Monday Morning Meeting 2 November 2015

  • USA Toll Free Number: 866-901-8266
  • USA Toll Number: +1-203-566-3863


  • Socorro: Jeff, George, Jim, Tak, Pam, Juergen, Susan, Kumar, Sanjay, Urvashi
  • CV: Andy, Akeem, Ville, Darrell
  • Garching: Julian, Justo
  • SCO:

News / Meetings / Visitors

  • Jason Kritter
  • Jen Donavan Myer

Build, Release, Testing

Cmake versions have been updated on the Linux build machines and in the Casa rpm repositories:

The most obvious change is for the Red Hat 5 users who need to install the package cmake28, and launch cmake with "cmake28" instead of "cmake". This is due to the package coming from epel repository instead of the standard Red Hat repository which still provides 2.8.5. The new EL5 version is, which is the last supported version for EL5.

The EL6 version is, which is the same version that is provided with the NRAO developer version install.

You also need to provide an additional cmake flag "-DBoost_NO_BOOST_CMAKE=1".



  • Warnings progress hadn't changed by Friday morning (see web page).


  • I've got the tests in MSVis building, but there was a problem getting them into the auto build. They'll go in today. A couple seem to be out of pattern for the automatic tests because they require the name of an MS to be input at startup.

--++++ Crash Recorder
  • Continued looking at the crash recorder which will capture information when CASA "crashes" (i.e. seg fault, unhandled exception, etc.). It's goal is to capture the stack trace and environment and then allow the user to send an e-mail to a CASA mailbox so we can track random crashed and maybe even fix them. A requirements document will be circulated soon.


  • Working on MMS support for OTF calibration trough mstransform. Some issues have been identify, possibly the need to add a reference to the top level MMS in the callib spec.


  • MSSection -


  • patches to C3R3: hifa_bpsolint, CASA interfaces
  • ALMA imaging pipeline testing
    • ALMA cubes are big: requires CASA 4.6 tclean channel chunking and / or HPC
    • feature finding issues: atmospheric lines residuals, edge effects
  • Long baseline campaign work
    • setup: machine configuration, initial runs
    • major issues so far: APDM changes, antenna position corrections


Developer Reports

Monday Meeting
  • Sanjay Bhatnagar
    • A-array data, finding some issues.
  • Sandra Castro
  • Lindsey Davis
    • pipeline task CASA interface fixes
    • followup on fixsyscaltimes behavior
    • long baseline campaign preparation
  • Pam Ford
    • CAS-8019 RTT - plotms autorange (max/nan issue)
    • CAS-7119 RTT - plot averaged weights (needed datacolumn set)
    • CAS-8018 - discussions with Justo about avg channel number instead of chanbin - now RTT
    • Work on export settings to improve plotms testing (CAS-7946) and other plotms export tickets (CAS-2502, 2579, 1977)
  • Kumar Golap
    • MS-UVBin as a flagger
  • Justo Gonz├ílez Villalba
    • Added support to retrieve input-ouput channel mapping in VI/VB2 and TVI
    • Working on MMS support for OTF calibration trough mstransform. See HPC section
    • Preparing list of targets for 4.6 and next year
  • Jim Jacobs
    • Renamed all of the derived duplicate classes that were playing a role in CAS-7733.
  • Jeff Kern
    • Investigated Structure-101
    • Various bits for release
  • Susan Loveland
    • Carta: Bug fixes & work on making color map settable on a per image rather than per stack basis.
  • David Mehringer
    • Discussion with Jim re: why build product gtest was appearing in source tree. Expressed my preference that no build products be placed in the source tree.
    • (CAS-8024) Move ImageAnalysis::fromRecord() to ImageFactory: RTT. Completed implementation.
    • (CAS-8027) Move ImageAnalysis::pixelvalue() to PixelValueManipulator: RTT. Implementation complete.
    • Fixed test_imregrid failure after being alerted by Julian. Apparently this has not been running automatically. Asked Ville and Akeem to please ensure it gets run automatically. Akeem says he has taken care of that.
    • (CAS-8049) move ImageAnalysis::toworld to ImageMetaData: RTT. Implementation complete.
    • (CAS-7449) MS transition: support per spectrum statistic calculation for MS: Some discussions about possible inefficiencies that seem to be related to iterator choice.
    • (CAS-7980) uid://A001/X121/X1bf weblog crash in final makeimages. WIP. Could not reproduce reported issue. Will look at code where log message originiates this week.
    • (CAS-8048) listobs requires 6 minutes to list an ALMA dataset with 12 minutes worth of 4ms dumps. Pointed out 1. that this data set has 170M rows, so it is going to take a bit for listobs to collect the data it reports, and 2. that on my machine, I can run listobs on this data set in 4.5 minutes, so that a hardware upgrade would provide an immediate 30% performance increase. Noted there is a very, very long and deeply nested loop in MSSummary that could possibly be made more effecient, so have been working on that. By gutting the loop and substituting MSMetaData calls, I've been able to improve the performance in that part of the code by about 30%.
    • Kana sent me email about github issues she is having that I forwarded along to Jim.
    • Attended HR Policy training.
    • Attended benefits meeting.
    • Completed Employee Safety Survey
  • George Moellenbrock
    • Finished PEP.
    • CAS-8051: Problems with cal library in selected MS contexts: fixed applycal w/ selection case; 'split' case (incl. MMS) tbd.
    • Consulted w/ Lindsey, et al. on CAS-8009: Bad time meta info for ALMA Tsys in SYSCAL subtable.
    • Writing VI2/TransformingVI patterns and improvements doc
    • Other miscellany.
  • Dirk Petry
  • Martin Pokorny
  • Urvashi Rao
    • some user support emails + laptop re-upgrade + documentation
    • went after some warnings in the synthesis module (partially through..)
  • Darrell Schiebel
    • worked more on linux app creation script (up to trying to replace lib references with location relative paths)
    • tried to find a linux/mac KVM
  • Ville Suoranta
    • New version of libsakura
  • Julian Taylor
    • CAS-7979 committed
  • Takahiro Tsutsumi
    • Updated some of the cube tclean unit tests and worked on C++ side of the tests a bit
    • Investigated "wrong mask velocities ... (CAS-7121)" when it is generated from the region (CRTF) text file.
Friday NAOJ Meeting
  • Kanako Sugimoto
    • Question to Jeff: response to plotms development questions?
    • Akeem reported the casapy on OSX 10.10 when running functional tests of tsdbaseline. It's likely libsakra in package was built with avx2 support which caused the crash on hosts only with lower levels of simd support, e.g., avx1.0 and/or sse4. It should be avoided by building libsakura explicitly with sse4 support only. Ville agreed with updating libsakura in dmg.
    • committed the changes to asap from Ville to support git repositories.
    • Slowness of beam size calculation ( CAS-7643): profiling of msmd.pointingdirection using Vtune amplifier (hung due to host issue).
  • Wataru Kawasaki
  • Masaya Kuniyoshi
  • Takeshi Nakazato
  • Renaud Miel
    • CAS-7339 No progress due to system administration priority. Will refocus on CASA development starting next week.
CONFLICT original 10:

-- KanaSugimoto - 2015-10-29-- KanaSugimoto - 2015-10-29
CONFLICT version 11:

CONFLICT original 8:
CONFLICT version new:

-- KanaSugimoto - 2015-10-29-- KanaSugimoto - 2015-10-29

-- KanaSugimoto - 2015-10-29
CONFLICT version 10:

-- KanaSugimoto - 2015-10-29-- KanaSugimoto - 2015-10-29
CONFLICT version new:

CONFLICT original 12:
-- KanaSugimoto - 2015-10-29
CONFLICT version 14:

-- KanaSugimoto - 2015-10-29
CONFLICT version new:

-- KanaSugimoto - 2015-10-29
Topic revision: r17 - 2015-11-02, 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