---------------------------------------------------------------------------- Bug AOCso02254 problems Submitted 010205 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010205 1 enclosure "severe error : unhandled mount type" Assigned engineer: gmoellen ************ CR_Description (Added 010205 by jbenson) *********** Attempting to construct an imager tool on filename : /home/pecos2/aips++/3C120_171.ms. VLBI data from AIPS via fitstoms. This data file has antenna mount types 'ALAZ','EQUA','ORBI' as listed by AIPS PRTAN. Even though visibilities from the 'ORBI' antennas are not present in 3C120_171.ms, the antennas are present in the AIPS AN table and I suspect are somewhere in aips++. 'ORBI' is a legal mount type, it indicates a space vlbi tracking station. ---------------------------------------------------------------------------- Bug AOCso02241 problems Submitted 010126 RESOLVED defect report AIPS++, version 1.5 (bui Assigned 010126 2 enclosures Opened 010205 Resolved 010205 "Visplot failure using Field Id's" Detection method: customer use Assigned engineer: tcornwel ************ CR_Description (Added 010126 by jbenson) *********** Attempting to plot /export/home/pecos2/aips++/ab973test.ms. Visplot doesn't correctly interpret the Field Id indices specified in the Select Ranges window. Five sources in ms, field id's 1,2,12,13,53 Will successfully plot field id 1 or 2 or all 5 only. ************ Resolution (Added 010205 by tcornwel) *********** Please describe your resolution to this problem: John, I could not reproduce this problem. I'm checking in a new version this week that may have the problem fixed if it ever was there (how's that for weaseling?). It'll be called msplot.g and will be available in weekly next week. Cheers, Tim ---------------------------------------------------------------------------- From kgolap@zia.aoc.NRAO.EDU Wed Feb 14 16:27 MST 2001 Date: Wed, 14 Feb 2001 16:27:40 -0700 (MST) From: Kumar Golap To: jbenson@zia.aoc.NRAO.EDU cc: gmoellen@zia.aoc.NRAO.EDU Hi I have submitted a bug as it is really 'image' that is failing to accept a union of regions. Kumar --------- Hi Kumar, Thanks for the advise on how to get around the union bug. It worked. While I was experimenting with functions in the regionmanager tool, I believe I found a bug in 'pixeltoworldregion'. The example in the Users Reference Manual does works. Thanks, John. # # Simple test of drm.pixeltoworldregion function. # include 'image.g'; include 'coordsys.g'; include 'regionmanager.g'; # x3C273im := image('3C273XC1.restored'); # # cs := x3C273im.coordsys(); ok := cs.summary(); ok := drm.setcoordinates(cs); # r1 := drm.box(blc=[100,100],trc=[150,150]); #r2 := drm.box(blc=[151,100],trc=[200,150]); # r11 := drm.pixeltoworldregion(x3C273im, r1); # ---------------------------------------------------------------------------- Bug AOCso02301 problems Submitted 010222 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010222 1 enclosure "Errors in u,v,w and exposure and interval columns" Detection method: customer use Assigned engineer: gmoellen ************ CR_Description (Added 010222 by jbenson) *********** Running fitstoms on a VLBA AIPS FITS file, 3C120UVCOP171.FITS, the u, v, w values in the ms are incorrect by 10's of kilometers. The exposure and interval times are all 0.00514984131 secs. They should be 8.39 secs. ---------------------------------------------------------------------------- Bug AOCso02305 aips2-namerica Submitted 010222 RESOLVED defect report AIPS++, version 1.5 (bui Assigned 010222 2 enclosures Opened 010222 Resolved 010222 "During deconvolution GUI and Glish crash." DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: tcornwel ************ CR_Description (Added 010222 by jbenson) *********** Running imager.clean on VLBA data, during the deconvolution stage (FTMachine : xforming) the GUI and Glish disappear. The imager process continues to run. Have tried clean with 'clark','hogbom','wfclark'. Same crash. ************ Resolution (Added 010222 by tcornwel) *********** Please describe your resolution to this problem: John, As I said, the swap space on your machine is too low. We recommend at least 512 MB to keep a reasonable safety margin. I'd recommend that you put in a helpdesk request to get this done asap. Cheers, Tim ---------------------------------------------------------------------------- Bug AOCso02302 problems Submitted 010222 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010222 1 enclosure "Applies incorrect amplitude corrections to visibility amplitudes" Detection method: customer use Assigned engineer: akemball ************ CR_Description (Added 010222 by jbenson) *********** After running fitstoms to create a new measurement set, on the first construction of an imager tool, a new CORRECTED_DATA column is loaded. The visibility amplitudes in CORRECTED_DATA are different from the amplitudes in the DATA column by several percent. The phases are identical. This is a VLBA dataset that has been calibrated in AIPS. ---------------------------------------------------------------------------- Bug AOCso02303 problems Submitted 010222 ASSIGNED enhancement AIPS++, version 1.5 (bui Assigned 010222 1 enclosure "Antenna indices are zero based." Detection method: customer use Assigned engineer: akemball ************ CR_Description (Added 010222 by jbenson) *********** The antenna indices displayed in the tablebrowser viewing a measurement set are zero based. This is confusing as antenna indices are 1 based every where else (I believe). ---------------------------------------------------------------------------- Bug AOCso02316 problems Submitted 010225 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010226 1 enclosure "Fails to clean properly through multiple masks" Detection method: customer use Assigned engineer: gmoellen ************ CR_Description (Added 010225 by jbenson) *********** If you set mask = 'maskfile', clean searches for clean components only in the mask = transparent region, as it should. If multiple masks are specified, mask = 'maskefile1 maskefile2 maskefile3', clean reverts to cleaning the inner quarter of the image. This is with algorithm = 'clark'. ************ TO KUMAR: (Added 010227 by gmoellen) *********** As we discussed, this has to do with DOimager::clean forcing the number of masks to be the same as the number of specified models, or else it uses the default mask, which is the inner quarter. ---------------------------------------------------------------------------- Bug AOCso02317 problems Submitted 010225 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010226 1 enclosure "Maximum residuals suddenly increase during deconvolution" Detection method: customer use Assigned engineer: gmoellen ************ CR_Description (Added 010225 by jbenson) *********** Running Clean with a single valid clean box or polygon, the maximum residuals reported suddenly increase. This generally happens when the function has cleaned down to several hundred microJy. Initial maximum residual: 2.30739 Iteration: 4, Maximum residual=1.64166 Iteration: 15, Maximum residual=0.838588 Iteration: 34, Maximum residual=0.345144 Iteration: 64, Maximum residual=0.130745 Iteration: 108, Maximum residual=0.0460069 Iteration: 177, Maximum residual=0.0159631 Iteration: 307, Maximum residual=0.00614339 Iteration: 592, Maximum residual=0.00296393 Iteration: 1161, Maximum residual=0.0015216 Iteration: 2259, Maximum residual=0.00109376 Iteration: 3568, Maximum residual=0.000629991 Iteration: 13568, Maximum residual=12816.3 Iteration: 13633, Maximum residual=3812.86 Iteration: 13795, Maximum residual=1071.07 Iteration: 14470, Maximum residual=571.372 Iteration: 15000, Maximum residual=367.589 Clean used 15000 iterations to get to a max residual of 367.589 Successfully deconvolved image ---------------------------------------------------------------------------- Bug AOCso02365 aips2-namerica Submitted 010308 NEW defect report AIPS++, version 1.5 (bui 1 enclosure "selectcalibrator(calcode="*" ) prevents calcode=" " sources from filling" ************ CR_Description (Added 010308 by jbenson) *********** Running vlafiller.selectcalibrator(calcode="*" ) prevents sources having calcode=" " from filling. Run this script get these results : include 'vlafiller.g'; myvlafiller:=vlafiller(host="pecos" , forcenewserver=F); ok:=myvlafiller.tapeinput(device="/dev/nst0" , files=1); ok:=myvlafiller.selectband(bandname="X" ); ok:=myvlafiller.selectproject(project="AB973" ); ok:=myvlafiller.selectcalibrator(calcode="*" ); ok:=myvlafiller.output(msname="ab973a.ms" , overwrite=F); ok:=myvlafiller.fill(verbose=F, async=F); ok:=myvlafiller.done(); > include 'ab973_vlafill_a.g'; Starting server vlafiller on host pecos Server started: /home/cluster/aips++/weekly/linux_gnu/bin/vlafiller (AIPS++ version: 1.5 (build #242)) Positioning the tape to the start of file 1. Will read file 1 from /dev/nst0 (tape) Selecting any data in the 2800 MHz band around 8200 MHz Selecting data from project AB973 Selecting all calibrators Writing to measurement set ab973a.ms Copying the data. Be patient, this may take some time. The measurement set contains 155142 rows. The antenna sub-table contains 26 entries The field sub-table contains 6 entries <************* 6 cal sources ******** The polarization sub-table contains 1 entries The spectral window sub-table contains 2 entries Successfully closed empty server: vlafiller The calibrator sources are filled, the program sources are not. Now eliminate the selectcalibrator function. include 'vlafiller.g'; myvlafiller:=vlafiller(host="pecos" , forcenewserver=F); ok:=myvlafiller.tapeinput(device="/dev/nst0" , files=1); ok:=myvlafiller.selectband(bandname="X" ); ok:=myvlafiller.selectproject(project="AB973" ); ok:=myvlafiller.output(msname="ab973b.ms" , overwrite=F); ok:=myvlafiller.fill(verbose=F, async=F); ok:=myvlafiller.done(); # > include 'ab973_vlafill_b.g'; Starting server vlafiller on host pecos Server started: /home/cluster/aips++/weekly/linux_gnu/bin/vlafiller (AIPS++ version: 1.5 (build #242)) Positioning the tape to the start of file 1. Will read file 1 from /dev/nst0 (tape) Selecting any data in the 2800 MHz band around 8200 MHz Selecting data from project AB973 Writing to measurement set ab973b.ms Copying the data. Be patient, this may take some time. The measurement set contains 667602 rows. The antenna sub-table contains 26 entries The field sub-table contains 12 entries <************* 6 cal + 6 pgm srcs The polarization sub-table contains 1 entries The spectral window sub-table contains 2 entries Successfully closed empty server: vlafiller ---------------------------------------------------------------------------- Bug AOCso02376 problems Submitted 010314 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010317 1 enclosure "Using a ref_ant in solve, ref_ant = dummy in gcal table" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: gmoellen ************ CR_Description (Added 010314 by jbenson) *********** I can apparently successfully run calibrater.solve with refant set to a legal antenna index because the phase of the solutions for refant are 0.0 degrees. However, the 'REF_ANT' column in the gcal table seen in the tablebrowser says REF_ANT = 'dummy'. ---------------------------------------------------------------------------- Bug AOCso02394 aips2-namerica Submitted 010317 NEW defect report AIPS++, version 1.5 (bui 1 enclosure "Argument t must be t=value, not just value." ---------------------------------------------------------------------------- Bug AOCso02390 problems Submitted 010316 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010317 1 enclosure "Solution row times in gain tbl are incorrect" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer:akemball ************ CR_Description (Added 010316 by jbenson) *********** Running calibrater.solve with the solution restricted to only the calibrator sources in program source - calibrator source switching, I find that the solution time reported in the gcal table viewed in plotcal is at the center of the solution interval. It should be some- thing like the weighted mean time in the data. The reported solution times are actually outside the time range of the calibrator source scans. ---------------------------------------------------------------------------- Bug AOCso02444 aips2-namerica Submitted 010330 NEW defect report AIPS++, version 1.5 (bui 1 enclosure "Applies uvrange restriction to all data in ms." ************ CR_Description (Added 010330 by jbenson) *********** Running imager.uvrange after imager.setdata(fieldid=3), imager.uvrange(uvmin=150000.0, uvmax=2000000.0) The uvrange function goes through the entire ms and 'flags' rows for all field_id's. Too slow. I think there should be a uvrange argument in imager.setdata ---------------------------------------------------------------------------- Bug AOCso02413 problems Submitted 010321 RESOLVED defect report AIPS++, version 1.5 (bui Assigned 010322 3 enclosures Opened 010420 Resolved 010420 "Background is grey instead of white like the other GUIs" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: gvandiep ************ Resolution (Added 010420 by gvandiep) *********** Please describe your resolution to this problem: It is now possible to specify the background color in the aipsrc variable logger.background. It defaults to xing which means use the default background (which is usually lightgrey). ---------------------------------------------------------------------------- Bug AOCso02445 problems Submitted 010402 RESOLVED defect report AIPS++, version 1.5 (bui Assigned 010404 3 enclosures Opened 010418 Resolved 010418 "Failed; Could not find polar. type for a valid antenna" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: mwiering ************ Resolution (Added 010418 by gmoellen) *********** MSFitsInput::fillAntennaTable() fills the entire UVFITS AN extension table, and MSFitsInput::fillFeedTable() was filling rows for only those antennas that are actually present in the data, which is always a smaller number for VLA data since one antenna is always "VLA:OUT" and now "VLA:VPT" (Pie Town) is included. MSFitsInput::fillFeedtable() now writes a row (all the same, since UVFITS can't be more complicated than this) for each row in the (MS's) ANTENNA table. This part of the code will be revisited soon in another defect that I am working on: adding sensitivity to multiple AN tables (for multiple subarrays). ---------------------------------------------------------------------------- Bug AOCso02451 problems Submitted 010403 RESOLVED enhancement AIPS++, version 1.5 (bui Assigned 010404 2 enclosures Opened 010404 Resolved 010404 "A new statement to gracefully exit a Glish script." DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: drs ************ Resolution (Added 010404 by drs) *********** Hi John, Thanks for the input, but I don't think such a function will be added. The script files are just loaded and then executed without having any mapping between executing a particular piece of code and which file it came from. The best way to handle this is to return a value (see http://tarzan/docs/reference/Glish/node71.html#fail-stmt-2). If functions are made to be sensitive to these values, control should return all the way back to the CLI. Darrell ---------------------------------------------------------------------------- Bug AOCso02453 problems Submitted 010403 POSTPONED defect report msplot, version 1.5 (bui Assigned 010404 3 enclosures Postponed 010406 "Will not plot weights or flag_row" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: rmarson ****************************** batchbug 010405 140723 Appended via email by "rmarson@lucy.aoc.NRAO.EDU" Hi John, I have investigated this defect and there are two distinct problems here. The first one occurs in in msplot and happens because the weights are one dimensional. msplot has code to plot scalar quantities like antenna number, field_id or u or two dimensional quantities like data[npol, nchan] but not one dimensional quantities like the weights. Fixing this is relatively simple. However this would not solve the problem! There is also a defect (#2457) in the ms tool that is used inside the msplot tool. This defect means that the weights corresponding to all polarizations are provided, even when the user has only selected some polarizations. Even though all polarizations are returned there is no way to work out which ones to plot as the weights may have been permuted. In the short term I have disabled the plotting from the msplot tool. It is unlikely that this defect will be fixed in time for the release and if you strongly feel that the msplot tool distributed with the upcoming release should be able to plot weights please tell me. The alternative is to plot all weights and issue a warning that polarizations selections are not available when plotting weights. Ralph ************ I'll remove FLAG_ROW (Added 010406 by rmarson) *********** Hi John, I'm going to remove the plotting of FLAG_ROW. Plotting this quantity doesn't make sense. Cheers Ralph ---------------------------------------------------------------------------- Bug AOCso02458 problems Submitted 010404 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010404 1 enclosure "TIME(UTC) column contains incorrect times" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: gmoellen ************ CR_Description (Added 010404 by jbenson) *********** Running tablebrowser on an 'gcal' table, the TIME(UTC) column contains the same value everywhere : -5884202/03/16/12:00:00 ---------------------------------------------------------------------------- Bug AOCso02489 problems Submitted 010410 RESOLVED defect report msplot, version 1.5 (bui Assigned 010411 2 enclosures Opened 010411 Resolved 010411 "Will not write the 'new' flags into the MS" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: rmarson ************ Resolution (Added 010411 by rmarson) *********** Hi John, I fixed this problem earlier today. The new version is avalaible in daily, or after the weekend it will be in weekly. ---------------------------------------------------------------------------- Bug AOCso02497 problems Submitted 010411 RESOLVED defect report AIPS++, version 1.5 (bui Assigned 010411 2 enclosures Opened 010411 Resolved 010411 "Autoregister button doesn't work" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: mmarquar ************ Resolution (Added 010411 by mmarquar) *********** This problem has been resolved earlier this week. Should be availabel in the next weekly build. All displaypanel invoked from the file menu now have autoregister as their default. Malte. ---------------------------------------------------------------------------- Bug AOCso02513 problems Submitted 010412 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010412 1 enclosure "Long spasms of refreshing the plot" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: gmoellen ************ CR_Description (Added 010412 by jbenson) *********** Antenna chooser goes into a long series of plot refreshing. It actually makes it difficult to use. ---------------------------------------------------------------------------- Bug AOCso02556 problems Submitted 010417 ASSIGNED defect report AIPS++, version 1.5 (bui Assigned 010417 1 enclosure "image is shifted the wrong way for new phasecenter" DETECTION INFORMATION LABORATORY INFORMATION Detection method: customer use Assigned engineer: kgolap ************ CR_Description (Added 010417 by jbenson) *********** I think the they sign of the image plane shift caused by setting phasecenter is wrong. For example, I imager.clean a point source and it peaks at pixel 158 on the y-axis (nx,ny=256). Using the viewer statistics I find the ra, dec of the peak in the image and enter that in the setimage phasecenter argument, like so : phasecenter=dm.direction('J2000','14:13:14.836','-03.12.23.291'), doshift=T, ...... The peak in the resulting image goes to y-pixel 187. I see the same problem on the x-axis as well. ---------------------------------------------------------------------------- Bug AOCso02561 aips2-namerica Submitted 010419 NEW enhancement AIPS++, version 1.5 (bui 1 enclosure "Constrain mean gain amplitude to unity" ---------------------------------------------------------------------------- Bug AOCso02563 aips2-namerica Submitted 010419 NEW enhancement AIPS++, version 1.5 (bui 1 enclosure "Solution interval boundries on even number from 0000 UTC" ---------------------------------------------------------------------------- Bug AOCso02565 aips2-namerica Submitted 010419 NEW defect report AIPS++, version 1.5 (bui 1 enclosure "Calibrater server fails when setapply field differs from setdata" ---------------------------------------------------------------------------- Bug AOCso02567 aips2-namerica Submitted 010419 NEW enhancement AIPS++, version 1.5 (bui 1 enclosure "Please install fieldid and spwid as function arguments" ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- ---------------------------------------------------------------------------- ----------------------------------------------------------------------------