CARTA: Mostly testing and bug fixes, some region work, and lots of correspondence with others (South Africa/Taiwan).
CAS-9227: Add OBSERVATION table check to msmd.open(): RTV. implemented
CAS-9146: impv complains about non-square pixels when pixels are square: RTV. Fixed. Increased tolerance of Double comparison check.
Scratched my head more on how to edit "summary" field in task docs in plone without overwriting complete description. Need to talk to Jason. Apparently, there is a tutorial coming which has not been written yet about how to do this.
CAS-7687: CVEL and LSRK: IR. Requested a data set which illustrates the possible exportuvfits issue (it sounds like the title of the ticket may be misleading...)
CAS-9249: test_imsmooth[test_conserve_flux] fails on OSX Sierra: Resolved. Fixed. Slightly increased tolerance for float comparison.
Spent time fiddling with my Mac that would no longer boot. Handed it over to Josh Malone who fixed the OS upgrade issue.
Spent time doing binary revision search to determine which commit broke the ppdisk test and successfully isolated the specific commit (it wasn't mine).
CAS-9124: Reimplement mode='pola' in immath: RTV. Reimplemented/refactored.
CAS-9123: Reimplement mode='poli' in immath: RTV. Reimplemented/refactored.
CAS-9162: asdm2ms fails with error code 6 (was originally and msmd ticket): Not an msmd bug. Some sort of file system issue caused MS to be corrupted (no rows in OBSERVATION table). I added a check to msmd.open() to check for that (see CAS-9227).
CAS-9122: Reimplement mode='spix' in immath: RTV. Reimplemented/refactored.
CAS-9121: Reimplement mode='evalexpr' in immath: RTV. Reimplemented/refactored.
CAS-7173: typos in the help text for msmd.exposuretime: Resolved. typos corrected.
CAS-9062: Weight column is the sum of all channel weights in uvfits: IR. Still scratching my head over this as user provided data set with incorrect values for WEIGHT_SPECTRUM but claims importuvfits was responsible. However, I see no obvious change in the code that writes these weights in at least the past several years. It is the case that the WEIGHT column written by importuvfits is the sum of the absolute values of the channel weights. There seems to be uncertainty as to what exactly this should be (suggestions have been the first unflagged channel weight, or the mean or median of the channel weights, but this is not my decision to make, so someone with authority needs to provide a specific answer so I can implement it). And anyway, the WEIGHT column shouldn't be used at all if the WEIGHT_SPECTRUM is present, according the the MS2 spec. But user claims he has talked with various CASA deconvolution code experts who assure him WEIGHT, not WEIGHT_SPECTRUM, is being used in at least some cases. So this would seem to be either a miscommunication, or if it is indeed true, an incorrect implementation of the deconvolution code according to the MS2 spec.
CAS-9098: requantizer gencal slow, apparently due to inadequate MSMetaData cache---How to set this sensibly?
CAS-9096, sub-tickets: VI2 in solve continuing
Inhabited JIRA, sorted/scheduled tickets
Finished science PEP, started functional PEP
Judged Socorro 5th Grade Science Fair
Measuring memory usage for various options in tclean and (a) verifying that there are no leaks, (b) trying to predict the mem use and match it with measurements and (c) checking scaling.
Emails and minor code edits about an exit criterion based on detected divergence.
CAS-8413: difference between iomanip between libc++ (clang) and libstdc++ (gcc)
looked into CAS-4022
built all third-party package updates for EL6 and EL7 via docker
Testing and fixing the developer Jira/Bitbucket/Bamboo workflow with Pam. First pull request merged to master.
Fixed Carta Mac builds in Taiwan
Learned that Bamboo doesn't cleanup agent workspaces properly. Wrote an initial version of a cleanup script.
Testing and fixing master builds & tests
Worked on the new automask algorithm C++ implementation
Friday NAOJ Meeting
migration of sampling interval codes to c++ ( CAS-8942): nearly done.