Talked to Tony R and student regarding "autoboxing"/boxit.
Reinstated some tests in imageanalysis.
ALMA large cube imaging
Fixed problem with modelvis prediction with chunking (with KG)
Enabled the addition of a 'last chunk' whose size is different from the rest when nchan%nchunks!=0
Added methods to SynthesusUtilMethods to print out (and write to a file) mem usage when called. Sprinkled calls all over the refactored imager code. Did some rough memory profiling with this. Some discussions about all this.
Trying to understand a minor memory leak between runs (maybe python? need to test) and looking into the role of loop ordering in SynthesisImager (data iterators vs chunk/model iterators).
User support emails with Claire and Frazer + minor code edits + jira to-do list updates in response to feedback.
worked on distribution creation with $ORIGIN
tried to find how to download and run Understand C++ for use on linux build server but failed (will try again)
Debugged a Casacore test crash when compiling with XCode 7 and submitted a patch, which Ger merged today.
The new Mac is out for delivery to NRAO today
Fixing IERSPredict table issues
Fixed a bug in SetJy s manual mode when spix=[alpha] along with IQU fluxes specified (CAS-8113)
Updated doxygen.cfg to remove deprecated modules from its build list and added some
Worked on updating C++ unit tests
Friday NAOJ Meeting
Reported an issue of calibration codes of Cycle 3 data and a work around. The issue has been handled by Takeshi this week.
Slowness of beam size calculation ( CAS-7643): modified msmd.pointingdirection method to access a cell instead of whole column. This had significant gain in performance. Getting directions of 450,000 pointings now takes 110 sec (previously estimated to take 25hrs). I will resolve the ticket on Monday.
New war against build warnings: done fixing code/singledish. Also sent TelCal team patches to fix build warnings and issues of ATM codes. Asked NAOJ developers for help on ASAP codes (due: end of the next week).
Removing deprecated single dish codes ( CAS-8114): a half done. plan to do the rest today.
Asked Akeem add new and remove deprecated SD functional tests from the critical tests invokes on nightly build. Akeem has done it. I'll check the test results of the last night to confirm it.
Created and updated tickets based on the plotms discussion on last last Friday.
Fixed CAS-8097, CAS-8098
Does anyone know how to fill FEED table properly? Does anyone know an example data set for multi-beam MS?