;----------------------------------------------------------------------- ; Copyright (C) 2000-2001 ; Associated Universities, Inc. Washington DC, USA. ; ; This program is free software; you can redistribute it and/or ; modify it under the terms of the GNU General Public License as ; published by the Free Software Foundation; either version 2 of ; the License, or (at your option) any later version. ; ; This program is distributed in the hope that it will be useful, ; but WITHOUT ANY WARRANTY; without even the implied warranty of ; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ; GNU General Public License for more details. ; ; You should have received a copy of the GNU General Public ; License along with this program; if not, write to the Free ; Software Foundation, Inc., 675 Massachusetts Ave, Cambridge, ; MA 02139, USA. ; ; Correspondence concerning AIPS should be addressed as follows: ; Internet email: aipsmail@nrao.edu. ; Postal address: AIPS Project Office ; National Radio Astronomy Observatory ; 520 Edgemont Road ; Charlottesville, VA 22903-2475 USA ----------------------------------------------------------------------- This file must be used to record any changes made to the Master version of AIPS kept in Charlottesville. An entry is: line 1: Date Program name Your name line 2: What was done (may span multiple lines) line n-1: Where has this change been moved (e.g. 15JUL94, nowhere) line n: Blank The details of line 1 are important since programs are used to find specified strings and do conversions on them. It MUST be typed as follows: 1 blank, 4 digits, a period, two blanks, then the Month, day, comma, year, some space, a title, some more space, your name. *********************************************************************** *********************************************************************** Changes while 31DEC01 is TST *********************************************************************** *********************************************************************** 10713. December 12, 2000 Update Eric Update: CHANGE.DOC, GETRLS, SETPAR, INSTEP1, MAKE.MNJ, update.pl, install.pl, AIPSPATH.SH and AIPSPATH.CSH to change 31DEC00 to 31DEC01. NEW INSTEP1 updated from 31DEC99 to 31DEC00. Moved NOWHERE - important not to go to NEW. 10714. December 12, 2000 PAPC.INC Eric Changed the shipped copy to include the larger sizes of AP but the one that is not commented away is still 5 Mbyte. Moved to NEW (31DEC00) this date. 10715. December 12, 2000 HGEOM Eric HGEOM has had an error reading the input whenever it was using only a subset of the rows and was doing more than one plane. It would read 1 too may rows in each plane when the rows were long enough - for some reason it avoided this error on smaller rows. It worked on 512x512 and failed on 1024x1024. GEOM, LGEOM and PGEOM also had this error. Changed all of them to put out 1/10'th as many progress messages. Moved to NEW (31DEC00) this date. 10716. December 12, 2000 COMB Eric Fixed typo in help file. Moved to NEW (31DEC00) this date. 10717. December 12, 2000 VPLOT Eric It reverses the sign of the VLA Y antenna position twice. Models work much better with only once. Moved to NEW (31DEC00) this date. 10718. December 12, 2000 FLATN Eric Changed to do multiple fields of multiple pointings. If there is > 1 pointing, the output image is primary beam corrected. Changed PBCALC to use the LAMBDA value as a clue to init or re-init the parameters. Moved to NEW (31DEC00) this date. 10719. December 12, 2000 VPLOT Leonia/Eric There was an error addressing the X coordinate when plotting a model at the data points in the presence of time averaging. This caused a bad and only partial model plot. Moved to NEW (31DEC00) this date. 10720. December 12, 2000 INSTEP3 Eric Added IMAGR and FITAN to support Y2K as well as DDT. Moved to NEW (31DEC00) this date. 10721. December 12, 2000 PBPARM Eric The new FLATN has a new adverb (to go to PBCOR, FACES, SETFC at least in time) call PBPARM for primary beam parameters. Changed POPSDAT.HLP, NEWPARMS.001, and created a help file. Moved to NEW (31DEC00) this date. 10722. December 13, 2000 INCS.SH Eric $SYSSGI was missing a suitable INCS.SH. Moved to NEW (31DEC00) this date. 10723. December 21, 2000 SETJY Jim Added the 1999.2 flux scale supplied by Taylor and Perley, and changed that to the default. Added a warning about 43-GHz fluxes to the help file. Moved to NEW this date. 10724. December 21, 2000 Message levels Eric Repetitive informative messages should be at a fairly low level, e.g. 2 or 3, so that they may be suppressed. ERror messages must be > 5. Changed $QNOT routines for the former: ALGSTB ALGSUB GR2DUV GRDMEM GRIDUV MAKMAP OLDMAP SDGRID UVGRID UVGRTB And the latter: APLOAD APWRIT UVTBUN UVTBUW VISDFT VSCDFT VSFDFT Moved nowhere - not important enough for NEW. 10725. December 24, 2000 $APGOOP/SPECR Eric/Leonia SPECR has not worked with compressed data. It has been fixed up. Moved to NEW (31DEC00) this date. 10726. December 28, 2000 IMAGR Eric/Bill Following suggestions made by Bill, added code to decide to skip a field in OVERLAP 2 mode when, after re-imaging, its objective function (peak + some part of average) is < the objective function of other field(s). This will remake a bunch of fields until a good one is found to Clean. Moved nowhere - experimental algorithm. 10727. December 28, 2000 $YPGNOT/FRPLT, POSSM Leonia FRPLT has never worked with non zero SHIFT. When FRSCL is calculated the two asteriks' was used erroniosly: one at the end of the line; another at the beginning of the next line. Such two asteriks' are interpreted as exponent. The value of the exponent power is too huge to calculate the exponent and FRPLT blew up. Another small bug has been fixed at FRPLT and POSSM. Both tasks use UVGET which takes care about the frequency reference pixel if bchan .ne. 1. But FRPLT and POSSM carry out it also. So the frequency pixel shifted twice at UVGET and at the task itself. Moved to NEW (31DEC00) this date. 10728. December 29, 2000 ZTRLOG Eric ZTRLOG has been called from both Fortran and C, the former with Fortran Character strings. We CANNOT do this - and have been lucky that all compilers we currently use put the string lengths at the end of the calling sequence. Changed $APLGEN/ZTRLOG.FOR to translate to hollerith and call ZTRLO2, renamed $APLUNIX/ZTRLOG.C to $APLUNIX/ZTRLO2.C Changed to call ztrlo2_ the following C: $APLCRI: ZTPOP2 $APLLINUX: ZMOUN2 $APLHP: ZMOUN2, ZTKOP2, ZSTRTP $APLSOL: ZMOUN2, ZTPOP2 $APLSGI: ZCREA2, ZMOUN2, ZMOUN2.CI5 $APLAXLIN: ZMOUN2 $APLCVEX: ZTPOP2 $APLSUN: ZMOUN2 $APL3090: ZCREA2 $APLULTRX: ZMOUN2, ZCREA2 $APLDEC: ZMOUN2, ZCREA2 $APLIBM: ZMOUN2, ZCREA2, ZTPOP2 $APLBERK: ZCREA2, ZSTRTP $APLUNIX: ZCREA2, ZM70O2, ZPATH, ZSTRTP, ZTPOP2, ZTKOP2, ZTTOP2 Changed $SYSUNIX/PP.FOR to include ZTRLOG.FOR, RFILL, CHR2H, H2CHR and created ZTRLOP.C containing ZTRLO2, ZPUTCH, ZGETCH. Changed $SYSPERL to handle ZTRLOP rather than ZTRLOG. Moved nowhere - let's hope 31DEC00 gets away with this. 10729. January 16, 2001 $QPGNOT/BPASS.FOR Leonia BPASS has shifted spectra at the VLBA case to remove the shift to the center of the Earth carried out by the correlator. The shift has been calculated for the first antenna of the visibility. It is wrong for the cross correlation case, because the visibility is the product of the two antenna signals being shifted differentely. At this new version the visibilities are nit shifted, but the solution for each antenna is shifted by the relevant value. The auto correlation case has not been changed. Moved nowhere 10730. January 18, 2001 $QPGNOT/FRING.FOR Leonia FRING has typed the warning: "Frequencies do not lie on a uniform grid.." when it is not true. It was because the tolerance 1E-7 of the frequncy channel space is too small and contradicts with the accuracy of the frequency set given as 'REAL'. So I have changed the tolerabce to 1E-3 and it should help. Moved nowhere 10731. January 22, 2001 PCAL Steve Myers / Eric The computation of the rms has been incorrect for a long time but the answers were reasonable until the weights became more nearly 1/sigma**2. Most weights were sqrt (number of observations) too small and some matrix elements also had to be scaled (increased) by the sum of the weights. LPCAL had typing corrected but never summed the weights for any real purpose. Moved nowhere. 10732. January 22, 2001 SCMAP, SCIMG Eric Both SCIMG and SCMAP passed the UVTAPER on as UVTAPER - this is no longer correct. It is called WTTAPER now and is applied during weighting. SCIMG also did not have parameters set properly to determine the number of channels to average (and did not put that value everywhere it is needed). Moved to patch area. 10733. January 23, 2001 QCLEAN Eric Three places did a METSCA and then applied that scaling to another variable witout checking if the first was pure zero. Fixed to METSCA on the other in that case. Moved nowhere. 10734. January 23, 2001 SCMAP Eric Failed to set NBOXES=1 when it was zero - so no cleanable signal and no Clean... Moved to patch area. 10735. January 26, 2001 FILLM Eric A typo meant that the obs mode in scans after the first was taken from the source name (chars 9-10) rather than the actual mode. This could cause Solar mode to be falsely asserted if the source name char 9 was S. It could also cause Solar mode to be missed when it should have been invoked. Moved to patch area. 10736. January 27, 2001 CATALOG Eric Changed CATLST so that it changes format when the highest slot number to be listed is > 9999. Affects CATALOG, MCAT, UCAT, PCAT verbs. Moved nowhere. 10737. January 27, 2001 SETGDS Eric SETGDS was filling NITER with the max number of CCs in the file rather than the number about to be used. Most tasks did not care (NCLNG ended with the desirec value) but VPLOT and CLPLT used the returned value. Changed CLPLT and VPLOT to handle 25000 CCs rather than 10000. PCAL put the output into the HI - should have used NSUBG-1 like all the other tasks. Moved nowhere. 10738. January 27, 2001 help files Eric The NMAPS section of a lot of help files was not updated to show the new name format, the support for starting at a class > 1, and the continued support for the old name form. Changed help files for UVCON, BLCAL, BPASS, KRING, FRING, IBLED, FLATN, PCAL, LPCAL, UVSUB, UVADC, CLPLT, VPLOT, CALIB, CPASS, GPHAS, BLING. Moved nowhere. 10739. January 29, 2001 ZTPMID Eric One of the main problems with FITS disk is detecting the end of file condition. END= is not legal in a READ where the BLKNO is specified, but the eof IOSTAT has varying values both with time and architecture. Made an $APLIBM version using END= to get around a recent change that we can no longer test or debug. Moved nowhere. 10740. January 30, 2001 Primary beams Eric Rick Perley completed the beam measurements for P and 4 Bands. Changed PBCALC.FOR to drop the old P and 4 band models and add the P and 4 band parameters provided by Rick. Changed the help files of FLATN, PBCOR, SETFC, FACES, and PATGN to account for this. Changed the help and Fortran files for PBCOR, SETFC, FACES, and PATGN to add the adverb PBPARM replacing other adverbs that functioned as beam parameters. Changed PBCOR to replace GPOS with COORDINA as well. Moved nowhere. 10741. January 30, 2001 ZDCHIN Eric Changed the initial value of CRTMAX to -1 so that, when first set up, installations will respond to the size of the CRT window. Moved nowhere. 10742. January 31, 2001 PBEAM.FOR Leonia Some final changes have been put to the task codes. They include change of the formats to work with very wide beams at low frequencies (327, 73MHz); exclude vector averaging of the "identical" points separated by time; final hair dressing of the codes. Moved nowhere. 10743. February 1, 2001 UVCON.FOR Leonia UVCON has not worked with CMODEL='COMP' since last summer, when the new version of the subroutine QXXPTS was installed. UVCON using the private copy of the old QXXPTS has a problem in an array address. I have modified the private version of QXXPTS using its last version at the system. Additionally the maximum number of antennas has been increased to 2000 to allow UVCON simulating SKA. Of cource antenna selecting does not work with the simulated data, because 2000 exceeds the upper limit 50/90. Some formats are updated to the new number of antennas. Moved nowhere. 10744. February 1, 2001 Tools Eric Changed PRTAC to allow stand-alone usage to summarize all of the Hosts in the local SITE in HOSTS.LIST. It also can initialize the AC files leaving all entries after a requested time. Added stand-alone usage description to the help file as an Explain section. New program DRCHK to check the agreement between HOSTS.LIST, NETSP, and the DADEVS.LIST file(s). It is only a stand-alone program but I wrote a help file anyway. Moved nowhere. 10745. February 5, 2001 $QPGNOT/BPASS.FOR Leonia The recentely modified BPASS (see #10729) has not worked with SOLINT=-1, because the source ID was not determined at this case. I have fixed it. Moved nowhere. 10746. February 7, 2001 DRCHK Eric Fixed a formatting bug - WRITE cares about trailing blanks fitting in the string. Moved nowhere. 10747. February 8, 2001 WTMOD Eric WTMOD could not handle packed data even though it could modify weights without ever unpacking and repacking the data. Fixed that and added clarifications to the help file. This task may be needed by people who have nominal sensitivity scaled weights to be combined with old simple weights. Moved nowhere. 10748. February 8, 2001 $YPGNOT/CONFI.FOR Leonia Made several modifications: 1. Increase number of antennas to 2000, in order to make possible using CONFI for SKA optimization. The relevant formats have been changed. 2. If the number of antennas exceeds 100, only cofiguration is plotted now without uv coverage, because the plot file is becoming huge for large number of antennas. 3. The output file can be prepared now in two formats: to continue of optimization of sidelobes by CONFI (the exsisted version) and to use it as an input file for simmulation task UVCON. Moved nowhere. 10749. February 12, 2001 ZLWIO Eric The expanded file name local variables were not long enough. Moved nowhere. 10750. February 13, 2001 OMFIT, HF2SV, CL2HF Eric These also called ZFULLN with short name strings. HF2SV needed a lot of cleaning up of its handling of the string. Moved nowhere. 10751. February 13, 2001 UVFLG Eric Added trap for APARM(4) > APARM(5). Users were flagging all the data when the elevation range was not specified properly. Moved nowhere. 10752. February 13, 2001 VBGLU Eric CHNDAT was called with FREQID unset - this could only have worked when there was no FQ table. Set FREQID=1 and hope that there is only 1 or that all have the same IF frequency order. Moved nowhere. 10753. February 14, 2001 HF2SV Eric Linux compilers got confused whan a character variable was passed in, correctly declared (CHARACTER FNAME*(*)), and then used in an expression which called for an intermediate argument (e.g. WRITE(MSGTXT,2000) FNAME(1:NCH) // 'XXX'). The compilers go on at length about array arguments when FNAME(1:NCH) is a scalar. Made local variables of known length to avoid this egcs compiler error. Solaris, HP, SGI all accepted this code. Moved nowhere. 10754. February 14, 2001 FITLD Eric Changed it to notice that the correlator (AN file array) name is not VLBA and to turn off all VLBA options (divide RE/Im by weight, DIGICOR, DELCORR, artifact correction) although some were not done in this case previously. Fixed history and messages to honor this. Moved nowhere. 10755. February 20, 2001 $APGNOT/FACES.FOR Leonia NPARMS should be 168 instead of 171 at the fortran code. This difference made a problem at some machines. Moved nowhere. 10756. February 23, 2001 $YPGNOT/IRING.FOR, HLP Leonia Fixed the problem with very small radii given at argsec. The help file has improved including CPARM(6). Moved nowhere. 10757. February 23, 2001 VLBAUTIL.001, VLBACALA.HLP Jim Changed VLBACALA in the VLBAUTIL procedure to do INTERPOL='SELF' on the a-priori amplitude calibration. This prevents interpolation between different sources at different elevation angles, which can cause significant errors in antenna gains. Also added a sentence mentioning this in the VLBACALA help file. Moved nowhere. 10758. February 27, 2001 MATCH Eric Added DOALL option to control which of the AN, SU, and FQ parameters are matched and which left alone. For use with VBGLU, the frequency parameters must be left alone. Moved nowhere. 10759. February 28, 2001 Plots at the pole Eric Changed: IAXIS1 to move the INSIDE strings in only a little TICINC to look over the full subimage for a range of coordinates and to use a better list for RAs. ITICS to use a simpler/better algorithm for the full axis drawing. CTICS as ITICS TKTICS as ITICS Moved nowhere. 10760. March 1, 2001 BPASS Eric The code to shift VLBA data after solving for the bandpass rather than before forgot about the possibility that one or more channels would be blanked (~ 10^^5). Changed the interpolation routine to allow data arrays with NP polarizations rather than a hard-coded 2. Then added a call to do the interpolation before the shift. Moved nowhere. 10761. March 2, 2001 DISKU Eric One of 2 arrays that should be dimensioned by number of disks only went to 9. Frazer (needless to say) went over this. Changed both to be parameterized at 35. Moved nowhere. 10762. March 5, 2001 CROSSPOL Amy Added defaults so that CROSSPOL would not die because of "hidden" inputs. Changed procedure so BASELINE 0 would work. Added checks to force a valid REFANT and GAINUSE to be entered. Added a CLCAL at the end to apply copied SN table to CL table. Corrected help file. Moved nowhere. 10763. March 6, 2001 $YPGNOT/CONFI.FOR; ..HLP Leonia The two options have been added: The hexagon tiling type of the initial configuration is added. The task gives a warning now if the spacing between antennas less than the given limit. The help file updated. Moved nowhere. 10764. March 6, 2001 DEFER Eric Added adverb to POPSDAT.HLP and NEWPARMS.001. Also new help file. Moved nowhere. 10765. March 6, 2001 MDESTR Eric Fixed it to look for a lot of files to destroy and destroy any it finds when the header is missing or empty. Moved nowhere. 10766. March 7, 2001 UVFLG Amy Changed subroutines FLAGEL and FLGPCL so that an input REASON will override UVFLAG's REASON. Edited help file stating that if REASON is left as the default in these two specific cases UVFLG inserts it's own REASON. Moved nowhere. 10767. March 8, 2001 SORT Eric There seemed to be some confusion of character length for the 'SORT' keyword in QUVUTIL and QCLEAN. Changed all to be *2. Moved nowhere. 10768. March 8, 2001 LISTR, UVSRT Eric Changed LISTR to display Tant as well as Tsys and changed the scaling of GAIN displays to show another digit by default and to apply FACTOR incl changes of a factor of 10 - formerly forbidden!) to both kinds of GAIN display. Added DEFER option to UVSRT to defer creating the output until the sort has been done and just after one of the scratch files has been deleted. Moved nowhere. 10769. March 9, 2001 UVFLG Amy Removed bug in IF statements I added 2 days ago. Moved nowhere. 10770. March 20, 2001 ACCOR Leonia Add the new option to SOLINT. Now if SOLINT<0 then the data are averaged over the scans. The help file has been updated. Moved nowhere. 10771. March 20, 2001 SETFC, FACES Eric Fixed the flux scaling in FACES - it was not corrected for the change back to mJy in the tables. Also moved the flux rescaling to before the flux cutoff so that the cutoff is in Jy at the actual frequency. Changed help file to reflect this and to drop the warning about large angles. Changed SETFC help file to drop that old warning and to fix the beam parameter section that was missing stuff. Moved nowhere. 10772. March 20, 2001 MDESTR Eric The call argument to MSGWRT was missing when the file was too busy to delete. Moved nowhere. 10773. March 22, 2001 FACES, SETFC Eric The number of fields allowed by the 512 limit combined with CELLSIZE and IMSIZE may not cover the requested central area. Added warnings to both tasks. Also change FACES to try up to 20 times 512 fields for sources so long as no more than 512 get written with sources. Moved nowhere. 10774. March 23, 2001 SPLAT Leonia SPLAT has not worked at time averaging mode if both cross and auto is used, because the total number of baseline was calculated specifically for only cross correlation. I fixed it calculating the total number of baselines in dependance on APARM(5). Moved nowhere. 10775. March 23, 2001 CROSSPOL Amy Fixed up the tget/vput and vget/tput statements so when CROSSPOL is run the previous tputs for the tasks in CROSSPOL are saved. Unfortunately this will not work if the procedure crashes. Moved nowhere. 10776. March 26, 2001 FACES, SETFC Eric The FLYEYE routine did not work well in the extreme cases. Fixed both to attempt 20 times (or more) MAXFLD fields and to try more Y positions than X since DY is about 1/2 DX. Suppressed coordinate error messages and changed the handling of failed coordinate computations to skip them entirely. All flyeye fields must fit (easily) into a -SIN projection. Moved to AOC this date. 10777. March 26, 2001 BLING Eric BLING forgot to move BADDISK from the Inputs object to the DFIL.INC common for use. Moved nowhere. 10778. March 27, 2001 UVPLT Eric The QUAL adverb was not put in the SELQUA Common variable for use by the calibration package and so was ignored. Moved nowhere. 10779. March 27, 2001 SETFC, FACES Eric The routine added for FLYEYE did not guarantee that (0,0) was at the center of a field. Added code to insure this. Moved nowhere. 10780. March 28, 2001 KRING.HLP Amy Changed reference to BPARM(3) in help file to CPARM(8), since BPARM do not exist. Moved nowhere. 10781 March 28, 2001 VLBAUTIL.001,VLBAUTIL.HLP Amy Added procedures VLBAPCOR, VLBAFRNG, VLBAKRNG, VLBAFRGP and VLBAKRGP to VLBAUTIL. VLBAPCOR runs PCOR, CLCAL and sometimes SNCOR and FRING. VLBAFRNG and VLBAFRGP run FRING and CLAL. VLBAKRNG and VLBAKRGP run KRING and CLAL. Moved nowhere. 10782 March 28, 2001 VLBAPCOR.HLP Amy Added new procedure help file. Moved nowhere. 10783 March 28, 2001 VLBAFRNG.HLP Amy Added new procedure help file. Moved nowhere. 10784 March 28, 2001 VLBAKRNG.HLP Amy Added new procedure help file. Moved nowhere. 10785 March 28, 2001 VLBAFRGP.HLP Amy Added new procedure help file. Moved nowhere. 10786 March 28, 2001 VLBAKRGP.HLP Amy Added new procedure help file. Moved nowhere. 10787 April 2, 2001 VLBAFRGP.HLP and VLBAKRGP.HLP Amy Fixed reference to CALSOUR in help files. Moved nowhere. 10788 April 2, 2001 VLBAFRGP, VLBAKRGP, VLBAKRNG, VLBAFRNG Amy Added IF statement so if GAINUSE is equal to the default the highest CL table is used as GAINUSE in KRING/FRING and as GAINVER in CLCAL. Also changed VLBAKRGP/FRGP to list what sources were being referenced to what calibrator. Moved nowhere. 10789 April 2, 2001 VLBACALA Amy Added OUTSEQ=INSEQ to the two TACOP runs in VLBACALA. Moved nowhere. 10790 April 4, 2001 FILLM Leonia/Eric Claire Chandler pointed out the data loaded by FILLM which (CH0) made any AIPS task blow up! I have found the bug in FILLM which cuased the corruption of the last visibilities of a scan if the next scan data has different frequencies and as a result requires creating a new files (stream). Having finished the buffer to record the previos scan (by UVDISK), UVINIT store the last portion of the data (the UVDISK actually records the integer number of blocks) at the beginnig of the buffer. The information of the following scan is added to this beginning. But the subroutine recording history information used the same buffer for its operation and as a result corrupted the beginning of the buffer that represents the end of the previous scan. I have changed the buffer used at the history subroutine to the small integer buffer of 256 elements. Additionally I changed using the main large buffer for the small buffer in input of many other routines: CATIO, UVCREA, CATDIR, DIE, GTPARM, RELPOP, NXTMAP, MAPCLS, CATKEY. The question is why have not people seen the problem before? The answer is: First, The problem appears when the next scan has different frequencies or something yet to cause the creating a new stream(file) of the data. It is not usual case. Second, the corruption of the last visibilities can not be so dramatic as in Clair's case. Instead it can cause the change of data or skipping the corrupted visibilities. So the people can not pay attention on this. Moved nowhere. 10791. April 4, 2001 UVCON Leonia UVCON has not allowed using images with a digit at the fourth character of its CLASS name. It was a correct precaution for the old representation of the images classes. Now it is wrong. So I have removed this constrain. Moved nowhere. 10792. April 9, 2001 PBCOR Eric PBCALC reset the input array to the beam model, but did not reset the units properly (the 10**3, 10**7 etc). When the frequency changed the user array was picked up again and the units were scaled down by another 10**3, 10**7 etc. The affect was to fix only every 16'th or 14'th or ?? plane. Moved correction the 31DEC00 patch area. 10793. April 9,2001 IMAGR Eric IMAGR was having trouble deciding that it is done. Changed the routine that evaluates the "objective function" to refuse to examine any field below MINFLX and added a check on the reuturn for no field found => FINISH = true. Also added IMAGRPRM(20) to allow the user some control over how far apart the "current" and "second-best" must be to do a retry. The default is a factor of 1.005; it was 1.0. Changed IMAGR.HLP and fixed a typo in it as well. Moved nowhere. 10794. April 9, 2001 CCOPTS.SH Eric Changed the $SYSLINUX version of CCOPTS.SH to have it specify 64-bit file I/O for site COAOARN (NRAOAOC backwards). This seems to be all that is required in the new Linux kernel to do large files. The option is -D_FILE_OFFSET_BITS=64. Moved nowhere. 10795. April 9, 2001 QUVUTIL Eric When making images without making beams (ALLOKAY >= 1) the image header common used for coordinate checks was not initialized. Changed OUVIMG to set DOINIT true at the start and for beams, bit allow it to be true for the first image. Moved nowhere. 10796. April 9, 2001 SHOW/TELL Eric Changed AU1A to display NUMTELL as well as OPTELL. Changed AU2 (STQUEUE) to use NUMTELL <= 0 to mean all as well as DOALL > 0 meaning all. Changed the STQ header to show POPS numbers included. Moved nowhere. 10797. April 9, 2001 VLBAPCOR/FRNG/KRNG/FRGP/KRGP Amy Add subarrays to help files. Procedures should not need changing. Moved nowhere. 10798. April 10, 2001 FPARSE Eric For some reason this parsing routine was not updated when the others were in 2/2000! The addition of EQUINOX pointed out the failure of using table position numbers in such routine - the character values are now used everywhere. Made GETI a new $APLSUB subroutine removing it from FITLD and UVLOD. It gets the numeric part of a string skipping over non-numeric characters. PRTTP and AIPS (TPHEAD) were affected by this error. Moved nowhere. 10799. April 10, 2001 SETFC, CHKFC Eric Changed SETFC so that the size of the circular Clean box is taken into account in the separation of the fields. This means that overlaps of a few not ~15 are now appropriate. CHKFC is a new task - it makes the fields described in a BOXFILE and fills them with 1's in the Clean boxes and blanks or 0's elsewhere. Running FLATN on the output will show where you are Cleaning. Moved nowhere. 10800. April 10, 2001 VLBAUTIL Amy Add VLBACRPL, VLBASNPL and VLBACPOL. VLBACRPL and VLBASNPL are plotting procedures, basically a simplified POSSM and SNPLT respectively. VLBACPOL is CROSSPOL with an added check on TIMERANGE (no default is allowed). moved nowhere 10801. April 10, 2001 CROSSPOL Amy Added check to TIMERANGE, the default is not allowed. moved nowhere 10802. April 10, 2001 VLBACRPL.HLP, VLBASNPL.HLP & VLBACPOL.HLP Amy Added the help files for these procedures. moved nowhere 10803. April 10, 2001 SPFLG Eric Added a progress message for large gridding ops. Moved nowhere. 10804. April 10, 2001 CALIB Craig Walker/Eric Changed Fortran and help to add the option to restrict gain normalization to elevations greater than APARM(10) degrees. Moved nowhere. 10805. April 10, 2001 SPECR Eric Fixed: a. Read mode needs to be RRAW since it does no data selection and READ removes the FREQID (at the least). b. Needs to copy any table in the input EXCEPT BP tables. c. Also fixed the FFT handling when the output spectrum is smaller than the input. Moved nowhere. 10806. April 11, 2001 VLBAUTIL, VLBASNPL, VLBACRPL Amy Add DOTV to VLBASNPL and VLBACRPL. Fix VLBAFRPL help file. Moved nowhere. 10807. April 11, 2001 CALIB Eric At Craig's suggestion, move APARM(10) to CPARM(1) since CPARM(2) is the logical that turns on the mean gain modulus constraint. Moved nowhere. 10808. April 12, 2001 PRTTP, TCOPY Eric Fixed counters in both to be double precision to avoid overflows at 2 Gbyte. Added counters to display the byte size of each file in PRTTP as well. Moved nowhere. 10809. April 12, 2001 FITLD Eric The use of FREQSEL to select 1 of several FREQIDs caused problems. Sveral tables were not rearranged and edited to have the new numbers and not have the unselected data. The logic was simplified to put the FQ selection and the FQ renumbering in one place and to apply that in all cases. Moved nowhere. 10810. April 15, 2001 CHKFC Eric Added optional field numbers in the center of each Clean box and a pixel at each end of the included portion of a row at a higher values as well to mark the Clean Boxes. Moved nowhere. 10811. April 17, 2001 ALGSTB Eric ALGSTB is required to deal with all data that comes in even if they are at u,v values outside the allowed range of the gridded model (1/(2*cellsize)). It does this by altering the u,v values to something legal while zeroing the data and weights. UVLIMT was supposed to count and report these. But it stopped counting at 10 and no summary was printed. Changed UVLIMT to count even when it does not print any more, to issue a warning on the first unprinted overflow. Changed ALGSTB to print a prominant message when samples are deleted. Moved nowhere. 10812. April 17, 2001 BPASS, CPASS Eric The use of models has never worked. There is an array called FACGRD that must be initialized to 1, but was left unset at 0. All CC fluxes were multiplied by FACGRD(1) (= 0) before being used in the model. Needless to say the model failed. Added adverb CMETHOD to allow the users to select the method. Moved nowhere. 10813. April 17, 2001 PRTTP Eric Changed it to NOT print summaries of a file when the file has 1 record <= 512 bytes. The analysts did not like seeing the outputs for the label files. Moved nowhere. 10814. April 17,2001 VLBAUTIL, VLBASUMM Amy Added new procedure to VLBAUTIL called VLBASUMM. This runs PRTAN and LISTR of the user. The VLBAUTIL help file was modified. Moved nowhere. 10815. April 17, 2001 SETFC Eric Changed SETFC to have BPARM(7) => radius of Clean circle if it is > 0 with a min of 5 and a max of fldsize/2-6. The Sun subroutine did not handle its call sequence properly. Moved nowhere. 10816. April 18, 2001 VLBALOAD Amy Add NFILES as an input. Moved nowhere. 10817. April 22, 2001 VLBAUTIL, VLBASUMM Amy Remove SUNARRAY adverb from VLBASUMM and make it loop over antenna tables if there is more than one present. Moved nowhere. 10818. April 22, 2001 VLBACPOL.HLP Amy Expand explaination of DPARM in VLBACPOL help/explain file. Moved nowhere. 10819. April 23, 2001 $YPGNOT/CONFI.FOR Leonia I have improved the proximity of antennas to the 'bad' topography. The old version has the discret equaled to the cell size. Now it is half of the cell size. This change is important when the roads are interpreted as a 'bad' topography. Moved nowhere. 10820. April 24, 2001 CHKFC Eric CHKFC used WINDF from IMAGR but that expected an array IMSIZE(2,nfields) all filled in. Moved WINDF to CHKFC and made a local version that computes the iMSIZE for each field as it goes and if there is no Clean box makes one with FLDSIZ. Fixed filling in of rectangles as well. Moved nowhere. 10821. April 24, 2001 UVMOD Eric Changed this old clunker to use GMAX, GPOS, et al to set up to 4 sources in a uv data set. One at a time is too slow. Moved nowhere. 10822. April 24, 2001 DBCON Eric Changed help file to recommend that the first file be the one with all the subarrays already in it. Otherwise, adding 5 days and 0.01 subarray can make cumulative floating-point error to the level that subarrays are not correctly recognized. There is more to do - this is the finger in the dike. Moved nowhere. 10823. April 26, 2001 $QPGNOT/UVCON.FOR Leonia The task UVCON has had a nice feature puting the model center at RA=0 and the given declination independent on the actual model center. It was achieved using the private version of the routine GRDAT. Erroniously this private version exluded the relevant shift of the UV data for the used fields of the model. As a result UVCON has excluded the space shift between the model fields and has produced wrong UV data for multi field model. I have fixed the problem introducing the shift relatively the first field. Moved nowhere. 10824. April 26, 2001 Subarray number Eric The subarray number is kept as the (subarray-1)/100.0 added to the baseline number (256*A1+A2). This strains the accuracy of single-precision numbers and makes absurd the attempt in some VLB routines to use the 1/1000 part for an important code. Repeated DBCONs with the multiple subarray data set second cause 0.01's to be added repeatedly to the baseline numbers. The accumulated error made subarray 5.9 instead of 6.0 and only 0.1 was used in the roundoff. Changed that to 0.5 everywhere: $APLNOT: DATFLG DATGET DGETSD AVERT $APGNOT: UVFIX AVSPC INDXH MULTI PHSRF PRTUV SWPOL USUAB UVCOP UVFLG UVIMG VBCAL FXVLA RESEQ DBCON TI2HA MATCH $APGOOP: INDXR $QPGNOT: CPASS SDVEL BPASS FRING $YPGNOT: IBLED VPLOT SPFLG TVFLG I wonder why the subarray is not used more places. Moved nowhere. 10825. April 26, 2001 IMAGR Eric Created adverb FGAUSS to set a flux level to Clean to for each of the resolutions in IMAGR. Changed IMAGR to sort and default these andput themin each image; QCLEAN to pick them up, use them in selecting and Cleaning fields; IMAGR.HLP, WGAUSS.HLP, FGAUSS.HLP. Also changed QCLEAN to ignore fields that currently seem done when computing the histogram to select pixels for the next Clean round. Changed POPSDAT.HLP and NEWPARMS.001 as well. Moved nowhere. 10826. April 26, 2001 LISTR help Eric DPARM(4) is important to ordinary listings even if no averaging takes place. Otherwise microscopic time differences are treated as different in the display. Added info to the help about this. Moved nowhere. 10827. April 27, 2001 SNSMO help Eric The info about SUBARRAY was wrong - 0 -> all. Moved nowhere. 10828. April 30, 2001 SETFC, FACES Eric These two were not prepared for a RADIUS smaller than the maximum good area. Fixed them to allow sizes up to 16384 for recommended image size and to take the user's radius into account. Moved nowhere. 10829. April 30, 2001 CCEDT Eric Added BOXFILE and NBOXES/CLBOX as alternative input methods for specifying the boxes - these are in cells and are converted internally. Moved nowhere. 10830. May 1, 2001 VPLOT.FOR, .HLP Leonia DBCON concatanating data with different subarrays adds 5 days to time for each subsequent subarray. Until now VPLOT has deleted these 5 days returning the data to the actual observation time. Now people, in particular Claire Chandler consider this feature is not good. So I have removed the 5 days deleting making VPLOT plot the data shifted at time by 5*(SUB-1). The relevant change of the calculation of the geometry for the model has been done. Users have to pay attention on this 5 days time shift selecting TIMERANG. A message reminding users this fact has been added. The HELP file has been changed at TIMERANG desription. Moved nowhere. 10831. May 1, 2001 SLBINI Eric SLBINI did not test to see if the coordinate rotation applied to the slice or not. Thus it would go to relative labeling of a spectrum if the RA/Dec were rotated in axes 2/3. Added test for pure x or y slice into this labeling control. Moved nowhere. 10832. May 1, 2001 COMB Eric Changed APARM(5) nad (6) for OPTD so that any value incl serious negatives is allowed. Moved nowhere. 10833. May 2, 2001 CALIB Eric CALIB scales the CC file by FACGRD based on the SU table flux divided by the sum of the CC fluxes. It failed to use 1.0 when the source was not in the SU tabel and when the source flux was 0 in the table. Added warning and setting to 1.0. Made note in help file too. Moved nowhere 10834. May 2, 2001 AIPSMEM105 Eric Put back an update of the VLBAUTIL procedure Memo. Moved nowhere. 10835. May 3, 2001 $QPGNOT/PCAL.FOR Leonia PCAL has not averaged IFs if CPARM(1)=1 repeating instead the first IF. Even more, PCAL has not averaged frequencies inside of an IF. Instead it has taken the first frequency at each IF. Erroniously both the number of IFs and frequencies to be averaged was put equaled 1 through the CATBLK(KINAX+..) which was fixed to one at the beginning of the routine PCLCOP. Moved nowhere. 10836. May 4, 2001 PCAL.HLP Leonia Because now (see the previos message) PCAL averages frequencies before finding solution I have added a paragraph to the HELP file reminding that the data should be phase calibrated before PCAL. Moved nowhere. 10837. May 5, 2001 UPDCONFIG Eric Made the chkout version of UPDCONFIG specify ssh1 for MPI. This is the only way for a site mentioned in this file to get anything done. Moved to MPI to get the damn MNJ working again. 10838. May 8, 2001 FRING Eric The dynamic memory allocation and addressing of the fringe search algorithm depnded on the number of frequencies returned by the routine that checks the distribution of the frequencies. In the case of seriously irregular distribution of frequencies that number was less than the number actually fed to the search routines. Thus the addresses used were not correct. Changed to use the max of the returned number of frequencies and the other frequency counts that may be used. This might waste some memory but will never be too small. Moved nowhere. 10839. May 8, 2001 VLACALIB.HLP VLAPROCS Amy Added DOPRINT and OUTPRINT to VLACALIB. The structure has also been changed so that messages are only cleared if DOPRINT>=0. Added FLAGVER as an input. Moved nowhere. 10840. May 8, 2001 CALIB.HLP Amy Added SNVER to the description of adverbs in help file. Moved nowhere. 10841. May 9, 2001 CPASS Eric Changed the usage of BCHAN and ECHAN so that all data are read but the channels outside BCHAN-ECHAN are given 0 weight. Made the fitter use a weighted solution and gave 3 options for weights: 1, data weight, apparent 1/sigma^^2 from pre-averaging (tempered a bit by the data weight). Moved nowhere. 10842. May 9, 2001 IMAGR Eric Changed QCLEAN to allow 1-pixel loads when there is only one field. It was setting APCLCLN(LFIELD) = 0 when 1 pixel was loaded which then forced a remake of all fields in OVERLAP 2. An alternative would be to laod more pixels if this is not ok. Moved nowhere. 10843. May 11, 2001 VLBAUTIL, VLBACPOL Amy Fixed bug in VLBACPOL. In the first run of CLCAL GAINUSE was set to the number of highest CL table rather than the highest CL+1. Moved nowhere. 10844. May 11, 2001 VLAPROCS, VLACALIB Amy Added DOWAIT=1 to the beginning of VLACALIB. Moved nowhere. 10845. May 11, 2001 VLBAPCOR.HLP Amy Expanded the explanation of what to do if the PC table does not correct certain antennas. Moved nowhere. 10846. May 14, 2001 FILLM Eric Found a subtle bug - the program never changed the number of antennas even if it changed. Further, it tested the antenna ID before that had been changed. This caused the task to treat an autocorrelation data area as cross-correlation. The Solaris system did not like 2.0 **(-31040) fortunately. Changed to reset antenna number and to set the MCINIT logical in the MCANT routine when the antenna info is known. Also fixed a large number of erroneous message levels. Moved nowhere. 10847. May 14, 2001 CPASS Eric Fixed the history writing to use the new BCHAN/ECHAN values and to do the averaging time in the claimed units. Moved nowhere. 10848. May 14, 2001 FACES, SETFC Eric In FACES.HLP changed a .23 to .023 in the inputs - the help and the code had it right. In SETFC.FOR, changed it to write the full line of the Clean box for NVSS and Solar fields. changed SETFC.HLP and .FOR to insert BPARM(3) to scale FLUXes so that FLUX refers to the flux at the observing frequency rather than the table frequency. Moved nowhere. 10849. May 15, 2001 $YPGNOT/CONFI.FOR, HLP Leonia I have enhanced the task adding hexagon tile with the arbitrary number of antennas as an initial configuration. This option can be usefull for design of a compact array configuration and for design of a bank of antennas as an array element. Moved nowhere. 10850. May 18, 2001 DEFLG Eric New task: flags data based on the ratio of the vector average amplitude to the scalar average amplitude in a sliding window. Moved nowhere. 10851. May 18, 2001 VPFLG Eric New task: copies a data set converting any partially flagged IF sample to fully flagged. It also copies and alters the FG tables. This is of use for VPOL observers. Moved nowhere. 10852. May 18, 2001 CPASS Eric Added BPVER to HI file. Moved nowhere. 10853. May 18, 2001 UVFLG Leonia Using INFILE as a flagging data source, UVFLG has not paid attention to the flagging data given as input parameters such as TIMERANG, BIF, EIF..and flagging because of elevation and pulse cal. Now I have added flagging because of elevation and pulse cal working together with INFILE. The relevant warning is provided if user uses INFILE. The help file has been updated. Moved nowhere. 10854. May 18, 2001 PRTUV, UVPRT Amy Changed the limits on NPRINT, so that if OUTPRINT was and DOCRT<0 then there are no limits on NPRINT. Previously NPRINT was limited to 2000 visabilities if DOCRT<0, now that only applies if OUTPRINT=''. Moved nowhere. 10855. May 21, 2001 FILLM Eric Cretaed adverb DOACOR (new help, changed POPSDAT.HLP, NEWPARMS.001) to control whether autocorrelations are read into line data sets by FILLM. Also changed FILLM to compute its own channel zero rather than use the (suspect) on-line one. CPARM(2) = 32 instructs FILLM to use the on-line one. Moved nowhere. 10856. May 22, 2001 UVSORT Eric Fixed the calls to catio - both were WRIT when the comments said one would be UPDT. Also suppressed the messages and made UPDT be first. Moved nowhere. 10857. May 23, 2001 $YPGNOT/PBEAM.FOR Leonia PBEAM has been used the fixed line dimension (21) for the measurment array as I had been told. Recentely Rick Perley tried (unseccesfully) PBEAM for the measurments 37x37. I have increased the dimension of the measurment array till 61x61. Moved nowhere. 10858. May 24, 2001 AVSPC Eric Fixed help file: ICHANSEL was spelled witout the I in the examples which were otherwise correct. Moved nowhere. 10859. May 25, 2001 TICINC Eric My February change messed up the handling of time and RA increments in the range f 2 to 30 minutes of time. Moved nowhere. 10860. May 25, 2001 SNFLG Eric AIPSified the task provided by Mark Reid and Lincoln Greenhill to write a flag table for places in the SN tables where, on a baseline basis, the phase jumps too much between solutions. Moved nowhere. 10861. May 26, 2001 DEFLG, CLIPM Eric Changed logic on FLAGVER to have them copy the input FLAGVER to a new file and append to that. Moved nowhere. 10862. May 27, 2001 ABOUT files Eric Updated ZZ*.HLP files: ADVERB AP CALIBRAT EDITING GENERAL IMAGING PROCEDUR RUN TASK UTILITY UV VLBI Moved nowhere. 10863. May 28, 2001 VLBACPOL.HLP Eric The pre-cursor remarks are used in the CookBook and this help file had them all wrong although the rest told truth. Moved nowhere. 10864. May 29, 2001 SAD, IMFIT, JMFIT Eric Changed the error reporting so that the errors are not divided by sqrt(2) when the widths are fixed. That applies only in the large wrt the beam case which generally does not apply. Moved nowhere. 10865. May 29, 2001 PBEAM Leonia PBEAM has not worked with polinom degree more than 3 for such a long wavelength as meters. The primary beam is very wide at this case. That is why the elements of matrice prepared for least square exceeds the limit of 2**128~10**42. To get around the problem I switched the relevant variables to double precision. The least square routine with double precision inputs/outputs is written. Moved nowhere. 10866. May 29, 2001 EPOCONV Eric New verb: converts COORDINA adverb between 1950 and 2000 coordinates. Moved nowhere. 10867. May 29, 2001 DLESQR Leonia DLESQR is the double precision version of the least square routine LEASQR. Moved nowhere. 10868. May 30, 2001 PBEAM.FOR Leonia I have detached the routine DLESQR I put in the system yesterday. Some minor bugs which prevented working the task with high polinom degree and low threshold have been fixed. Moved nowhere. 10869. May 31, 2001 Bandpass cal application Eric Changed BPASET to set ISVLA (as not VLBA and not EVN) - ISVLA was actually referenced without ever being set. Fortunately, that was only in the case of bandpass shifting which is only done for the VLBA. Changed DATBND to measure the average BP amplitude applied to the data (non-VLB central 75% of channels, VLB all channels) and to apply that average to the data weights on DOCAL=2. Now gain calibration may be done with BPASS as well as CALIB. Moved nowhere. 10870. May 31, 2001 PLOTR Eric Added lots of functionality: axis functions, up to 10 point types, and corrected the line plotting so that it works. Moved nowhere. 10871. May 31, 2001 EXTLIST Eric Added code to support the new PLOTR in AU8A. Moved nowhere. 10872. June 1, 2001 ANTAB.HLP Eric Droppped the ' ' code for blanking a Tsys column - the parser would not recognize it so there was no point trying to allow it. Moved nowhere. 10873. June, 1, 2001 CookBook Eric Updated chapters: 0 Table of Contents changed 4 VLACALIB change; add DEFLG, VPFLG, SNFLG; FILLM channel-0; CPASS use 5 Mention CHKFC, FACES, DEFLG, SNFLG 9 New VLBAUTIL procs; DEFLG, 2 Gbyte limit change 13 Regular update A VLACALIB DOPRINT option B VLACALIB DOPRINT option C VLBAUTIL procs I Index changed Recipe rearranged usage - more unused now Moved nowhere. 10874. June 1, 2001 IMAGR Eric Added code to allow scaling the weights by channel. A W or w card in the BOXFILE gives the weight, channel (0 -> all), and IF (missing or 0 -> all) of the weight. Later cards prevail over earlier. Changed IMAGR.FOR and IMAGR.HLP only. Moved nowhere. 10875. June 1, 2001 VLAPROCS, VLACALIB Amy Changed DOPRINT so that VLACALIB will only print if DOPRINT>0. Previously VLACALIB would print if DOPRINT>=0. Also edited HELP file to reflect this change. Moved nowhere. 10876. June 1, 2001 VLBAUTIL, VLBALOAD Amy Added a message to at the end of VLBALOAD which warns of permanently flagged data. When FITLD is started it states this but a novice user is unlikely to notice the message as it is one among many at the start of the program. Also edited HELP file to mention the automatic flagging. Moved nowhere. 10877. June 4, 2001 FILLM Eric Changed scaling of Van Vleck correction per request of Bryan Butler. Moved nowhere. 10878. June 5, 2001 $APGNOT/ACCOR.FOR, HLP Leonia Using a fixed value for SOLINT, ACCOR can meet a small interval of averaging at the scan end because of the gap between scan. The measurements can be very noisy at the small intervals. Now the intervals less than 0.4*SOLINT are not recorded to the output SN table. Several sources can be averaged together if SOLINT.GT.0. So the source ID to record to the output table is not determined at this case. Until now the source ID of the first scan was recorded to all rows of SN table. Now the source ID as well as frequency ID equaled zero are recorded to all rows of SN table. Zeros at source ID column makes the following task CLCAL non sensitive to the adverb CALSOUR. The help file has been updated. Moved nowhere. 10879. June 6, 2001 AIPSASSN Eric Changed default for MSG_EMULATOR to xterm from none for AOC and VLA. Fixed $ONLINE for AOC. The change to START_AIPS means that remotely started servers no longer know what the user wants and I like the message terminal. Moved nowhere. 10880. June 6, 2001 FC Eric Changed FC so that it searches first $SYSLOCAL, then $AIPS_VERSION/$ARCH/SYSTEM, and finally $SYSUNIX for the FDEFAULT.SH. Made ALPHA, SOL, SUL, HP, HP2, SGI, LINUX, and AXLINUX versions of FDEFAULT.SH. Stripped these systems out of the $SYSUNIX version of FDEFAULT.SH leaving 9 others. Moved nowhere. 10881. June 6, 2001 XAS Eric The change to START_AIPS means that XAS no longer will have a "stderr" to which to write messages starting sometime after the beginning. With rsh it did not abort even if it could not write to its former stderr. With this damn ssh, it aborts leaving shared memory and TVSERV processes clogging the system. Moved the message about being told to exit to after the clean up operations so that the users will not be too fucked. Moved nowhere. 10882. June 6, 2001 VLAPROCS (VLACALIB) Amy Changed the handling of DOWAIT so that the users input DOWAIT is saved rather than changed to 1, as in last version, or -1, in previous versions. Moved nowhere. 10883. June 7, 2001 CLCOR Leonia Correction of the AN or/and the SU table, if the relevant correction of the antenna or/and the source is carried out (OPCODE=ANTP) has been added. The history file has not included information about the source position correction. It has been fixed together with the important information about AN or/and the SU table correction. The help file has been updated. Moved nowhere. 10884. June 10, 2001 ANTINI,PRTAN,UVFIX Leonia The subroutine ANTINI has had the misprint reading POLARXY. POLARX has been read as POLARY, and POLARY=0. The misprint has been fixed. The two task: PRTAN and UVFIX which definitely use POLARXY have been recompiled with the corrected ANTINI. Moved nowhere. 10885. June 10, 2001 TVFLG Eric Found a typo in the clip by form section - it used I inside the loop as if it were the loop variable instead of LSTOKS. Moved nowhere. 10886. June 11, 2001 PRTAN Leonia POLARXY have been erroniosly listed in meters. They should be in arcsec. I have fixed it. Moved nowhere. 10887. June 11, 2001 UVFIX Leonia UVFIX has erroneously considered that POLARXY are given in meters. So POLARXY have been recalculated to arcsec for using in the routine JPRECS. This recalculation has been removed because POLARXY is given in AN table header already in arcsec. Thanks to Lincoln Greenhill for pointing out to the mistake. Moved nowhere. 10888. June 11, 2001 UPDCOPYSUB Eric This routine was creating .old but then leaving it on successful conclusion. Added an rm to delete it. They are piling up. Moved nowhere. 10889. June 11, 2001 CCOPTS Eric Our users are not building the load modules with 64-bit capability even when their machine has it. Changing: $SYSLINUX, $SYSSUL, $SYSAXLIN, $SYSPPLIN CCOPTS.SH to specify COMP="$COMP -D_FILE_OFFSET_BITS=64" all the time. Moved nowhere. 10890. June 12, 2001 POSSM Eric Oh no - Phil code!!!! 1) It was plotting all n IF plots to the TV with no pause on NPLOTS=0. Added a pause. 2) On NPLOTS=0, it set PBIF=PEIF=1 for convenience in plotting the first time interval. Then it set BIF=PBIF and EIF=PEIF somewwhere during the second plot interval! Dropped the setting to 1 and handled the inconvenience another way. 3) The UV data array was summed from BIF to EIF as 1 -> (EIF-BIF+1) but then addressed as BIF to EIF. If BIF was 2 the last plotted IF was all zero. If BIF was 3 of 4 then a zero divide occurred. 4) The NPLOTS > 0, reset the IF selection based on the IF header coming out of UVGET. If BIF=EIF=2, there is 1 output IF so BIF was reset to 1 causing it to plot IFs 1 and 2. 5) Many places the plot parameter KLOCF(LOCNUM) was used to address ordinary headers - either in or out of UVGET. It neads JLOCF for output and KLOCFY for input (with new code to make sure the KLOCs get set). I am amazed it worked in labeling since JLOCF and KLOCF need not be the same and they were thoroughly confused. 6) On single-source files, it was not calling CHNDAT (early enough) but used the FINC, FOFF outputs of it. Moved the CHNDAT call to do it for all files and re-ordered the code so that the label initialization was called ahead of the use of these parameters. The frequencies and velocities output to the file were wrong. Moved nowhere. 10891. June 12, 2001 FITS disk Eric Changed FITS disk to C type I/O (except VMS). This should allow FITS disk files larger than 2 Gbytes. ZTPOPN $APLGEN:Fixed call to ZTPOP2 and ZTPOPD to send Hollerith file name. ZTPOPD $APLGEN: Fixed call seq for hollerith file name ZTPOPD $APLUNIX: C open routine for C type I/O ZTPMIO $APLUNIX: .FOR same as $APLGEN except a new call sequence for ZTPMID. ZTPMID $APLUNIX: C routine like zmi2 ZTPWAD $APLUNIX: C routine like zwai2 ZTPCLD $APLUNIX: C routine to close down like zdacls Removed IBM version ZTPMID, many many versions of ZTPWAD. Moved nowhere 10892. June 16, 2001 TVFLG Eric Found two more bugs in CLIP BY FORM. The first was a constant being passed down through several subroutine calls and eventually being set in a bottom level routine. It was 1 being set to MAX (1, MIN (NGREY, IPL)), i.e. 1, but some smart compilers protect dumn programmers from changing the value of 1. The second was a failure to set the main flagging Stokes pattern to the one desired at the moment. That pattern, not the one for the FC file controlled what was done. So 1011 was done for LL as well as RR. Moved nowhere. 10893. June 18, 2001 UVLIN, UVLSF Eric Changed remarks about FLUX to clarify the meaning - it does not refer to 10 seconds any more but to real weights. Also mentioned UVLSF in UVLIN help file. Moved nowhere. 10894. June 18, 2001 SETGDS Eric Improved the message when the non-3D image is offset. Moved nowhere. 10895. June 18, 2001 BPASS Eric Changed BPASSPRM(10) so that value 1 is the VLB option - normalize amplitudes using all channels, value 2 is normalize amplitude using ICHANSEL channels, and value 3 and up is normalize amplitude and zero phase using ICHANSEL channels. With these options, one may not need to divide by channel zero at all (unless phases are crazy and rapidly varying). Moved nowhere. 10896. June 21, 2001 VLBAUTIL:VLBACRPL Amy While trying to use VLBACRPL on a non-phase referencing experiment, I realized that the POSSM input SOLINT was necessary. So I added it. Changed both VLBAUTIL.001 and VLBACRPL.HLP Moved nowhere. 10897. June 21, 2001 POSSM Eric Which header is which was still confused and the use of KLOCF(LOCNUM) was a way out of it for Phil but most misleading. It did not get frequency right in UV plots, BP were okay. Moved nowhere. 10898. June 21m 2001 BPLOT Eric The character labeling had wrong variables for string length. It seemed to work for DOTV but not plot files. Changed the time labeling on LTYPE=3. Moved nowhere. 10899. June 22, 2001 FRPLT.HLP Leonia The adverb ANTENNAS has not be described clearly. It was not clearly told that the data are averaged for all selected baselines (included ANTENNAS) and the one plot is plotted indepedent on ANTENNAS. A paragraph has been added to clarify it. Moved nowhere. 10900. June 22, 2001 UVPRT, PRTUV Eric Made then systomatically check for high fluxes and use an appropriate ste of formats for them. Moved nowhere. 10901. June 22, 2001 AIPS Eric Updated the copyleft. Moved nowhere. 10902. June 26, 2001 ANTNUM, SCANTIME Amy Added line to the EXPLAIN files to demonstrate how to get these procedures to print the antenna number or scan time range, respectively, to the terminal. Moved nowhere. 10903. June 27, 2001 Model subtracting Eric Changed: CCSGRD - it was changing NSUBG (next cc to subtract) based on the requested, not actual, number of CCs. It also looped for more even if the first negative or too low flux was found. Former = flat error, latter harmless since all fluxes will be too low. GRDCRM Changed to set NSUBG and also, when DONE due to too low flux or first negative to reset NCLNG to the number that actually were used rather than the number requested (by the user or the CC file size). VISDFT Reset NCLNG on done. VSCDFT Reset NCLNG on done. VSFDFT Reset NCLNG on done. Moved nowhere. 10904. June 28, 2001 TVFLG Eric Failed on 1-polarization CLIP BY FORMs - accessing the "2nd" polarization in some odd and wrong way. Moved nowhere. 10905. July 2, 2001 SNPLT.FOR Leonia When STOKES='DIFF' and OPTYPE='AMP', SNPLT has calculated amplitude of vector differences of the two polarizations. Therefore the plotted amplitude has not been equaled zero for identical amplitudes of the two polarization, if the phases are not identical. Besides that the difference has not been multiplied by GMMOD, as it is done if a single polarization amplitude plotted. I have fixed it. Moved nowhere. 10906. July 2, 2001 GETCRD Eric GETCRD was changing an input variable if it was < 1. The new Solaris compiler is too smart for its own good and complained even though the input was always >= 1. Changed it to a local variable. Moved nowhere except U Colorado. 10907. July 2, 2001 FILLM Eric Changed the VLA center per Barry Clark note to aips++ giving the center used in 1995 coordinates in 1999. Also changed the default behavior of CPARM(7) for spectral line data: = 0 gets a new FQ on a big change, < 0 gets only 1 FQ in all cases. Also changed Chapters 4 and B of the CookBook. Moved nowhere. 10908. July 3, 2001 SNPLT Leonia Return back vector differences of the two polarizations. (see message 10905). The help file has been revised to emphasize that the difference may not be equal zero if the both polarization amplitudes are identical (OPTYPE= 'AMP'). Moved nowhere. 10909. July 3, 2001 FILLM Eric Changed it to rename a source if its RA and/or Dec differ from those of an identically named and qualified source. Added bit to CPARM(2) for planetary observations where position changes are to be ignored. Moved nowhere. 10910. July 4, 2001 TVDEVS.SH Eric The TV number can be > 9 in extended hex, so it must be REHEXed before comparing to 0. Vincent McIntyre of the ATNF for pointing this out. Moved nowhere. 10911. July 10, 2001 TVFLG, SPFLG Eric Added option to expand the time range flagged by DPARM(5) sec in each direction as the FC table is applied. Moved nowhere. 10912. July 12, 2001 AVSPC Eric Changed the default channels for "channel 0" to (n+1)/8+1 to n - (n+1)/8. This matches FILLM. Note that FILLM gave the wrong reference channel for channel 0 prior to the recent change to compute channel 0 internally. Moved nowhere. 10913. July 12, 2001 FDEFAULT.SH Eric/Shoshana The $SYSUNIX version had a string terminator mismatch. Moved nowhere. 10914. July 12, 2001 UVCOP Eric It uses the NX table to predict the size of the output file - a very bad idea for TASAV files. Changed to use min (nvis, prediction). Moved nowhere. 10915. July 12, 2001 INSTEP4 Eric Philippe Brieu of UCLA pointed out that a grep -v ^- could use quotes around the ^- => they are no harm and apparently are needed sometimes. Moved nowhere. 10916. July 13, 2001 AVSPC Eric Clarified BIF, EIF used only for AVIF mode. Moved nowhere. 10917. July 17, 2001 $TMPDIR Eric $TMPDIR is used by window managers etc and so should not be used by AIPS. Changed $SYSUNIX/ RUN, XASERVERS, START_AIPS, TVDEVS.SH, DADEVS.PL, SUB, UNIXSERVERS to use AIPSTMP instead. Moved nowhere. 10918. July 17, 2001 OPTIMIZE.LIST Eric DECBIT fails on the latest Solaris compilers. It has failed under HP and SGI so set it to 0 level for all. Moved nowhere. 10919. July 27, 2001 VLACALIB.HLP Amy Added DOBAND and BLVER so VLACALIB could be used with spectral line data. Moved nowhere. 10920. July 27, 2001 WENSS Eric/Wendy Lane Added Cygnus A to the WENSS survey tables. Moved nowhere. 10921. July 30, 2001 SETFC, FACES Eric Used the wrong array reference when seek the FREQ axis - it worked so long as there was a reasonable number of random parameters. Moved nowhere. 10922. August 1, 2001 VLBAUTIL.001 VLBACPOL.HLP Amy Changed last CLCAL in VLBACPOL to apply the SN table produced by POLSN to all antennas, as opposed to just the ones listed in the BASELINES parameter. Clarified the part of the help file about selecting a REFANT. Moved nowhere. 10923. August 2, 2001 CLPLT Eric CLPLT, when averaging spectral channels and/or IFs, called AVGCHN with a call sequence it NEVER had. Then it did an average anyway over the size of the record. Had the code worked the average of all the data twice except the first channel/IF which would have had weight 1 would have been plotted. In fact, the call to AVGCHN was so bad, one wonders what did happen. Moved nowhere. 10924. August 3, 2001 $APLNOT/INVERT.FOR Leonia The new subroutine INVERT calculates the inverse matrix of sizes up to 64x64. Moved nowhere. 10925. August 6, 2001 $APLSUB/JPRENU.FOR Leonia When DIR=-1 the time (relative the JD time) was changed by sign. This is the wrong way returning back to the mean coordinates, because the parameters of the precession matrix are described by a polynom including the second power term. The inverse matrix is calculated instead for DIR=-1. Moved nowhere. 10926. August 7, 2001 VLAPROCS.001,VLACLCAL.HLP Amy Cleaned up VLACLCAL. Specifically remove all "hidden inputs" and input OPCODE, which is and always was set by the procedure. DOPRINT and OUTPRINT were also added to the inputs so the LISTR output could be controlled. Edited help file to remove references to inputs that did not exist and added an explain section which describes the procedure. Moved nowhere. 10927. August 7, 2001 $APLSUB/JPRECS.FOR Leonia Several changes have been made: 1.Having calculated position and velocity of the Sun and the rotation matrix for the two given times (given JD and JD+DELDAT), the interpolation for any other time was carried out wrong. For example if the time is equal JD then the weight of its measurements was calculated as zero and the weight of the (JD+DELDAT)'s measurements was one. Obviously it should be vice versa. 2.I have changed the calculation of the relativistic light bending using the AIPS routine GRD which has been used at B1950 routine PRECES.FOR. Using GRD does not change the result but makes returning back (DIR=1 => DIR=-1) much simpler. 3.In the abberation I have simplified the coefficients checking that this change does not change the result. The abberation effect at DIR=-1 is estimated as the solid solution relatively the date coordinates. Such a solution guarantees the returning back (DIR=1 => DIR=-1). With these changes and changes at JPRENU (#10925) running JPRECS twice (DIR=1 and DIR=-1) returns coordinates back to the mean ones with accuracy better than 0.0001 marcsec. Moved nowhere. 10928. August 8, 2001 FITLD Eric The handling of CQ tables was never correct. At each new tape file, the old CQ table was read and each line was treated as a different integration-time/filter pair although in fact they were usually the same pair and different FQs or subarrays. Then the program tried to append at the end for new int-time/filter pairs, but with pointers that caused it to do nothing. Now the read extracts the pairs and at the end of each file the CQ table is rebuilt from scratch. Removed the DELCORR adverb - for the VLBA the CQ should always be written and it will not be written for the other correlators. Corrected DELDAT to 1.D-6 in the call to JPRECS. Moved nowhere. 10929. August 8, 2001 CLCOR Eric Changed the call to JPRECS also to 1.d-6 Moved nowhere. 10930. August 9, 2001 $APLSUB/JPRECS.FOR Leonia As it is indicated at the previous message (#10927) the abberation was represented by simple coefficients without relativism. It gives a small difference (~0.4mas). Now I have implemented the abberation with relativism. The inverse (DIR=-1) solution for abberation is still found by solving relevant system of equations relatively the date coordinates making the precise returning back (DIR=1 => DIR=-1). 10931. August 9, 2001 JPRECS, JPRENU Eric JPRECS and JPRENU were in $APLSUB but now call routines in $APLNOT so moved them to $APLNOT also. I cleaned up the code a tiny (I hope) bit too. Changed DELDAT in JPRECS calls in GETSOU and FRMAP to 1.D-6. Moved nowhere. 10932. August 10, 2001 SAD Eric Added option to loop over clip levels in selecting islands, fitting at a sequence of decreasing levels. Added option to scale printed peaks and fluxes by the primary beam. When the bandwidth smearing is selected, the printed peak values are increased by the smearing factor (in area). Changed the handling of deconvolution to allow partial solutions with 0 as the min value. Corrected the position angle of the bandwidth smearing and the deconvolved results by 90 degrees. Changed DECONV to be friendlier about allowing partial fits and forcing a sensible position angle. Changed IMFIT and JMFIT to understand the frienlieer DECONV. Moved nowhere. 10933. August 10, 2001 $APGNOT/UVFIX.FOR Leonia UVFIX uses the private version of JPRECS which allows to apply rotation and not rotation effects under control of input parameters. I have update the private version of JPRECS to the system version of JPRECS changed recently. Moved nowhere. 10935. August 20, 2001 $YPGNOT/VPLOT.FOR Leonia Averaging at time VPLOT has stopped averaging meeting the scan end if SOLINT>SCAN. I have added the new option at SOLINT. Now if SOLINT<0, then SOLINT= ABS(SOLINT) and time averaging does not pay attention on SCAN averaging until end of SOLINT. Moved nowhere. 10936. August 25, 2001 $YPGNOT/VPLOT.FOR Leonia Negative SOLINT allows averaging through scans if |SOLINT|>SCAN (see #10935). Therefore averaging different sources is possible. I have added a warning about this. At the edit option (APARM(1)=1) if autocorrelation is selected only the selected autocorrelation has been flagged. I have changed it. Now all baselines with selected antenna (OPTYPE=AUTO) are flagged at the found bad points. Moved nowhere. 10937. August 29, 2001 IMAGR Eric Changed IMAGR.FOR - FLUX negative does not care what the value of FLUX is but ABS(FLUX) was used as the default for WFLUX causing odd behaviors if users really paid no attention to the value in that case. Clarified help file too - I was confused. Moved nowhere. 10938. August 29, 2001 SAD, IMFIT, JMFIT Eric Changeed the deconvolution section to add 1 more deconvolution attempt to the setting of the max deconvolved result: the Clean beam itself plus 1 sigma. This should help with setting as better upper limit to the sizes. Moved nowhere. 10939. September 4, 2001 VLBACPOL.HLP Amy Changed help file so that it states when VLBACPOL should be run while reducing data. Also fixed the description of VLBACPOL's output. Moved nowhere. 10940. September 5, 2001 SAD Eric Inserted DPARM(2) option to cutoff on total component flux. Added asterisk to deconvolved sources smaller than the deconvolved Clean-beam + uncertainty (in major axis). Correctd deconvolved display -> did --- on 3/4 of the displays using AND rather than OR. Moved nowhere. 10941. September 6, 2001 JMFIT, IMFIT Eric Added the corrections for primary beam to peak and flux and for delay beam for peak to both tasks. Fixed display of cutoff level in PBCOR - it was not changed when the meaning changed to 1/previous meaning. Moved nowhere. 10942. September 13, 2001 FITLD Eric Changed it to forgive a missing a CQ table with warnings and setting DELCOR false. Then a stronger warning appears if it turns out that this is VLBA data. Moved nowhere. 10943. September 19, 2001 $APGNOT/ACCOR Leonia The default for SOLINT has been equal 60 min. Long time it has been supposed that the behaviour of the BBC's digitizer is stable in time. Now when it is clear that it is not right SOLINT=60min is too big. So the default changed for 2min. Moved nowhere. 10944. September 20, 2001 $APGNOT/ACCOR Leonia The negative SOLINT implemented this year (#10770) mean the averaging is carried out at the scan length independent on the value of the negative SOLINT. It is changed. Now If SOLINT.LT.0 then the ABS(SOLINT) is used as the average interval and averaging is terminating at the end of the scan. Because of the gaps between scans small intervals can appear at the scan ends. The intervals less than 0.2*MIN(SOLINT, SVAN) are not recorded to the output SN table. Formely the intervals less than 0.4*SOLINT were not not recorded to the output SN table. Default for SOLINT has been changed from 2min to -2min. Moved nowhere. 10945. September 21, 2001 $YPGNOT/UVPLT Jim Modified UVPLT to add an option allowing users to plot against uv distance in a user-specified position angle. This has been implemented in a new option 17 for BPARM(1) and BPARM(2), as well as addition of the adverb ROTATE which allows users to set the position angle on which they want to project the baseline. Moved nowhere. 10946. September 21, 2001 FILLM Bryan Butler/Eric 1. Added the creation of a weather (WX) table with subroutine FLMWX which writes to the table and a parameter in FLMTAB which tells it whether to call FLMWX to add an entry and timing variables to check entry timing. Created: WXINI which creates/opens the WX table TABWX which writes/reads the WX table DWXV.INC for common WX variables 2. Added the creation of a source position (PO) table *only* if the 'planets & Sun' bit is set (bit 4 of CPARM(2)) and if the source has coordinates of date (EPOCH=-1). Added subroutine FLMPO which writes to the table, a parameter to FLMTAB which tells it whether to call FLMPO to add an entry, timing variables to check entry timing, and a variable which indicates which sources are moving ones and which are not - in FLMSU and FLMPO. Created POINI which creates/opens the PO table TABPO which writes/reads the PO table DPOV.INC for common PO variables 3. Added modification of the CL table entries, based on opacity and gain curves (of course it can be turned off if desired). Added BPARM input parameter (reluctantly) to allow some flexibility. Added several common variables to indicate parameters and the type of opacity and gain curve correction and slightly modified input parameter handling (FLMIN) to deal with these new adverbs and parameters. Added a subroutine to read in a gain curve file. A standard VLA.GAINS file will be provided in $AIPSIONS. Added mechanics in FLMCL (the CL table writer) to check for the type of opacity and gain curve correction and do the necessary modification of the amplitude (real part) of the gain correction value. Added a subroutine to calculate the opacity given the weather information. 4. Other changes (while I was at it) - changed the shadowing check: if the 'default' (25 m) is asked for, the shadowing is now retrieved from the shadow bit on the archive tape. This is useful because the on-line version knows about *all* antennas (not just the ones in a given subarray) when it does its shadowing calculation (Ken tells me that this even includes the AAB!). This only required a small modification to the input parameter treatment, and a small modification in FLMUV where the shadowing is checked (instead of calling MCSHAD, the on-line shadowing bit is checked in this case). - made a new 'bit' logical in CPARM(2): . bit 8 makes it possible to ignore all source qualifier numbers, treating all sources with the same name as truly the same source (sets the qualifier for all to 0). This only required a small modification to the input parameter treatment, and a check in FLMSU. - Took out the reassignment of CPARM(8) and CPARM(9) in the table timing since the initial values may be needed later. 5. 2nd round of changes i made to FILLM (after eric AIPSisized it): - made references to light speed and pi (and 2*pi) come from INCS:PSTD.INC at various places throughout. - in MCREAD: some cosmetic changes (got rid of variable GOTEOF [used ISEOF in its place]); set ISEOT properly in the case of 2 EOFs encountered in a row. - in MCHEAD, added a check for crossing file boundaries, so an informational message can be printed at these. to facilitate this, needed to add 2 of the variables which are used in MCWANT (TSDATE and TSIAT0) to a COMMON in MC2.INC (and modified the use of these variables [TSOFF and TSIAT] in MCWANT cosmetically). - in MCRECO, changed the way that NOFILE is updated. As programmed previously, it updated twice for every EOF encountered (since MCHEAD was effectively called twice at every EOF [the first time hrough, MCHEAD traps ISEOF, and just falls through]). Also, added an update to NOFILE in MCREAD when tape headers are encountered. - in MCH1: modified the interpretation of the 'nibbles' for bandwidth code, front-end filter code (added code and variable [MCFEFW] for this), and correlators per baseline (i made these use ZGTBIT, which I think is infinitely more straightforward than what was in there before); modified setting of bandwidths in cases of bandwidth code of 8 or 9 (this was incorrectly set to be equivalent to 50/2**8 MHz in both cases, when in fact it should be set to be 50/2**9 MHz in both cases); also modified the bandwidths in cases where the front-end filter code indicates that the front-end filter bandwidth is less than the selected 'bandwidth'; took out a duplicate assignment to the calibrator code variable (TCALCD); - in FLMGN, added the parameter 'READ' to the 2nd call of ZTXIO, i.e., changed: CALL ZTXIO (LUNGN, INDGN, INLINE, IRET) - in PARSEL, changed II after the call to GETNUM, II is already set to the next white space, so doesn't need to be incremented by 2, only by 1 or 0. If it's incremented by 2, then the first date read in is missing its first digit and is wrong. - in FLMPO, made a check that this particular entry has not already been written into the PO table. For datasets with integration times less than 10 sec, consecutive entries can have the same exact values for the position (because the position is only updated every 10 sec tick in the online system), and we don't want duplicate entries. so, have to keep track of times+positions already written (in array TIMED), and check that the current entry hasn't already been written before adding it to the PO table (use logical variable ALREDY). - in FLMWX, added a check that the tape revision is >= 4 before adding a weather entry to the WX table (weather data was not written to tape prior to revision 4). Moved nowhere. 10947. September 24, 2001 PRINT Eric Added some digits to PRINT verb in KWICK.FOR and changed FRMT.FOR to make more low digits when useful. Moved nowhere. 10948. October 2, 2001 TECOR.HLP Amy Change the sentence in the help file that said dispersive delay corrections should not be turned on. I said that they depended on the accuracy of the input file, and should be checked carefully. Moved nowhere. 10949. October 9, 2001 CL2HF, ANTINI Eric CL2HF ignored the fact that table keywords are now all double precision and got DATUTC wrong. But a check of ANTINI, showed that ANTINI read it wrong also, getting the type handling right but using a pointer not properly set. UT1UTC was returned rather than DATUTC. Moved nowhere. 10950. October 10, 2001 FITLD Eric If the first FQ found is not number 1, it is switched to be number 1 to guarantee that there are data for number 1. The routine FR2FQ was prepared to find the frequency offset from the new FQ 1 and apply it to the other FQs. Unfortunately, if the offset did not occur in the first FR record (which it would not in this case) then the offset was zero for the first record(s) in the output FQ table and then non-zero for the later record(s). Added code to find the offset first, then to write the table. Moved nowhere. 10951. October 15, 2001 FILLM Brian/Eric Reset the planetary motion array when streams/files change. Moved nowhere. 10952. October 16, 2001 LISTR Eric Added DPARM(14) to get weights in a MATX display. Moved nowhere. 10953. October 23, 2001 TECOR Amy Changed TECOR so that it can handle multiple files. Added adverb NFILES. If NFILES > 1 then TECOR will read the day number and year from the file name given in INFILE, if the file name is in a standard format. It then will increment through as many day numbers as in NFILES. The standard format is CCCCdddC.yyC, where C is any character, ddd is the 3 digit day number and yy is the 2 digit year. The file in INFILE must also be the first day, the C's must be the same between days and the files must all be in the same directory. Moved nowhere. 10954. October 23, 2001 IMAGR Eric Fixed the TELL section's handling of FLUX. The FGAUSS cutoffs for each field were not updated. This caused the histogram to not be recomputed but did not stop the Clean - which caused numerous bad things. This error only applied if FLUX was set to something and then lowered with TELL and applied only to OVERLAP < 2. Moved nowhere. 10955. October 25, 2001 SLCOL Eric Changed the format and type of data displayed. It now does X, Y wrt the true ref pixel, Distance wrt user-spec pixel (ref pixel default) with a sign set by X wrt that point. All models and all components are displayed. Moved nowhere. 10956. October 25, 2001 FRING Eric Changed GRDCRM to test Gaussian widths with less precision than a total match. Fixed FRING to use a proper CMETH of 4 characters - it was failing to use the input string. Moved nowhere. 10957. November 1, 2001 FITLD Eric Changed the size parameters at the start to standard ones from PUVD.INC. IT was limiting the number of channels to 2048 and MAXCIF to 4096. Moved nowhere. 10958. November 1, 2001 ISPEC Eric Used REAL for frequencies - DOUBLE pRECISION is needed at high frequencies with narrow channels. Moved nowhere. 10959. November 2, 2001 CHKFC Eric Changed the character writing into the facets to allow the option of large characters (same lousy font). Moved nowhere. 10960. November 2. 2001 FITLD Eric I guess Phil never figured that AIPS "bytes" (1/2 word) were not the same as real tape bytes. He tested real tape bytes against buffer sizes in AIPS bytes. For the user with large data this will not do. Moved nowhere. 10961. November 5, 2001 MWFLT Eric One 2048 sized buffer remained -> MAXIMG. Moved nowhere. 10962. November 6, 2001 TECOR.HLP Amy Added paragraph mentioning all the IONEX files available on CDDIS. Moved nowhere. 10963. November 15, 2001 COMB Leonia COMB has the two opcode 'MEAN' and 'SUM' which carry out similar operation with the two given map. Both make the linear combination of the input maps (MEAN do not have the free term C3). But MEAN normalises the coefficients A1 and A2 by their sum. The difference between SUM and MEAN has not been clearly described at the help file. So the relevant description has been added. SUM and MEAN have treated the blanked pixels differentely. By request of Lincoln the new OPCODE='SUMM' has been added. The difference between SUMM and SUM is only at blanked pixels treatment. SUM puts blank at output if the MAP1 or the MAP2 pixel is blank. SUMM puts blank at output if the both MAP1 and MAP2 pixels are blank. If one of the map's pixel is blank, SUMM puts this map's term at the linear combination as zero. Moved nowhere. 10964. November 15, 2001 FITDISK Amy Procedure using FITTP to write files to disk using an automatically generated name based on the file name in AIPS. Internal blanks are changed to '_' and trailing blanks omitted. Moved nowhere. 10965. November 20, 2001 UVCOP Eric Failed to apply any flag table entry with non-zero source number to single-source data. Changed it to apply all source numbers to such data. Moved nowhere. 10966. November 21, 2001 SCANTIME (part of VLBAUTIL) Amy SCANTIME used to calculate the beginning and end time of a scan by taking the time in the NX table and adding and subtracting the time on scan ('TIME INTER) in the NX table. This is incorrect, the time on scan should be divided by 2 and added and subtracted to the time in the NX table. Moved nowhere. 10967. November 21, 2001 FITLD Eric For some reason the correlator sometimes writes data with FQIDs having identiacl frequencies. This confuses things no end when concatenating to files that have 1 FQID of the same value. Made another Band-aid patch to translate the "duplicate" FQIDs to the lowest one before any of the other translations. Moved nowhere. 10968. November 21, 2001 UVLOD, FITLD Eric UVLOD (and its copy in FITLD) attempt to convert abnormal Stokes and Complex axes to normal ones. There was no attention paid to a change in the axis length affecting how much to read. All converted Stokes axes were length 4 even if there were not 4 incoming values to do the conversion. Changed to avoid the conversion in some more cases. Moved nowhere. 10969. November 23, 2001 SLCOL Eric Fixed it to honor the replace text file option as well as append. Changed how it is requested. Moved nowhere. 10970. November 23, 2001 FITAB Eric The logic testing the vis range could result in an infinite loop where IPIECE < NPIECE always. Fixed. This showed up on a file with no vis recs. Moved nowhere. 10971. November 23, 2001 VLBAUTIL.001 Eric The CPARMs were not set right for SNSMO clipping bad amplitudes. The clip number was put in the phase smooth time instead of CPARM(6). Moved nowhere. 10972. November 23, 2001 SPLIT Eric When calibrating a single-source file it changes the velocity type and reference based on the non-existent source table. Changed it to change the type only when multi-source. Moved nowhere. 10973. November 25, 2001 USAVE Eric New pseduo-verb: SAVEs to another user. Changed POPSDAT.HLP, STORES.FOR, USAVE.HLP. Moved nowhere. 10974. November 26, 2001 XAS.SHR Eric The order and placement of the RGB bytes in the 24- or 32-bit word for TrueColor displays depends on the byte order of the display machine, not the byte order of the machine running XAS. The Image structure may be used to give this byte order, while shared memory usage guarantees that the two are the same. Moved from the AOC this date. 10975. November 28, 2001 ELINT Leonia Rick Perley requests printing error of the polynom coefficients. I have modify ELINT to do it. Print of source factors for each antenna has been removed. The help file has been updated to clarify PRTLEV=0,1,2. Moved nowhere. 10976. November 29, 2001 IMAGR Eric Clarified the handling of BOXFILE - now leading blanks are ignored in lines of all types. Changed IMAGR, FIELDF, WINDF, WRIBOX, AU5C (verb FILEBOX) to handle the files this way. Fixed QCLEAN to be forgiving about errors when prompting for a field number (in force a field or select a field). Moved nowhere. 10977. December 4, 2001 SNPLT Amy. Changed SNPLT so that it checks the weights of all IFs, rather than only the first. Also changed SNPLT to set the amp, phase, rate and delay for a specific IF to FBLANK if the weight for that IF was negative. Moved nowhere. 10978. December 4, 2001 GETTIM Eric Changed subroutine to regard GMRT as a VLA-like antenna and to set the error code for unrecognized arrays only if the array center is zero (i.e. VLB). Affects CVEL.' Moved nowhere. 10979. December 5, 2001 MAKE.MNJ Eric Added lines to allow it to accept an argument that only contains the date without the time. This save fixing update.pl and install.pl and puts the correct .000000 that one would want in those cases. Moved nowhere. 10980. December 5, 2001 MNJ Eric/Pat Changed putbck system's addrec.c to use UT (gmtime) rather than localtime in the *.UPD files. Changed $UPDUNIX/AIPSUPD to fetch the time string as UT in form YYYYMMDD.HHMMSS to name files and to use in the LAST*.DAT* files. Changed $UPDUNIX/UPDLSTDAT.c to build a comparable string from the transaction files for comparing with the current time. Thus we will not have to wrry about crossovers of dates etc between the aips repository (currently CV) and the MNJ site. This will force the MNJ sites to recompile UPDLSTDAT.c. Moved nowhere. 10981. December 5, 2001 SNEDT Eric The EDITSN subroutine in QEDIUTIL had the logic wrong on ENTER IF when the IF is predictable. It is not predictable when crowded. Moved nowhere. 10982. December 6, 2001 PCAL, LPCAL Eric Both tasks play games with BIF/EIF internally and then wrote them to the HI file. Fixed the HI file to get the actual full range used. Moved nowhere. 10983. December 6, 2001 help files Eric Updated INTERPOL.HLP with info from the tasks that use it. Updated TVCPS.HLP by dropping all the film recorder info. Moved nowhere. 10984. December 6, 2001 FRING.HLP Amy Changed the default of APARM(1) from 6 to the correct value, 3. Moved nowhere. 10985. December 10, 2001 ABOUT etc Eric Updatedthe ABOUT files ZZANALYSIS.HLP ZZCOORDINA.HLP ZZFITS.HLP ZZIMAGE-UT.HLP ZZMODELING.HLP ZZPOPS.HLP ZZPROCEDUR.HLP ZZTASK.HLP ZZTV.HLP ZZUTILITY.HLP ZZUV.HLP ZZVERB.HLP ZZVLBI.HLP Also the tab completion file: HLPIT.LIS Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC01 is TST *********************************************************************** *********************************************************************** 10987. December 18, 2001 DFTPL Eric The code that finds the I Stokes was really messed up! It counted the correlators that make up the I and then divided by that count. But the count was zeroed way outside the loop and the other parts were in the innermost part of the loop. And there was no test for zero divide. Moved from 31DEC02 Dec 18. 10988. December 19, 2001 VLBACALA, VLBASUBA, VLBAFQS Amy The biggest change was made to VLBACALA where I changed the CPARM(6) (max deviation for the amplitudes) in the second run on SNSMO to 1.0. Since the max deviation is in real amplitudes (i.e. not fractions or percents) 0.02 was much too small. Changed VLBASUBA and VLBAFQS so that the default of CLINT is actually 1 minute as claimed in the help file. Previously it had been left to the default in INDXR which is 5 minutes. Moved to 31DEC01 this date. 10989. December 28, 2001 update.pl Eric Fixed to edit PAPC.INC into $INC/NOTST/$ARCH whether or not it started there. Previously it edited it in $INC! Moved from 31DEC02 this date.. 10990. December 29, 2001 CPASS.HLP Eric The help file lied about the auto normalization after averaging which was also done in CPARM(8)=0. Fixed that and added comments about normalizations. Moved from 31DEC02 this date. 10991. December 29, 2001 UVWAX Eric/Joe Lazio/Wendy Lane New task - like UVNOU but deletes data near 0 in u and/or near 0 in v and/or at high values of w. Moved from 31DEC02 this date.