing an internally inconsistent MS in this case.
    • Brief email discussion with Ger re: the variants of ImageUtilities::openImage()
    • Resolved (CAS-7328) imregrid does not rotate beam in J2000-GALACTIC conversion. Introduced new ia method, rotatebeam() to deal with this issue from the UI level, and modified relevant casacore classes.
    • Discussed with Susan (CAS-7370) beamArea in viewer statistics needs more general calculation.
    • Modified MSKeys.h, MSMetaData.h to add $Id: Mar0915.txt,v 1.14 2015/03/09 14:57:32 JeffKern Exp $ tags and to forward declare structs rather than classes to remove Clang warnings.
    • Created Safari Online 10 day trial account which I'm currently using to read Modern C++ (ie C++11/14) and Algorithms in a Nutshell, as well as various Python algorithm/data structure oriented books. Suggest management consider investing in a team account to foster professional development of software developers. This is a great resource and would cost significantly less than, eg, sending a developer to one conference/training opportunity per year.
  • George Moellenbrock
    • CAS-7285,7286: Completed final wiring of CalibratingVi2 to the synthesis calibration framework, thereby enabling non-trivial calibration via cal library. Also checked in upgrades to MSTransformManager so as to support seamless use of CalibratingVi2 there. Testing 'real' calibration in plotms hijacked by dbus issues: Vectors (e.g., spwmap) in the callib are not transmitted over the dbus.
    • Various code build/maintenance associated with the casacore migration, including work with Pam to sort out some mysterious bugs unwittingly introduced in the merge.
    • Investigated discrepancies in im.sensitivity. Instead of using the WEIGHT column on assumption that they are properly calibrated, it is instead using a ~arcane formula from TMS (that implicitly depends on weight uniformity) to approximate the net sensitivity. I will be checking in a new version that actually uses the WEIGHT column info directly, after some testing.
  • Dirk Petry
    • Working on ephemeris and wvrgcal bug fixes.
  • Urvashi Rao
  • Darrell Schiebel
  • Julian Taylor
  • Takahiro Tsutsumi
    • Fixed setjy unit tests as the format of 'Origin' string written to the history table has been changed (possibly due to the casacore merge?)
    • Wrote a helper function to map the internal array to store sum of weights in correlation basis (sumWeights) to those in Stokes plane corresponding to output images (related to CAS-7362)
Friday NAOJ Meeting
  • Kanako Sugimoto
    • sdimaging: set theoretical beam size to image ( CAS-5488) ... DONE
    • Update ATM library in CASA 4.4 ( CAS-7220): updated atmosphere using ATM codes in 2015-2 branch. Asking Dominique if that's it. ... DONE?/ONGOING?
    • Per spectrum fitting parameter in new baseline task ( CAS-7003): introduced a new blfunc ... ONGOING
    • works related to casacore merge: I believe asap is in good shape except for mapping of vector<bool>
      • merged Shinnosuke and Darrell's changes to asap at ATNF server. Coordination with Malte.
      • Darrell reported an issue in mapping of std::vector<bool> using (newer) boost on OSX 10.10. The definition is currently activated in asap. I'll try to setup 10.10 and reproduce the error on Fri.
  • Wataru Kawasaki
    • CAS-6997, CAS-6998: implemented basic unit tests for tsdbaseline, committed tsdbaseline codes and relevant updates for sdms tool to the trunk (code: r32551, gcwrap: r32552, casa-data:r10366), then set these tickets RTT.
    • deleted some obsolete data files which are no longer used in sdbaseline and sdmath unit tests (casa-data: r10373, r10374), and updated sdreduce unit test code (gcwrap: r32578).
  • Masaya Kuniyoshi
    • working on the unite tests of CAS-6995.
    • got a lecture for a new ticket CAS-6999 from kana & wataru.
    • started to work on CAS-6999: MS transition: enable blfunc='cspline' in the new baseline task for MS].
  • Takeshi Nakazato
    • CAS-6996 (otfraster sky calibration) is RTT
    • discussion with Bunyo about apply feature relevant to CAS-6993
    • discussion on CAS-5295 and CAS-7372 (PYTHONPATH and PATH issue on Mac release)
  • Shinnosuke Kawakami
-- KanaSugimoto - 2015-03-05
Topic revision: r14 - 2015-03-09, JeffKern
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