Socorro: Susan, Jeff, Juergen, Tak, Sanjay, Urvashi, Pam, Lindsey, Kumar
CV: Mark, Andy, Ville, Darrell, Dave
Garching: Julian, Sandra, Justo
News / Meetings / Visitors
No Meeting next week (6/27)
Build, Release, Testing
Github: Short term solution in place
Git Migration: First version of document out
Comments to Mark by next week
C++11 Can be used in code and gcwrap
RH 6.5 has a problem compiling because of boost version, but RH 6.6 does work.
Kumar has updated his directory and changed the flags
Issues with CASA-4.4
Unflagging of data in PolCal and BLCal
Added support for MMSs with FLOAT_DATA column inside importasdm (createmms section).
Added the tclean entries in taskwrapper, casa and casa_in_py.py to find the _pg files when run in parallel. This was making the pipeline fail.
Full successful run of the ALMA pipeline using Tier-0 parallelisation for calibration imaging. tclean paralell run in a local disk is 2.8 times faster than in sequential for X291 (3 EBs).
Verified correct behaviour of a EVLA parallel run prior to science testing verification.
tclean performance problems when running on MMS may be due to openMP settings in James tests.
Added a new summary view in flagdata, with field breakdown to support mosaic-like observations.
importASDM issue with ephemeris objects.
Several developers on vacation
Prepared initial CASA 4.3.1 / pipeline user distributions, tested Linux distribution
Final acceptance testing of CASA 4.3.1 / ALMA interferometry pipeline, waiting for final JAO acceptance
On-going CASA 4.3.1 / EVLA interferometry pipeline acceptance testing
Pipeline flagging refactor integration and testing
ALMA single dish acceptance testing, major efficiency improvements, pipeline execution times now linear with number of EBs
Areas of concern: antenna position corrections, tclean data selection, ephemeris target support
Fixed CAS-3080: flagmanager mode='save' -- default behavior should not destroy existing flagversion. Created functional tests for flagmanager.
Implemented the first version of time averaging in flagdata's clip mode, which uses the Time Average Iterator. The implementation is still under testing.
Finished the parallelisation of clearcal (CAS-4108).
User support on .casarc memory settings and clean's openmp settings.
Fighting to build CASA on RHEL 6 C++11 version. My advise when doing such big changes in CASA is: 1) post the list of RPMs needed; 2) post the full cmake options to build casa; 3) distribute the information to the developers.
tbuff parameter discussion with single dish developers
upgraded hif_antpos task, antenna position corrections discussion
tclean tests, data selection issues
reviewed / assigned pipeline jira tickets relevant for long baseline observations
Resolved several old plotms tickets that are no longer issues
Plotms tickets ready to test: CAS-3263, 7700, 7725, 7731
Tested CAS-7699 in plotms for Justo and passed along to the test group
making imageconcat standalone do virtual concatenation...fixed some stuff on the way along with extending imageOpener to deal with putting images inside one directory. Made the viewer load the virtual concatenation images
Migration of multiple builds to git and gcc4.9
Justo González Villalba
Implemented new flag summary mode to list a separated summary per field
Implemented interface changes in mstransform to use OTF calibration, and cross-validated it vs applycal
Implemented a workaround to by-pass MSSelection problems affecting correlation selection in mstransform
GUI Save as png image for desktop version
Zoom and pan reset
Miscellaneous bug fixes
Started work on contours
CAS-7531: Poisson Fitter for 2D Histogram does not appear to be working (ready to test)
CAS-6232: Writing version numbers in GUIs (ready to test)
CAS-7717: txt file created by the spectral viewer has wrong units (ready to test)
Met with Eduardo Miranda with local team for a total of 8 hours.
Open CAS-7723 re: github and NRAO svn casacore out of sync
Reported semi-infinite number of warnings re: Block.h when building under RHEL5
Got new machine and upgrade to RHEL6, lost about two days sorting out configure and build issues
Open CAS-7724, ImageExtend out of sync between github and NRAO svn casacore
Discussed various issues of stats framework with Martin
Talked with Darrell to find out where the current authoritative files re ownership for various tests live.
CAS-7319 MSMD additions: committed remaining bits and marked RTT
Communicated with Jim re: some casacore unit test failures I observed
Began migrating some code I was working on to C++11
CAS-7704 interaction of box/chans/stokes with region parameter: Completed implementation of requested change in rules, handed off to Mark to update documentation.
Responded to first round of tclean testing by CB/JO (connected a second clark minor cycle as the old clean apparently uses two different methods and I had earlier picked one that wasn't as robust as the other, changed pb norms to 1 at the location of the max instead of center, added internal T/F masks correctly for output images, made cube norms happen per channel, some interface parameter rearrangements (and xml experiments) ).
Wrote and tested a continuum subtraction script for the pipeline folks + some user support emails
Broke my build when updating before checking in any of the above, due to changes from CountedPtr to SHARED_PTR that had been done to the code tree and which my laptop environment isn't set up to handle. Could not resolve this quickly.
refactored part of flagcmd(action='plot'), added splitting plots into multiple pages for > 28 antennas (CAS-5187)
fixed difference between mstransform and cvel, mstransform did not use reference timestamp for reference frame.
fixed casacore build with gcc >= 5
Worked on fixing spw channel selection with stride > 1
Investigated tclean's seg fault. Probably this is caused by selectData() being not handling selection correctly when intent is involved.
Per CAS-5755, wrote a function to return a solar system object's angular diameter using me.comentangdiam() and JPL-Horizons ephemeris table. Its value seems to be slightly different from one obtained using solar_system_setjy function
Tried and struggled to setup casa build environment on OSX10.10