NRAO Applications Users Group Meeting - DRAFT MINUTES Date: 2007-01-17 (Wednesday) Time: 1300 MST (1500 EST) Video Hub: SOC-conf Rooms: SOC317/CV-ER311 ******************************************************************* Send corrections or additions to smyers 1. NAUG News o general news (Steve) We are continuing to have the weekly NAUG meetings to maintain momentum from the EVLA test. Next meetings are scheduled for Jan 24 (standard NAUG slot), and Jan 31. 2. CASA Status Report Standard items: o CASA status (Joe) - see main event o release status (Joe) - The new Alpha patch (previously scheduled for Jan 15) should be available today at the AOC, and Friday (or Monday) in CV. o use and testing status (Steve, Mark, Joe) - "User Support Fridays" are a GO for kickoff this Friday. Testers will be guaranteed support, although Joe has chosen this Friday for one of his Army-sponsored vacations. Users should try out regression scripts, their own data, and in particular functionality marked Resolved on the spreadsheet. o development targets (Joe) - See project office page: http://projectoffice.aips2.nrao.edu 3. ALMA - From last week: Debra says there is a new draft of the ALMA Offline Requirements document. Main changes are the addition of ACA and Single Dish requirements. See the SSR Wiki: http://almasw.hq.eso.org/almasw/bin/view/SSR/WebHome The document is under SoftwareSubsystemsDetailedRequirements: http://almasw.hq.eso.org/almasw/pub/SSR/SoftwareSubsystemsDetailedRequirements/offline.rqmts.ALMA-70.10.00.00-006-B-SPE.pdf Standard items: o ALMA Test status (Joe, Steve, Debra) FROM LAST WEEK: - NAUG pre-testing for the ALMA test plus simulator testing will be held in Febrary at the AOC. Crystal will be out here Feb 4-10 and Rob Reid Feb 4-17 (with some time spent at the ATF). Steve will be at the AOC from LANL for all of February for testing also. - Debra says that external testers have started booking their flights for March. Joe quipped that at least the GBT Orion data set would be testable. - upcoming test ALMA2006.10-6 (+single dish analysis). See: https://wikio.nrao.edu/bin/view/Software/AlmaTest5 http://projectoffice.aips2.nrao.edu/ALMA2007.03/ALMA2007.03.html o NAASC issues (John, Crystal) - CV will get CASA updates again starting Jan 19, including the Alpha patch. o ALMA development targets (Joe, Steve) - mosaicing and simulator developments (Kumar) FROM LAST WEEK: Kumar is working actively on the simulator tool. Some outstanding bugs are being fixed. Currently only the construction of the ms given times, configuration and setup, use of images for the signal, and addition of noise are supported. Gain corruption (e.g. from atmosphere) is pending Georges calibration framework improvements. Once the toolkit is working, Joe will construct tasks. 4. EVLA Standard items: o EVLA test results (Joe, Michael) - see Main Event - Joe would like scripts from the testers so regressions can be done. - See: http://projectoffice.aips2.nrao.edu/EVLA2006.12/EVLA2006.12.html - Testing spreadsheet: http://spreadsheets.google.com/pub?key=p9Q-V26qhzpmEtAF5AD4gOg o EVLA development plan & requirements FROM LAST WEEK: - Sanjay is working on the pointing calibration and correction. Some possible test data sets have been identified. Stay tuned for when this should be available for general testing. 5. Main Event - Presentations and Discussion Items *************************************************************** - Joe McMullin : This will be a follow-up from last week, with reports on further progress. See: https://wikio.nrao.edu/bin/view/Software/2007-Jan-17 See also NAUG minutes from Jan 10: http://www.aoc.nrao.edu/~smyers/naug/minutes/naug-2007-01-10.txt See also the Testing spreadsheet: http://spreadsheets.google.com/pub?key=p9Q-V26qhzpmEtAF5AD4gOg Some issues / discussion that came up: o Joe has crossed off some items on the Alpha Issues page and on the Test Issues spreadsheet. Look these over and when you get a chance try them out to verify that they in fact have been resolved o Joe is actively editing the Cookbook and will try to get it checked back in to the Wiki soon. o calibrator models are now available for almost all bands/sources. These are casa model images made from the clean components and correctly scaled to the observing frequencies of the image header. Later on, the total fluxes will be scaled on-the-fly to the setjy values, for example when observing other frequencies, as is done in AIPS. Use of models and uvrange will be illustrated using Crystal's test example data in the Cookbook. Currently models used for fluxscale. Frequency scaling not yet implemented (soon). Will eventually have IQUV models for polarization calibration, although there are some issues to be worked out. o selection - more features implemented field still uses auto-minmatch rather than explicit wildcarding so beware of spurious matches. Can use the selectfield function to see what the selection will return. antenna name matching has fixed so that "1" doesnt match "10 11..." in addition. Use selectantenna function to check. refant is now a name and not a number (refantid) in tasks so not backward script compatible. o There was much discussion on handling of EVLA, in particular the transition handling of EVLA vs. VLA antennas by name convention, e.g. EA# vs. VA#. AIPS uses the AN table to allow choice. There was also some concern that we ramp up preparation for issues that might come up for EVLA support. First hurdle is when the new online computers come in June 2007 and vis are written as raw correlation coefficient and thus need to have calibration applied (as in VLBA data). Note the new FILLM in AIPS will handle this through checking the toggle in the export data. The EVLA folks (Michael, Bryan) will keep us up-to-date on this. o "makemask" features and task cleanbox parameter has been added to clean and mosaic tasks. This takes a BLC,TRC list (to make easy pipelining) or a file name. This will support AIPS boxfiles, though currently it just takes lists of blc, trc in pixels or ra-dec. Support for the AIPS BOXFILE cards like C and F will come eventually. Joe will make it so that these are skipped in the meantime. Circles are on the way. There was request for support for MIRIAD style boxfiles also. There is also now a makemask task, which now supports the cleanbox param. Will eventually include polygons and interactive region setting. Other stuff on the way: 3D regions, facets/fields (C,F cards). Further in future is way to step through the facets/fields during interactive clean as in AIPS. Note that implementation is wrapped up in issues for calling of viewer within tasks, which must be worked out. o task history should write all parameters so you can rerun the task. Presumably explicit tool calls (ie. not through tasks) will do the same. There was a request for version info (presumably what release was called?). o flagging proposal - got some comments on Urvashi's flagging tool proposal. See link on wiki page. https://wikio.nrao.edu/bin/view/Software/2007-Jan-17 She has incorporated comments received, still time for more. Some discussion ensued, including use of a circular median on the complex visibilities (google gives some interesting links on this). There was some concern expressed about the ease of use of autoflagging controls. Should be careful to name things as sensibly as possible. Ed pointed out antenna based flagging could benefit from forming quantities equivalent to what calbrater gets. For example, excluding outlier gain amplitudes. Should be user accessible through the flag task, even if implemented using calibrater tool underneath. Should probably be a topic in upcoming NAUG or NAWG meeting (particularly for algorithmic aspects). "tvflg" functionality should be available in the new Alpha (presumably through the viewer). o Joe also seemed inundated by all the excellent user reqests and helpful advice. Some gems (not sure I got this straight): "If it peed orange juice it wouldn't have enough pulp." "I now understand why Eric is a grumpy guy." These were nominated as the quotes of the week. I can't think of a better way of kicking off our User Support Fridays! As Joe said: "get the party started" I'll keep you posted on Joe's upcoming breakdown/meltdown ... *************************************************************** Future Topics o Flagging (ms) implementation plan (Urvashi,George) o Clusters and CASA o suggestions welcome 6. AIPS++ Developments - see latest targets and info at Project Office: http://projectoffice.aips2.nrao.edu/ 7. Upcoming meetings and deadlines: o 2007 Feb 1 ALMA pre-testing starts o 2007 Feb 5-9 NAUG pre-testing at AOC o 2007 Mar 5 ALMA external test starts The agendas for past NAUG meetings are archived at: http://www.aoc.nrao.edu/~smyers/naug/agenda/ The minutes for past NAUG meetings are archived at: http://www.aoc.nrao.edu/~smyers/naug/minutes/