From gtaylor@cv3.cv.nrao.edu Mon Mar 5 10:00:03 2001 Date: Wed, 28 Feb 2001 10:49:47 -0700 (MST) From: gtaylor To: Steven T. Myers Cc: AIPS++ Core Testing Group , dshepher@zia.aoc.NRAO.EDU, Frazer Owen , efomalon@cv3.cv.nrao.edu, gmoellen@zia.aoc.NRAO.EDU, gtaylor@zia.aoc.NRAO.EDU, gvanmoor@zia.aoc.NRAO.EDU, jbenson@zia.aoc.NRAO.EDU, jhibbard@cv3.cv.nrao.edu, julvesta@zia.aoc.NRAO.EDU, tcornwel@zia.aoc.NRAO.EDU Subject: aips++ testing log of Feb 25-28 Testing AIPS++ Detailed Log Feb 25-28, 2001, session: note: TTT = trouble BBB = bug QQQ = question BBB: Bug AOCso02329 Tried off and on for 2 days to get aips++ to come up. Symptom was to just hang at startup with no logger or other windows coming up after the welcome to aips++ message. Turned out to be a corrupted log file. It is absolutely essential that aips++ inform the user that their log file is corrupt. Otherwise there is no easy way to know what is going wrong. It took 2 days to get around this problem here at the AOC with lots of experts on hand. For somebody in a University setting it could take a whole lot longer and really set them against aips++ before they ever got going. BBB: Tried to fill some L band VLA data from tape but the filler was broken. Went to Ralph and he seems to have fixed it. Filled my entire set of VLA data to /home/pegasus/aips++/calsfeb19.ms Tried to use imager to have a look at the MS and BBB: AIPS++ defect AOCso02322: imager failed trying to open ms Caught an exception! Event type=create exception=Tue Feb 27 19:51:09 2001 SEVERE imager::imager(MeasurementSet &theMS) (file /aips++/stable/code/trial/implement/MeasurementEquations/DOimager.cc, line 229): Open of MeasurementSet failed I suspect (and Ralph confirms) that imager doesn't like measurement sets containing both line and continuum observations. Still it would be better to see some kind of warning than a complete failure. Went back and reloaded just the L band continuum data. Ralph showed me how to enter a frequency selection too. Started up imager and ran setjy. BBB: Bug AOCso0232 Setjy uses the old Perley-Taylor 95 coefficients. Newer, 1999 coefficients are available and should be put in place. Ideally Setjy would have some parameter which one specifies to choose between 1995, 1999, etc. coefficients depending on the date of the observations. Now I want to use the calibrater tool to generate some solutions. BBB: Bug AOCso02326 In AIPS I used just the inner 4 antennas/arm, but in AIPS++ this is a pain. There is a way to set a uvrange, but no way to specify a select subset of antennas to solve for. Somewhere (maybe visplot) there is a nifty little GUI that would be good to have here as well. Otherwise it is rather tedious to do this with msselect. BBB: The solver blows up sporadically if there are autocorrelations present in the data. It should not be necessary to flag the autocorrelations or to explictly exclude the autocorrelations with the msselect field -- the solver should be aware not to use autocorrelations when solving for the antenna gains. Finally I've got this little dataset loaded and gain calibrated. Next comes polarization calibration ... --- end of testing. Frustration meter reads: |********** | low high