Monday Morning Meeting Feb 02, 2015

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

Attendance

  • Socorro: Jeff, Joe, Lindsey, Pam, Tak, Susan, Kumar, Juergen, George
  • CV: Darrell, Dave, Mark, Akeem, Andy
  • Garching: Julian, Sandra
  • SCO:

News / Meetings / Visitors

  • ALMA Polarization Meeting
  • Newsletter

Build, Release, Testing

  • import_asdm test failure options
    • ALMA problem probably just needs to be fixed
    • EVLA may be more substantial
  • OSX 10.10:
    • Getting time to switch to new compilers
  • Users Testers Meeting
  • libsakura is proceeding well
  • Need to explicitly install the mli-rpms
  • gc-casacore: progress

Architecture

  • Technical Forums:

Development

  • SetJy- Fix 4.3 release
  • Issue w/ self cal and model
  • VI/VB2: George progress on calibrating VI2 iterator

Pipeline

  • Heuristics release installed at NAASC (some issues with standard recipe versions), ESO, JAO (in progress), NAOJ(?)
    • ALMA pipeline, JAO catalog service still offline
    • EVLA pipeline, CASA setjy problem
  • Development plan for the next release in place
  • Testing underway on the cluster

AOB

Developer Reports

Monday Meeting
  • Sanjay Bhatnagar
  • Sandra Castro
    • Run the pipeline using a smaller MS and tested the MMS case. It runs and seg faults on clean. It seems that alma-m100-hpc used to also seg fault in clean. Maybe we should use a parallel version of clean for this test. Next, I will run this test in Socorro on lustre.
    • Testing mstransform with the new balanced mode of partition.
  • Lindsey Davis
    • Development (including imaging) plan discussions
    • Followup on pipeline problem tickets, investigate / reassign
  • Joe Depa
  • Pam Ford
    • Continued implementation and testing of MSTransform TVI in plotms
  • Kumar Golap
    • SDImaging Failure (exception handling)
    • msuvbin to do w-projection
  • Justo Gonz├ílez Villalba
    • Continued work with Pam to integrate TVI in plotms
    • Tested lazy mode in the MMS context
    • Provided support for translation of parallel imager from simple_cluster to mpi4casa
  • Jim Jacobs
  • Jeff Kern
  • Susan Loveland
    • CARTA links, animator, colormap work
    • CARTA histogram work with Amanda
  • David Mehringer
    • Completed adding "fit to half" algorithm functionality to ia.statistics()/imstat
    • Replied to Susan's query re: method to determine fraction in cumulative distribution function given a data value.
    • Fixed a couple of statistics code compile issues Ger found when using clang compiler.
    • Bemoaned long compilation times in debug-compile cycles when it was necessary to recompile all three trees and realized that in such instances, since I don't need the display nor the synthesis libraries, I can comment these and related libraries out of the code CMakeLists.txt file and decrease compilation times by nearly 50%. Still painful, but less so.
    • Spoke with Tracy Halstead and Darrell re: upgrading my machine to RHEL6. I thought the 4.3 release was the last release that CASA was going to support RHEL5, but Darrell tells me that's not the case and requested that I not upgrade so we can still have people developing on RHEL5.
    • Investigated but could not reproduce test_sdimaging failure in my sandbox, which showed a stats-related failure error message in the rpmbuild log. Looks like that test is no longer failing.
  • George Moellenbrock
    • Responding to Japanese queries re FEED in CalTables and general SD calibration
    • Various regression failure investigations (python plotting and GIL locks; asdm2MS)
    • JIRA triage of items slipping to v4.5
    • (CAS-7285,CAS-7286) Nearly finished initial CalibratingVi2 class (supporting OTF calibration), as well as supporting classes, including a generalized ViFactory that enables stacking the CalibratingVi with the (time) AveragingTvi2. Nearly ready to begin plugging it into mstransform, after which efforts shift to enabling the cal library to control how the data will be OTF-calibrated.
    • Prep for ALMA Polarization commissioning meeting (this week in Socorro)
  • Dirk Petry
    • still working on ephemeris and fits idi issues
    • migrated nightly sneeze machine to RHEL 6
    • with Adam Ginsburg: working on automating tests on Mac OSX
  • Urvashi Rao
  • Darrell Schiebel
  • Julian Taylor
    • fixed regressions introduced by releasing gil in plotting routines
    • setup local automatic regression test suite to prevent this from happening again
    • improved manual flagging performance by about x10 for 12k flags on 2GB data (CAS-7201)
    • removed a few hundred compiler noise warnings
  • Takahiro Tsutsumi
    • Just returned from vacation, worked for 3 days, cought up with e-mails
    • Fixed a SetJy bug (CAS-7298), started to work on fixes on SetJy for MMS support
  • Ger van Diepen
    • Merged in all NRAO changes till 2-Feb-2015; gave some feedback about compiler/doxygen warnings.
    • Run valgrind on all tests in casa,tables,scimath,measures,lattices. Fixed a few leaks and errors. Now all these tests are free of leaks and errors. Will do other parts soon.
    • Gave Shinnosuke the solution for the getcol problems in the CASA tests. Data type uInt is new in ValueHolder and is not handled by CASA.
    • Some email discussion with Darrell about move to gc-casacore.
Friday NAOJ Meeting
  • Kanako Sugimoto
    • sdms tool frame work ( CAS-6989): Done necessary developments for now. Ready to change status to RTT, but the codes are in branch and not testable until the branch is merged back. Holding merge until libsakura to be available for all developers.
    • multi-beam imaging issue of NRO data set ( CAS-6987): there was a bug. I believe I managed to fixed it locally. testing the fix now.
    • Discussed about user survey inquiries and sent list to Juergen.
  • Wataru Kawasaki
    • CAS-6702, CAS-6703 : modified rasterutil so that time range of detected raster rows/visits in the CASA logger become consistent with TIME value of spectra corresponding to edges of those raster rows/visits (committed to gcwrap:r32188), then asked testers to confirm the fix.
    • CAS-7001 : contacted with George M. about adding FEED_ID info to the NewCalTable. will Jiratize about this issue.
  • Masaya Kuniyoshi
    • CAS-6992: committed Unite tests for the tsdcal
    • CAS-6995: implementing the calmode 'apply' in the tsdcal
  • Takeshi Nakazato
    • CAS-6991: working on defining unit test and generating data for that
    • CAS-6993: e-mail based discussion with George
    • fix for sdimaging unit test, needs to be improved
    • supported Shinnosuke's work that creates Sakura rpm
  • Shinnosuke Kawakami
    • unittest/ regression test against compiled CASA(gc-casacore + code/gcwrap/asap) , Orion regression test failed, about 20 unittests related to getcol failed.
-- KanaSugimoto - 2015-01-29
Topic revision: r20 - 2015-02-02, 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