`;----------------------------------------------------------------------- ; Copyright (C) 2005-2007 ; 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 SOCORRO. 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 31DEC06 is TST *********************************************************************** *********************************************************************** 11990. December 4, 2005 Version update Eric Changed GETRLS, SETPAR, INSTEP1, install.pl, update.pl, ZXHLP2.PL, UPDCONTROL, AIPSPATH.*SH to put 31DEC06 in place of 31DEC05. Moved nowhere. 11991. December 4, 2005 TVFLG, SPFLG Eric The gridding routine needed a +0.001 to avoid numeric errors with computing the row to grid. Moved to 31DEC05 this date. 11992. December 6, 2005 APCAL Eric Bad call seq to PLCLOS caused Intel load modules to abort. Moved to 31DEC05 this date, nowhere else. 11993. December 6, 2005 IMAGR Eric MINPCH was allowed to change to match window sizes etc. Fixed to use a local value based on the user's MINPATCH without changing the user's value in routine CDECID in QCLEAN. Moved to 31EC05 this date, nowhere else. 11994. December 8, 2005 install.pl Eric Made grammar changes suggested by Chris Phillips (ATNF) and also added a read after an information display (of available OS flavors) so that the user has time to read the display. Moved changes to 31DEC05 this date. 11995. December 9, 2005 DSORC Eric Olaf Wucknitz (Jive) pointed out an error in the routine that calls the source changing so that the changing was done on the input not the output. Moved to 31DEC05 this date. 11996. December 9, 2005 CCRES Eric NEW TASK - like RSTOR except that it uses brute force methods. Thus it can subtract and/or add Celan components from/to any image without requiring the components to be on cells or the images to be powers of 2 in size. A model file of varying Gaussian widths may be subtracted or added also. Moved to 31DEC05 this date. 11997. December 12, 2005 TVCPS Eric Corrected it to read more than one image from disk when more than one is visible. Note that this works only when the images are on different image planes and are the same in general image size and coordinates. Moved nowhere. 11998. December 16, 2005 CPASS Eric Changed it to accept error code 3 from the solver as normal, error code 2 is reported but the answer accepted anyway (no converge), and so it dies only on error code -1 - input parameter error. Fixed help file to tell defaults and suggest that many iterations and larger tolerance is needed. Moved to 31DEC05 this date. 11999. December 16, 2005 Linux FDEFAULT.SH Eric Changed $SYSLINUX references to *g77 and *gcc into *g77* and *gcc* so that versions named e.g. g77-3.4 will be recognized in FDEFAULT.SH, LDOPTS.SH and CCOPTS.SH. Moved nowhere. 12000. December 27, 2005 UVCON Leonia The primary beam of an array can be variable at time if the array element is array itself (antenna station). It happens because the projection of the antenna station changes when the elevation of the source changes. LWA antenna station and VLA (when used at the phased array for VLBI experiments) are the typical examples. The new version of UVCON considers the primary beam as the circular two dimensional Gaussian with the given BMAJ=BMIN. Elevation (EL) and azimuth (AZ) are calculated for each visibility. BMAJ is recalculated as BMAJ=BMIN/sin(EL). BPA of the primary beam is calculated as BPA=AZ. This new option works (under control of BPARM(6)) only if the source model is given as a set of CC components (CMETHOD ='DFT', CMODEL ='COMP'). Moved nowhere. 12001. December 29, 2005 *GEOM helps Eric The explain files for LGEOM, HGEOM, and PGEOM used the APARM nomenclature for the order of interpolation in part and then used the actual order number in the rest of the explanation. Changed it to use the APARM(n) nomenclature solely. The order is APARM(n)/2 where n=1 for HGEOM and n=6 for the other two. Moved to 31DEC05 this date. 12002. December 29, 2005 TABHDR Eric Changed it to complain but not quit when a recognized keyword has unrecognizable value, e.g. CRVAL1 = 'Right ascension reference value' as found on an HST tape! Moved nowhere. 12003. January 4, 2006 VLBAUTIL, VLBALOAD, VLBAEOPS, VLBATECR Amy Fixed VLBALOAD, the final INDXR did not work if people used the new load data from disk option. Fixed VLBAEOPS and VLBATECR to set SYS2COM and SYSOUT to '' at beginning Moved nowhere. 12004. January 5, 2006 XASERVERS, UNIXSERVERS Eric There are too many sites for which the quote marks do not work in the xterm -e 'message-server ; sleep 3 ' commands and only one site that I am sure requires them. Of course the sleep 3 lets the sites that cannot use the quotes to read the message. I have dropped the quotes so some sites will now have an xterm flash by with no way to see what is wrong. Moved to a re-opened 31DEC05 this date. 12005. January 6, 2006 FILLM, PRTAN Eric Changed FILLM to detect EVLA antennas and name them as such ('EVLA:aNN') and changed the form of the VLA names to match ('VLA: aNN'). Changed PRTAN to recognize either as VLA along with the old form of VLA name ('VLA:aNN '). Moved nowhere. 12006. January 6, 2006 TVUTIL Eric Missed one variable that needs to be LONGINT - caused DOTV in IMAGR to abort on AMD 64s. Moved to 31DEC05 this date. 12007. January 9, 2006 VBGLU Amy The original VBGLU assumed a strict time order for the tables it was merging. This was a problem for the TY table which is in antenna-time order. Fixed that and added a feature where if less than 95% of the 2nd table is merged with the first a loud warning message is given and the output table is deleted. This seems to be a problem if the times of the 2 tables are not lined up. Moved nowhere. 12008. January 12, 2006 VBGLU Amy Wouldn't compile under Macs because of a couple superfluous commas in WRITE statements. Fixed. Moved nowhere. 12009. January 16, 2006 UVPLT Eric Changed labeling for antennas to take into account de-selection and to show 0 as * even on deselection. Moved nowhere. 12010. January 17, 2006 install.pl Eric Two places fixed where words ran together in messages to user. Moved nowhere. 12011. January 17, 2006 DBCON Eric Fixed code to handle FQ entries that do not have the same channel width (in magnitude or sign) as the header. This matters when the reference channel is changed to 1. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12012. January 18, 2006 FQ increments Eric Changed to allow for channel increments that differ between IFs in the FQ table (due to sideband or real differences) DGHEAD To set the increment of BIF into the CATBLK. This will force a major relink of tasks. SPLAT To adjust the output FQ table entries to account for averaging mode 3 with differing increments in BIF and higher IF numbers. SPLIT Similar change required. Changed SPLIT and SPLAT help files to clarify the differences and which tables are copied. Moved nowhere. 12013. January 18, 2006 tapes Eric $APLGEN versions of ZTAPR and ZMOUNR did not use the right format for the new LUNs. Moved nowhere. 12014. January 19, 2006 DSMEAR Eric This VLBI de-smearing routine did not handle FQIDs correctly. It should test for a match or for either the CQ table or the calling routine FQ <= 0. It only tested for a match so when there is only 1 FQ and hence no FQID parameter, the routine failed. Made fixes to BPASET, BPGET, CSLGET to require > 0 for both FREQ IDs or a match. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12015. January 20, 2006 tapes Eric Changed $APLUNIX/ZABOR2.C to ignore SIGPIPE so that AIPS and TPMON et al will not abort on certain tape errors. Let's hope this helps more than hurts. Moved nowhere. 12016. January 23, 2006 SAD Eric Fixed the routine that computes a string of the RA/Dec in sexagesimal. The round-off terms were not reduced when the accuracy was increased - the resulting display was 0.0005 seconds of time and 0.005 seconds of arc too large. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12017. January 24, 2006 CPASS Eric Lynn Matthews pointed out a bug in handling the HI cards so that all HI entries after ANTWT were omitted. Moved nowhere. 12018. January 25, 2006 ELINT Leonia Rick Perley erroniously put inputs for CALSOUR='' and got a wrong answer of ELINT. Actually ELINT requires explicit list of the CALSOUR. So ELINT is changed to terminate its job if CALSOUR='' with the relevant message. Help file is updated. Moved nowhere. 12019. January 27, 2006 CPASS Eric Added option to write out fit result as a normal BP table rather than the results of the fit. This allows time smoothing of real and imaginary rather than of Chebyshev polynomial parameters. Made this option the default. Moved nowhere. 12020. January 30, 2006 VBGLU Eric Fixed the memory allocation for pointers to four times the absolute minimum (stll << 1000000) and changed to use dynamic memory generously for the comparison buffer. Changed the logic that decided when and how to compress that buffer by putting the UVDISK call ahead so that the correct amounts would be known. Added test for overflow as well. Moved nowhere. 12021. January 30, 2006 WIPER Eric Corrected source number initialization to -1 since 0 is what one gets for single-source files and elevation etc require a source parameter lookup. Added a call to IAXIS1 to display axis labels whenever the full image is visible. Moved source numebr to 31DEC05 via patch 21 Feb 2006. 12022. January 30, 2006 CPYRT Eric Cleaned up typing of help file and changed the copyright statement for non-NVSS images. Moved nowhere. 12023. January 31, 2006 VBGLU Eric Found more addressing problems triggered by data present in data set 2 but missing in data set 1. These should be counted if possible but I just did a fix to keep it from screwing up royally. Moved nowhere. 12024. February 1, 2006 VLBAARCH Lorant Fixed some code to properly deal with bandcodes - a feature for unusual experiments that apparently was never used so far. Moved nowhere. 12025. February 1, 2006 LOCIT Eric Adjusted the VLA-specific parts of the station ID code to match the new (and old) formats of FILLM. Moved nowhere. 12026. February 7, 2006 CLCOR Leonia CLCOR has interpreted the leap second using MOD for the IAT-UT1 values read from the USNO file. It is wrong in some cases for example at 2006 when the leap second jumped from 32 to 33 second. Now the leap second added to the USNO data before interpolation. The leap second is picked up from the CT table (the first value). The problem can happen at the rare case of changing the leap second during the interpolation interval. Moved nowhere. 12027. February 8, 2006 UVCON Leonia The history file has not included the name of the model file. I added it. Moved nowhere. 12028. February 10, 2006 CLCOR Leonia Recent fix of the leap second problem (#12025) does not care the case of the leap second jump during the interpolation interval. The J. Benson's interpolating subroutine UT1COR has been modified adding the leap second input as an array picked up from the CT table. So the UT1-UTC is converted to the smooth function UT1-TAI before interpolation. The leap second of the observation day added to interpolated UT1-TAI to return back to UT1-UTC. Moved nowhere. 12029. February 10, 2006 UVCON Leonia UVCON has failued when APARM(7)= integration/separation time=0. It happens because the number of time intervals is calculated as (Hmax-Hmin)/APARM(7) + 1. Inspite of the fact that it is rather natural to give an explit value for APARM(7), some users still leave APARM(7)=0. So I added the default for APARM(7): 0 => 1.E6. Such a big value provides the number of time intervals equaled 1. Therefore now APARM(7)=0 means simulate snapshot observation near hour angle Hmin. The default is added to the APARM(10)-blockage as well. 0 => 1. Maximum fractional area blockage of one antenna by any other is equal 1 (the total blockage). Therefore the default (1) allows any blockage. Moved nowhere. 12030. February 10, 2006 LOCIT Amy The order of statements caused the pad name to be truncated. Using a second variable, repaired the problem. Moved nowhere. 12031. February 17, 2006 SNEDT Eric Changed SNEDT and QEDIUTIL to allow for editing of TY values. A number of problems, esp with crowded displays, were found in the existing routines. A SET VALUES function was added as was the re-writing of the TY table with the changed values. Moved nowhere. 12032. February 17, 2006 LISTR Eric In the gain averaging subroutine, the conversion of times from double precision to single caused enough loss of accuracy (when the times were 2.5 days or more) that each record appeared to be at a later time. This caused the task to print each gain on a separate line. A change to double precision was enough to avoid this problem. Moved nowhere. 12033. January 17, 2006 SPLIT Eric The first UVGET ('INIT' call for each source did not set UVOPEN until after branching on error. Therefore it did not close it. Therefore it could not open things for the next source. Moved the UVOPEN statement. Moved nowhere. 12034. February 19, 2006 SPLAT Eric Changed it to force the creation of a multi-sourrce file even when only one source is selected if the assemble option (APARM(7) = 0) is selected. Moved nowhere. 12035. February 20, 2006 SETFC Eric Found an error in setting the boxes around NVSS sources. The X position was computed by dividing the RA difference by the cos of declination. That should be a multiplication when converting to pixels. Moved to 31DEC05 via patch 21 Feb 2006. 12036. February 20, 2006 VBMRG Eric Correct VBMRG to (1) allow records longer than 2000 words, (2) to read/write compressed data, and (3) to be a bit more modern in typing/coding style. Changed VBMRG help file to warn that results are changed by compression. Moved nowhere. 12037. February 21, 2006 FLATN Eric Changed INTERPLATE.FOR - several intermediate variables needed to be LONGINT. Moved to 31DEC05 via patch 21 Feb 2006. 12038. February 22, 2006 IBLED Eric Changed: 1. UVGET had a bad error message which had mismatched variable list and format 2. IBLED did not set error returns in the init routine and did not check for them in main 3. IBLED did not provide a sensible default for NMAPS and used numerical values for call arguments that are in/out 4. IBLED perhaps due to changes in FACGRD failed to detect missing model images (it looks for I, R, and L) and quit instead if all 3 were not present. Moved nowhere - went to patch 31DEC05 23 Aug 2006. 12039. February 24, 2006 IMAGR Eric Changed IMAGR.FOR and .HLP to allow beamss of 4086x4096 when IMAGRPRM(10) is 1. Changed QCLEAN to do the beam histogram over 2060 rather than 1030 values and allow the maximum patch to be 2048. If that does not fit in the AP, adjusted the upper limit to be a bit less than before since the previous adjustment still led to an overflow. Fixed an error message that said F7.0 on an integer variable. Moved nowhere. 12040. February 24, 2006 VLOG Eric Changed status to WRIT on first CATIO and added another CATIO at the end of the history routine to reset the status. Moved nowhere. 12041. February 27, 2006 CLCAL, SN2CL Amy There is a feature of CLCAL where if a SOURCE does not exist in the input CL table that the SN table then the SN entries for CALSOUR are appended to the SN table. This caused a problem for a user who already had entries for that CALSOUR. I changed SN2CL so that it makes a table of sources, subarrays, frequencies and time ranges that exist in the input CL table and if the SN entry that it wants to append is inside this time range for the specific source (subarray and freq) it does not append that entry. BLAPP also used SN2CL was also recompiled. Moved nowhere. 12042. February 28, 2006 FILLM Eric Added a test for telescopes on the master pad (MPD) so that they will not be called "OUT". Moved to on-line FILLM today, nowhere else. 12043. March 1, 2006 install.pl Eric Added code early in the script to detect use as root and to forbid that, exiting cleanly but abruptly. Moved nowhere. 12044. March 2, 2006 CLCOR Leonia Correction of multi band delay is added at the case of the OPTYPE='ANTP'. Looking at the CLCOR's codes I found that multi band delay correction is implemented only for OPTYPE added recentely 'atmo', 'eops', 'suns' and 'antp' now. The multi band delay correction and clock correction have been implemented only for the first stokes, although formally CL table includes them for both stokes. So the multi band delay correction and clock correction for the second stokes are added. Moved nowhere. 12045. March 3, 2006 CLCAL, SN2CL Amy Added warning message associated with change of Feb 27. It will not tell the user it could not find the selected SOURCE, TIMERANGE, SUBARRAY or FREQID in the input CL table and therefore was going to attempt to append SN table. Moved nowhere. 12046. March 6, 2006 VBGLU Eric Rewrote the task to glue together all times from all files, discarding no data. The older algorithm was more complicated and difficult and lost any data found in files 2-4 but not in file 1. Moved nowhere. 12047. March 6, 2006 TABF3D Eric Fixed TABF3D subroutine so that the default column count is 1. It was returning zero. Linked FITLD, UVLOD, IMLOD, etc. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12048. March 7, 2006 VLARUN+HLP Lorant New VLARUN has many upgrades such as autoflagging, full primary beam, check on Fast-switching names and some imaging parameters. Determines theoretical rms, integration time and has an option for self-cal (which is still a bit fragile). More to follow. Moved nowhere. 12048. March 11, 2006 FILLM Eric PT was listed as EVLA since it is connected by fiber. Fixed this special case. Moved nowhere. 12049. March 11, 2006 LIBR.DAT Eric The $SYSLINUX LIBR.DAT had the QYPGNOT list rather wrong. curious that it seems to work. Other versions are fine. Moved nowhere. 12050. March 12, 2006 xxINI routines Eric The xxINI routines need to close the file after the first round of initialization, reopen it, find the diesired columns, and check for (but survive) missing or empty ones. This allows for tables that have all the desired columns but in a different order than "standard" and provides for warnings. Routines used only in FITLD that did the search for columns, but assumed that all were in order: AGINI ATINI BCINI CAINI CMINI FRINI SOINI BTINI Routine in FITLD that assumed all was in order with no search for columns: GAINI Routines with elaborate Chris Flatters code needing to be toned down and to allow missing columns to be reported but not kill things: FLINI GNINI PHINI TSINI More general routines, needing messages on number values and missing columns and more use of TABERR ANTINI CTINI FQINI POINI WXINI BLINI BPINI BSINI CALINI CQINI GCINI HFINI IMINI MCINI OBINI SNINI SOUINI TYINI General routines needing to test for columns FCINI MFINI CCMINI CSINI FLGINI GPINI IMINIT NDXINI OFINI OTTINI PCINI Generically - made all message levels 6 or 7 so they can be suppressed. Changed TABERR - it suppressed the message suppression. Changed it to honor suppression. Corrected the variable declarations - incoming strings should not have a length declared. It claimed 6 for the subroutine name which is usually 5 and so printed an extra character (or a null stopping the rest of the message). ANTINI will not warn on empty columns of ORBPARM, POLCALA, POLCALB. Moved nowhere. 12051. March 13, 2006 WIPER Eric Added a history writing subroutine with an elaborate attempt to list all sources included. Moved nowhere. 12052. March 14, 2006 CALIB, FRING, KRING Amy The loop that set ANTWT only went from 1-20 (i.e., only the first 20 were set to the input ANTWTs). The loop was changed to 1-30. Moved nowhere. 12053. March 15, 2006 INI routines Eric Changed AGINI to skip complaints about the ORBPARM column. Changed CTINI to omit complaints about the TIME INT column when the table has fewer than 12 columns. Changed SOINI to treat EPOCH and EQUINOX as equivalent column labels. Moved nowhere. 12054. March 17, 2006 Intel compiler Eric The Intel compiler is up to its trick of changing doubles to singles prematurely. Changed: POPSRD New routine to do the conversion between POPS 2-reals representation of doubles and true doubles OPTIMIZE.LIS Added POPSRD with no optimization AU7 Verbs ALTDEF and AXDEFINE adverbs RESTFREQ and AXVAL to use POPSRD AU7A Adverb KEYVALUE in verbs EGETHEAD, GETHEAD, GETTHEAD, PUTHEAD, PUTTHEAD, TABGET, TABPUT to use POPSRD PUTKEY Used by AU7A to put keyword values CATKEY Several statements checking the range of allowed types were incorrectly typed and so checked that the type was >= 1 twice rather than >= 1 and <= 5. TBHEAD Use POPSRD TBDATA Use POPSRD TABED Corrected to use POPSRD for double and KEYV(1) for single precision; it used the sum of KEYV(1) and KEYV(2) set back into the single and did not test for the two halves being equal (which we assume the user does not mean). Changed help file to be clear on this point. VLBIN Changed to use POPSRD for rest frequency SETJY Changed to use POPSRD for rest frequency Moved nowhere. 12055. March 19, 2006 IMMOD Eric Removed arbitrary -180 to 180 limit on FWIDTH in help file. Moved nowhere. 12056. March 21, 2006 More INI routines Eric Suppressed messages for polarization 2 when there is only one polarization in IMINI, MCINI, GCINI, TYINI. Corrected nasty typo in CATKEY created March 17. It caused catalog keywords to have only one word of data stored, messing up double precisions and characters. This has a bad affect on FITLD. Moved nowhere. 12057. March 21, 2006 ANBPL Eric new task to print and plot uv-data converted to antenna-based quantities. Weights, amplitudes, phases, reals and imaginaries computed in a couple of ways can be plotted against time, elevation, hour angle, azimuth, and parallactic angle. The plots are similar to ones from VPLOT and the prints are the LIST form of LISTR more or less with user control of the number of digits. Moved nowhere. 12058. March 22, 2006 Mac Intel Eric Added Mac Intel port to system by changing AIPSPATH.CSH Added cases for $ARCH MACINT with library INTELCMP AIPSPATH.SH Added cases for $ARCH MACINT with library INTELCMP AREAS.CSH Added $APLMACINT, $INCMACINT, $SYSMACINT; changed old MAC ones to $APLMACPPC, $INCMACPPC, $SYSMACPPC AREAS.SH Added $APLMACINT, $INCMACINT, $SYSMACINT; changed old MAC ones to $APLMACPPC, $INCMACPPC, $SYSMACPPC LIBR $ARCH case to MACPPC or MACINT and $ARCH test for MACPPC to MAC* START_AIPS $ARCH tests for MACPPC to MAC* TVALT Added case for $ARCH = MACINT ZLASCL $ARCH tests for MACPPC to MAC* ZLPCL2 $ARCH tests for MACPPC to MAC* UPDUPDATE special code for macs no longer needed. MAKE.MNJ Add MACINT to list MAKE.BMNJ Add MACINT to list UPDCONTROL Change to MAC* UPDERROR Change to MAC* install.pl Use $mach = i[34567]86 to recognize MACINT and add that to comments XAS.SHR Add MACINT to Makefile options In $SYSMACPPC: COMRPL.ADD Change to MACPPC and drop APLLINUX LIBR.DAT Change $APLMAC to $APLMACPPC INCS.SH Change $INCMAC to $INCMACPPC In new $SYSMACINT coped from $SYSMACPPC: CDVER.SH no real change COMLNK.ADD no change COMRPL.ADD Change to MACINT and drop APLLINUX INCS.SH change to $INCMACINT LIBR.DAT Change $APLMAC to $APLMACINT and change APLMAC directory name to APLMACINT (PPC will keep APLMAC since it is already in use) CCOPTS.SH Add Intel compiler options LDOPTS.SH Add Intel compiler options FDEFAULT.SH Add Intel compiler options In new $APLMACINT copied $APLMACPPC and changed: ZDCHI2.FOR Changed byte order code to 3, SYSTYP and SYSVER Copyright and version info updated in ZCPU.C, ZERRO2.C, ZFRE2.C, ZTACT2.C, ZTKILL.C, ZTXMA2.C and the FORTRAN stubs ZMOUN2, ZTAP2, and ZTPWA2. Moved nowhere. 12059. March 24, 2006 system areas Eric One of the update C routines requires directory logicals to be 8 or fewer characters. Changed $SYSAIPS/AREAS.DAT which I forgot and which is the most important and $SYSUNIX/AREAS.*SH to have the shorter names (APLMACPP and APLMACIN etc). Edited the old PUTBCK.UPD file to fix this. Moved nowhere - forced a MNJ. 12060. March 24, 2006 ANBPL Eric The solver routines are too fussy for this task. Changed ASOLVE Add FFLAST argument, change levels to be fussier until the last which can be set from outside - default 3.0 sigma. Change to put a floor on RMS so that it cannot reach too small so that math error looks like 10 sihma. VSOLVE Add FFLAST argument to set last cutoff level - default 2.5 sigma. Added an RMSMIN also. FINDR Changed all calls to ASOLVE and VSOLVE to add the control argument at its default value. FLAGR Changed all calls to ASOLVE and VSOLVE to add the control argument at its default value. ANBPL Changed calls to ASOLVE to cut at 4 sigma and to VSOLVE to cut at 3.5 sigma. Corrected to understand that <= 0 from ASOLVE means no good and (real,imag) = (0,0) means bad from VSOLVE. Changed ANBPL also to do multiple polarizations and to plot the IFs and polarizations together or separately under control of OPCODE. They are printed separately. Moved nowhere. 12061. March 26, 2006 Mac stuff Eric INCS.SH INCMACPPC -> INCMACPP also COMRPL.ADD INCS.SH INCMACINT -> INCMACIN also COMRPL.ADD UPDUPDATE Special code put back for $SYS$ARCH for Macs Moved nowhere. 12062. March 28, 2006 UVFIX Leonia Add 'VLBI' to the list of vlbi arrays: 'VLBA' and 'EVN'. The help file updated. Moved nowhere. 12063. March 28, 2006 DDBGR Eric NEW: Service program - dumps disk contents to debug the writing of disk files. Moved nowhere. 12064. March 28, 2006 BPLOT Eric Added PDIF operation to difference the BP phase with the average BP phase - the vector difference can get big differences. Fixed end handling so that it will view doing some plots and then hitting the end of data as normal. Moved nowhere. 12065. March 29, 2006 3C138_X Amy/Eric Put back Amy's new model of 3C138 at X band. It incorporates interfering sources from some distance away. Sorted CC file and ran IMEAN and wrote a sub-image with CALWR. Moved nowhere. 12066. March 31, 2006 BPLOT Eric Corrected tendency to label everything when that causes the Y axis labels to overlap. Added DO3COL option 2 to use color for intensity rather than time or baseline - allows closely spaced lines. Added TIMERANGE backwards to plot only TIMERANGE but average all times. Moved nowhere. 12067. April 1, 2006 KNTR Eric Fixed top axis labeling - needed more LOCNUM messing around to make labeling strings be properly determined. It labeled each image line with the same labels previously. Moved nowhere. 12068. April 1, 2006 PBCOR Eric Copy keywords to output image. Required new FILKCP subroutine for WaWa IO system. Moved nowhere. 12069. April 9, 2006 IMAGR Eric Major new option - to use a spectral index image to alter how bandwidth sysnthesis component subtraction is done. Changed: IMAGR Add FQTOL and IN3NAME and IN4NAME et al adverbs, create SPIX AND SPIXC object and check their existence, add to HI file info. IMAGR HELP file: Change description of IMAGRPRM(1) to mention standard antenna beam models for VLA and ATCA. Add IMAGRPRM(17) as indicator for spectral index correction and radius of an averaging circle when reading the spix image. Add IN3NAME, IN3CLASS, IN3SEQ, and IN3DISK to specify a spectral index image and IN4NAME, IN4CLASS, IN4SEQ, and IN4DISK to specify a spectral index curvature image. Add FQTOL to control how many spectral channels are done at one time when doing the primary beam and spectral index corrections. QUVUTIL Pick up SPIX object and call PBFSUB if there is a SPIX object and/or primary beam correction requested POSITION If another POSTION routine was called then PSNANG thought all was well when in fact the pointing position was not set. Changed all routines that call SETLOC for LOCNUM=1 to zero the pointing position and PSNANG to re-initialize if that position is zero or the class name is different. PBUTIL Changed PBFSUB and PBFDIV to pick up SPIX and FQTOL. The DO loop over channels was replaced with code to do as many channels together as FQTOL would suggest. If the array is VLA or ATCA, PBCALC is called to get the beam parameter numbers. PBFCCT and PBFSCI have new call sequences to receive the new parameters (PBFINC, PBPARM, SPIX) and to call SPVALU if SPIX is not blank. PBFACT has a new call sequence to receive the PBFINC, SINDEX, PBPARM parameters. It sums over the PBFINC channels, uses the PBCALC code if PBPARM(1) is positive, and applies the spectral index including curvature. The correction for spectral index is limited to the range 0.001 to 1000 to avoid gross overflows. The average value of the PBFINC channels is used compared to the average over all channels to set the scaling for the current set of channels. SPVALU is new to read the spectral index image(s) and get the value by interpolation or averaging over a circular aperture. SPIXR HELP file - removed some XMOM things, gave necessary help for OPCODE, and math formula for fitting. Moved nowhere. 12070. April 9, 2006 CPYRT Eric Added call to KEYCOP. Moved nowhere. 12071. April 10, 2006 IM2UV Eric Added alternative scaling options and made the default divide by the clean beam area if present in the image header. Fixed the handling of RA and DEC axes dropping the --SIN business which confuses UV routines. Moved nowhere. 12072. April 13-14, 2006 UVHGM Eric Added calibration and all data/calibration selection adverbs to the task. Changed allowed Stokes and made the maximum number of boxes larger. Added optional Gaussian fitting. Moved nowhere. 12073. April 14, 2006 UVHIM Eric New task: a variant on UVHGM to make an image of a 2-dimensional histogram. Axes may be vis real, imaginary, amplitude, phase, or weight or time or baseline u, v, w, length, or position angle. A choice of smoothing function is available for counting each sample on the grid. Moved nowhere. 12074. April 15, 2006 EXTLIST Eric Changed UVHGM to pass info re multiple subarrays and FQIDs. When a plot file was created on disk the header was destroyed (it used the output header from the cal routines not the real header of the data). Changed ANBPL to use a unique plot type number (32) and the correct number of adverbs (293 was 329) Changed AU8A to recognize new UVHGM format and do useful things with the new adverbs and find the old ones. Changed AU8A to recognize ANBPL and do something useful. Moved nowhere. 12075. April 16, 2006 DOCALIB Eric Changed all tasks to calibrate weights if DOCAL is > 0 and <= 99. One can still avoid calibrating weights but only by a deliberate setting of DOCAL to an odd value (100, 101 - actually 99.00001 to 101.0 - are allowed). Changed $APGNOT: ACFIT BLAVG CLIPM DECOR FLGIT LISTR RLDIF SHOUV SPCAL SPLAT SPLIT UV2TB UVFIT UVFND UVHIM UVHOL UVIMG UVMLN UVPRM UVPRT $APGOOP: FINDR FLAGR MAPBM RFI $QPGNOT: BLCAL BPASS CALIB CPASS DEFLG FRING HORUS KRING LPCAL PCAL $QPGOOP: BLING EDITR IMAGR OMFIT SCIMG SCMAP $YPGNOT: ANBPL CLPLT FRMAP FRPLT IBLED POSSM SPFLG TVFLG UVHGM UVPLT VPLOT WIPER Fixed copyright in AIPS.FOR to include 2006. Dropped all precursor remarks in JTRIM inside $AIPNOT/DDBGR Moved nowhere. 12076. April 18, 2006 STUFFR Eric New procedure and run file to stuff multiple-day observations into a compressed data set. Each data set is converted to hour angles rather than times, sorted into BT order, and concatenated with DBCON. Then UBAVG is run on the result. TI2HA needed to be fixed to drop the subarray code from the baseline ID when it drops it from the times. Moved nowhere. 12077. April 18, 2006 APCAL Amy Various changes: 1) Used to identify all single stokes as RR, fixed. 2) Added error recognition to opacity fits. If the receiver temp or tau0 was < 0 or greater than certain maximum values then croak and tell user to edit Tsys. 3) Added robustness to two of the fitting algorithms, GRID and OPAC, adding GRDR and OPCR. Moved nowhere. 12078. April 18, 2006 BPLOT Eric The X axis labeling when doing more than 1 IF confused LWPLA. Changed it to do all labels for 1 IF and then all for the next. Added a digit to the pixrange display at the bottom. Moved nowhere. 12079. April 20, 2006 VLBAUTIL, VLBATECR Amy There were some "background" adverbs in VLBATECR because I didn't default the verbs GETHEAD and GETTHEAD. Fixed. Minor changes: Added a TGET VLBACALA at the end of VLBACALA. Added some words about BCHAN and ECHAN to VLBAMPCL.HLP. Moved nowhere. 12080. April 23, 2006 Color contours Eric Added the ability to color each contour level separately: Changed: G3VCOL To write a plot command only if the color has actually changed. CONDRW New call sequence adds RGBLEV and changes DOCOLR to integer (0 no color, 1 use color already set, 2 use RGBLEV and BLKLEV). Corrected - dashed lines were not drawn in color previously when they should have been. Also changed to not draw black lines when the contours are colored. PCNTR Add adverbs and code to see if RGBLEV is ever non-zero. Changed call sequence to CONDRW. Corrected to force 3-color with B&W images if any of the other true color options are used. DODARK inhibited for contours if color contours are being drawn. KNTR As PCNTR. GREYS Change call sequence to CONDRW, dummy the color contour arguments. CNTR Change call sequence to CONDRW, dummy the color contour arguments. POPSDAT Add RGBLEVS and more scratch adverbs (II, JJ, KK, RR, BB, TT, XX, YY). NEWPARMS add RGBLEVS. RGBLEVS New help file SETRGBL New system RUN file: defines procs to set RGBLEVS by various prescriptions. CIRCLEVS New help file for new proc RAINLEVS New help file for new proc FLAMLEVS New help file for new proc STEPLEVS New help file for new proc Moved nowhere. 12081. April 24, 2006 PEELR Eric New RUN file and procedure to remove all image facets from a uv file. Then it loops over a list of facets - restoring the one facet (all resolutions), self-cal'ing, removing the model from the self-cal output, and undoing the self-cal SN table from the residual data. When the list is complete, it restores all facets to the output UV data. This should clean up the garbage around outlier, "interfering" sources. We may add options to image and repeatedly self-cal but that may not be necessary and will be complicated. Moved nowhere. 12082. April 26, 2006 STARPL Eric Skip call to PNTPLT on symbol types <= 0 - it plotted a type 1 in that case rather than nothing which we promised in the STARS help file Moved nowhere. 12083. April 26, 2006 PEELR Eric Add BCHAN and ECHAN for CALIB only. Moved nowhere. 12084. April 28, 2006 PEELR Eric A bad piece of logic made the looping over multiple fields not work. New logic should fly. Moved nowhere. 12085. April 29, 2006 UPDUPDATE Eric The make on XAS should not be done for binary jobs since the compilation is done at the AOC and may not work at the remote site. Moved nowhere. 12086. May 1, 2006 EDITA, EDITR Eric Changed QEDIUTIL to offer the choice of flagging the current source or all sources. Fixed it to avoid asking the flag number range on UNDO FLAGS when there are 0 or 1 flags currently. Got rid of extraneous message when listing 0 flags. Corrected nasty error that caused it sometimes not to find the nearest point when the user pointed exactly at the point. Moved nowhere. 12087. May 2, 2006 PEELR Eric Changed dimension of PPARM to 100 to allow more facets to be processed at once. The start and stop sections require all N facets to be UVSUBed while the inner loop deals with only 1 facet at a time and so is very much faster in many cases. Changed PEELR.001 and PEELR.HLP Moved nowhere. 12088. May 3, 2006 install.pl Eric Changed the method of finding Solaris memory size per a suggestion of Vincent McIntyre of CSIRO. Fixed the old code to test for the file's existence before opening it. It will not exist on non-Solaris machines. Also - the display of options does not update properly after the user enters compiler options - moved the SCREEN_10A point to force the update. Moved nowhere. 12089. May 3, 2006 IMEAN Eric Added a plot of the fit histogram on the requested histogram. Simplified the fitting code for this simple case. Reworded the DOHIST part of the help file. Moved nowhere. 12090. May 4, 2006 CCNTR Eric Forgot to change CCNTR for CONDRW call sequence. Moved nowhere. 12091. May 4, 2006 KNTR, PCNTR, EXTLIST Eric Corrected the number of parameters in these two tasks which have been 1 short for some time (and PCNTR added 180 while needing only 90 new ones). Changed AU8A to have EXTLIST know the number (corrected) of adverbs now and say something about color contours if they are used. Moved nowhere. 12092. May 4, 2006 PLGET Eric New verb to get the adverbs of a plot task from the plot file. Works with EXTLIST - list the PL files then plot them with TVPL and then get the KNTR (or whatever) adverbs that generated a plot file you like. Moved nowhere. 12093. May 5, 2006 stars Eric There is confusion on star type 0. Changed: STARPL Will plot a symbol for star type 0 if it does not plot a label string. Plots symbols for all positive star types and plots no symbol for all negative star types. STARS Changed to set star type 0 to -1 if there is a label string or +1 if there is not (also help file) GSTAR Changed to limit FACTOR to 1 - 24, fixed help file to list the current types. MF2ST Changed to limit types to 1-24 and do the scaling in the range. Updated help file. STFND Changed to limit star type to range 1-24 and changed the CPARM(10) parameter to make the user really request pixel coordinates (non-standard for ST files) Updated the help file for these. Moved nowhere. 12094. May 7, 2006 CALIB Eric Changed the reporting of individual closure errors at level 2 to also examine the expected error (1/sqrt(weight)) and only report those > 2.5 sigma. CLBPA Changed to recompute weights as needed after calls to GCALC et al. Changed the PRTLV passed to GCALC by 1. CLBSNR Changed to compute the average closure error with and without weighting. Changed to compute the expected closure error and report individually only those > 2.5 sigma (at level 2) or all > cutoff levels in PRTLV>=4. CALIB Changed help file to show the new print level meanings. UVUTIL recompute weights, set prtlev, handle error returns from GCALC as in CLBPA. Moved nowhere. 12095. May 8, 2006 CALIB Eric The print-level change of yesterday still does not allow proper flexibility. Changed CALIB to have CPARM(7) set the limit in sigmas for the display (0 -> 2.5). Changed CLBPA to receive and pass on the parameter handling the default. Changed CLBSNR to use the new RMSLIM cutoff. Changed UVUTIL to pass through 2.5 although the individual closure errors will not be reported in any case in the SCIMG/SCMAP self-cal. Changed the print level meanings back to what they were. Moved nowhere. 12096. May 9, 2006 SPLAT Eric Added adverb CHINC to be used with CHANNEL to select the channels averaged and output when APARM(1)=3. This allows keeping a multi-source file while keeping only, e.g. every other channel after Hanning smoothing. Moved nowhere. 12097. May 10, 2006 BLANK Eric Added opcode RADI to blank outside a radius in arc sec from the reference pixel. Moved nowhere. 12098. May 10, 2006 TVFLG Eric Added options LOAD NEXT CHAN and LOAD PREV CHAN to load the next or previous spectral channel using all of the other parameters now in use or set for the next LOAD. Moved nowhere. 12099. May 12, 2006 3C138_C model Amy/Eric Amy has completed another model. Moved nowhere. 12100. May 12, 2006 STUFFR Eric Changed STUFFR to sort the output data set back into TB order. Moved nowhere. 12101. May 14, 2006 delete boxes: verbs, IMAGR Eric Added two new verbs DELBOX and DFILEBOX in new subroutine AU5F with service routine GRBOXD. These allow interactive deletion of boxes from CLBOX and a boxfile, respectively. Added DELBOX as an option in IMAGR et al changing QCLEAN to offer the option (sets NBOXES(field) negative to call OTVBOX. Changed TVUTIL (comments only in OTVBOX) and TVDEVICE to handle a negative NBOXES as a call to GRBOXD. Changed AIPS.FOR to call AU5F and wrote new help files for DELBOX and DFILEBOX. Changed POPSDAT and NEWPARMS for new verbs. Moved nowhere. 12102. May 14, 2006 GETPOPSN Eric New verb - returns POPS number on stack. Chaned NEWPARMS, POPSDAT, AU8 and made new help file. Useful in procs. Moved nowhere. 12103, May 14, 2006 IMAGR Eric IMAGR and HLPCLEAN help files changed for DELBOX. Moved nowhere. 12104. May 15, 2006 PEELR, STUFFR Eric Added code to test for the intermediate and final file names and to quit if they already exist. Added info about those names to the help files. Fix the proc to get the proc's inputs back before exiting. Moved nowhere. 12105. May 16, 2006 MBDLY Eric A call sequence got messed up with an extra argument. Caused it to have bad antenna number, time, rate, source ID and source name. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12106. May 16, 2006 STUFFR, PEELR Eric Change VNUM to 33-GETPOPSN to keep the pops numbers from colliding. Moved nowhere. 12107. May 18, 2006 UVFLG, FLAGUP Amy UVFLG would ignore all other inputs but REASON when OPCODE was set to "REAS". So if a user wanted to unflag a certain reason, but only for one antenna, UVFLG would unflag that reason for all the antennas. I changed the subroutine FLAGUP so now it pays attention to other selections as well as the REASON. Moved nowhere. 12108. May 19, 2006 UVFLG Eric Upgraded FLAGUP a bit more following Amy's lead. Moved nowhere. 12109. May 20, 2006 MWFLT Eric Added OPCODEs MAX and MIN with a reference to Larry Rudnick's paper making interesting use of these functions. Fixed the history writing to record the window and weight parameters. Dropped the ill-considered option to change the header coordinates in arbitrary fashions. There are verbs to do that if the user insists. Also supplied defaults and checked that windows have odd number pixels. Moved nowhere. 12110. May 24,2006 $AIPSTARS/VLBA_UVCON Leonia The VLBA configuration file $AIPSTARS/VLBA_UVCON has had zero at the nineth column of each line. This column should be '1' because the antenna coordinates are given relatively the Earth center (see UVCON help). Otherwise the antenna elevations calculate wrong. Moved nowhere. 12111. May 24, 2006 APCAL Amy There was a bug that if the user set the initial TRec then TRec_max was never set. This then caused the program to die because of the new opacity solution reality checks added in April. Fixed. Moved nowhere. 12112. May 24, 2006 IMAGR Eric Added adverbs ANTENNAS and BASELINE to the task. They are applied in copying the data to the work file and are added to the history file. Moved nowhere. 12113. May 26, 2006 MWFLT Eric Larry Rudnick suggested an addition to the explain file on the MIN and MAX functions. Moved nowhere. 12114. May 26, 2006 CALIB Eric Changed help and code so that the output weights are not calibrated when APARM(10) > 99. Previously, they were calibrated except when DOCAL > 99. Moved nowhere. 12115. May 27, 2006 FLAGR Eric Changed it to allow any Stokes to be used for the flagging. DOSTOKES changes its meaning also. Moved nowhere. 12116. May 28, 2006 ANBPL, VPLOT Eric Changed both to set the axis scaling on the first plot of the page and retain that throughout the page. This removes the uncertainty of whether they are all Milli or Kilo or ??. Moved nowhere. 12117. May 28, 2006 SETSP, SETPAR Eric Dropped the UN-REGISTERED!!!! business and all the encoding and subterfuge. Moved nowhere. 12118. May 29, 2006 WIPER Eric Changed IAXIS1 to allow more axis types (9) like CLAB and to honor incoming LABTYPs when they are not recognized by lABINI. Changed ITICS to treat the new types 7-9 as time axes. Changed WIPER to allow STOKES='HALF' which forces all sorts of things to loop over polarization where only one polarization was allowed previously. Changed axis labeling leaving IAXIS1 to handle the metric scaling and making time in hours (type 11) be properly supported. Changed AU5B and TVDEVICE to set LABTYP=0 to get previous behavior with IAXIS1. Moved nowhere. 12119. May 29, 2006 MBDLY Ed/Eric Ed Fomalont made some suggestions for the help and explain file. Moved nowhere. 12120. May 29, 2006 CLCAL Eric Changed SN2CL to use the the WWTn weights (limited to -0.5 to 1.5) rather than the (time-time1)/(time2-time1) which can be very large. The 2PT method used the weights but several of the methods used the times for phases while using the limited weights for amplitudes. The CUBE method cannot be limited in this way. Added warnings about that to the CLCAL help file. Moved nowhere. 12121. May 30, 2006 VLOG Eric JIVE asked us to enlarge the L-band range to encompass red-shifted HI. Changed 1.35 to 1.20 GHz in the table in subroutine VLCAL. Moved nowhere. 12122. May 30, 2006 FITLD Eric Changed SOINI - the VLBA correlator had another column name spelled wrong (2 in one table!). The source number columns is supposed to be 'SOURCE_ID' not the VLBA's 'ID_NO.' Added a second synonym to the routine (along with EPOCH for EQUINOX). Moved nowhere. 12123. May 30, 2006 IMPOS, IMVAL, MAXFIT Eric Change AU5 and AU9 to return all parts of COORDINA as negative when the value is negative. This meets the definition used when COORDINA is used and avoids the problem of -00 degrees or minutes. Moved nowhere. 12124. May 30, 2006 IMAGR Eric Changed PBUTIL to read the frequency from the spectral index image header and use it as the base frequency. Images of pure spix do not care but when curvature is applied then the frequency reference is critical. Improved inputs section of IMAGR.HLP re IMAGRP(17). Moved nowhere. 12125. May 31, 2006 semi-annual update Eric Updated LSAPROPO since it no longer contains many of the task etc. name listings. Fixed HLPB.FOR so this should not re-occur. Updated the about help files: ZZANALYSIS ZZAP ZZCALIBRAT ZZEDITING ZZIMAGE ZZIMAGING ZZINTERACT ZZOOP ZZPLOT ZZPOPS ZZPROCEDUR ZZTABLE ZZTASK ZZTV-APPL ZZTV ZZUV ZZVERB ZZVLBI Updated the tab completion file HLPIT.LIS. Moved nowhere. 12126. June 1, 2006 plotting Eric CONDRW did not do negative color contours correctly. The INPUTS files for STEPLEVS, RAINLEVS, FLAMLEVS, CIRCLEVS were not typed correctly. Movwed nowhere. 12127. June 4, 2006 Plotting Eric New functions: GCOMNT Places a type 18 comment command in a plot file GCHDRW Draws A-Z, 0-9, plus some special characters using line-drawing commands (GPOS, GVEC, G3VEC) TXRGBL Specialized routine for drawing a LEVS= line in a plot file using GCHDRW. Changed STARPL To call GCOMNT ahead of the stars plots KNTR To use TXRGBL when appropriate, leaving out the usual LEVS= lines. Also to use GCOMNT. PCNTR To use TXRGBL when appropriate, leaving out the usual LEVS= lines. Also to use GCOMNT. TVPL To step over comment function data. TKPL To step over comment function data. TXPL To step over comment function data. LWPLA To put the comment function information as a comment in the output PostScript file. CNTR Calls to GCOMNT CCNTR Calls to GCOMNT GREYS Calls to GCOMNT IMEAN Calls to GCOMNT SL2PL Calls to GCOMNT ANBPL Calls to GCOMNT. Corrected scaling of phase when both phase and amp are plotted. Fixed failure to handle all 0.0 phases. VPLOT Calls to GCOMNT. Corrected scaling of phase when both phase and amp are plotted. Moved nowhere. 12128. June 5, 2006 EDITR, EDITA Eric QEDIUTIL had a typo. Help files were not updated on May 1 (#12086): EDITR, EDITA, SNEDT, HLPEDITS, HLPEDITY, HLPEDIUV, HLPEDICL, HLPEDISN. Moved nowhere. 12129. June 7, 2006 UVFLG Eric The OPCODE 'REAS' has no use and is hereby removed from the help. Moved nowhere. 12130. June 8, 2006 CookBook Eric Changed: 3 Add a sentence re PLGET next to the TGET stuff. 4 Added note on WIPER/FLAGR, rewording UFLG in UVFLG, add new options to TVFLG (but not figure), fix DOCALIB wording (it said DOCALIB=TRUE several places which ddid not cal the weights!), CALIB suppression of closure errors when not significant, EDITA new figure, models have added words and a new figure, paragraph on ANBPL, reword VLARUN. 5 STUFFR paragraph, PEELR section, IMAGR beam size, IMAGR and verb DELBOX, verb DFILEBOX, section on spectral index option in IMAGR, update EDITR options, CALIB display options, mention CCRES. 6 Add two new color plot pages (KNTR with RGBLEVS and SNPLT). TVCPS multiple images, SNEDT does TY, add ANBPL sentences, Add sentences for BPLOT, upgrade UVHGM and add UVHIM, update IMEAN including new figure, PLGET, DELBOX, DFILEBOX. 7 Add CCRES section and info re new BLANK and MWFTT operations 9 Emphasize parallactic angle correction for phase referencing, reword VBGLU, add VBMRG, reword SNEDT (some words are now true that were wrong before), reword UVFLG UFLG op (also wrong before), enhance APCAL description 12 Add mention of GETPOPSN. 13 Update for changed ABOUT files (12125). C Refer to HSA, upgrade parallactic angle correction D DOCALIB meaning shifted 0 As needed I, J As needed Recipe Consequent changes Moved nowhere. 12131. June 13, 2006 DEFAULT Eric Fixed AU2A for using the wrong loop limit when looking for special adverb names. On AMDs with the Intel compiler, some adverbs would be found this way and would take bad values. Moved nowhere. 12132. June 14, 2006 system stuff Eric Fixed the $SYSLINUX, $SYSMAXIN, $SYSMACPP versions of LIBR.DAT - they all referred to the system library too many time in places. Changed the CCOPTS.SH too for SYSMACIN to be ordered more normally. Some of the others retain COMP in a funny place. Moved nowhere. 12133. June 14, 2006 KNTR, PCNTR Eric Fixed them to plot the color RGBLEVS only if contours are being drawn. Moved nowhere. 12134. June 19, 2006 AIPSLetter Eric Put LET06A into $AIPSPUBL in TEX, PDF, and pS forms. Moved to ftpp site this date. 12135. June 24, 2006 ELINT Leonia ELINT has been died at the case when the last plot data are flagged. Actually I found that ELINT has a problem even when the bad data do not correspond to the last plot (at Solaris) although Rick has not seen this problem at his Linux box. Anyway the problem is fixed. Moved nowhere. 12136. June 27, 2006 DELZN Leonia The following modifications/corrections have been done. 1. RMS of the residuals actually was printed as the variance. SQRT of the variance is added to calculate RMS correctely The units of the RMS have been added 2. The having existed limit 10000 of the output file rows increased to 100000. 3. The plots of residuals have been added under control of APARM(9) 4. The option of the solution of clock only has been added (APARM(2)=0) Moved nowhere. 12137. June 27, 2006 ELINT Leonia ELINT failed with the new RICK's data because the OT table size exceeds the size of the relevant arrays in the ELINT. Increase the array sizes from 1000 to 10000 Moved nowhere. 12138. June 27, 2006 POPSDAT Eric Fixed error id TVCH=0 on input. Moved nowhere. 12139. June 29, 2006 DELZN Leonia I've found logical errors in connection of create/not create CL table <=> use CL time or even time <=+ create/not create the output file. The errors have been fixed. Moved nowhere. 12140. July 3, 2006 VLARUN Lorant Had an omission in running checkids in VLARUN (to correct for fast-switching source names automatically using DSORC). Moved nowhere. 12141. July 6, 2006 ELINT Leonia Rick Perley requested to add prints of the polynomial fitted to the data into plots. It is done. Moved nowhere. 12142. July 13, 2006 WTMOD Eric Added NOISE to scale each weight by NOISE(ant1) * NOISE(ant2). Moved nowhere. 12143. July 13, 2006 AIPSASSN Eric changed CV to have $DA01/FITS for the fits area. Adde PUBFITS to point at the older area. Moved to CV this date. 12144. July 13, 2006 UVPLT Eric Added option to plot one or bath axes in opposite order. Enter the code with a minus sign (BPARM(1) and BPARM(2)) and the plot order will reverse. Moved nowhere. 12145. July 14, 2006 VLAPROCS Amy Added to VLAPROCS: VLATECR - similar to VLBATECR, calculates day number and automatically downloads IONEX files and runs TECOR. VLALIST - runs LISTR a'la VLACALIB Calibrator models - to VLACALIB, if the flux calibrator is the only CALSOUR and UVRA and ANTE are 0 then VLACALIB now attempts to automatically download the flux calibrator models and uses them. VLAPROCS.HLP - created. Was disturbed to see that this didn't exist. I am sure there are users who if they hear of VLAPROCS try a HELP VLAPROCS. This now tells them about what VLAPROCS contains and how to load it. Moved nowhere. 12146. July 14, 2006 CALRD.HLP Amy Removed list of models, as it was outdated and added a reference to CALDIR. Moved nowhere. 12147. July 14, 2006 EXTLIST Eric Fixed several errors in the ISPEC section of AU8A. moved nowhere. 12148. July 18, 2006 VLAPROCS Amy Fixed bug I introduced in which there were two VLADUMMY procs to declare variables. Moved nowhere. 12149. July 19, 2006 UVPLT, WIPER Eric Made u axis be +u at left/bottom unless -6 type requested. Added the axis reversal code from UVPLT to WIPER. Moved nowhere. 12150. August 1, 2006 IMCENTER Eric New verb: finds centroid in a sub-image. Changed POPSDAT.HLP, AU9.FOR, WHATSNEW.001, and wrote a new IMCENTER.HLP. Moved nowhere. 12151. August 4, 2006 HELP Eric HELP a specific type (i.e. a list of VERBS or PROCS or ..) did not close the printer in AU1A. This then left LUN 3 open in AIPS for all later actions. Moved nowhere. 12152. August 7, 2006 VBGLU Eric The routine that figured out the order of the UV data times between the 2-4 data sets was very slow in some circumstances. Added tricks to speed it up. The table glueing routine did not take into account the different numbers of columns when there is 1 Stokes or more. Also the pointer used for the desired record number was the wrong variable. It is hard to see how this ever worked. I suppose fixed buffers were used in almost all the debugging to get around limitations in the debugger. Moved nowhere. 12153. August 9, 2006 More AT file follies Eric Changed ATINI, ATTAB to correctly handle NOPCAL = 0 which is one of its 2 valid values. The code assumed 2 but did 0 when allocating space. Changed VBGLU to use 0 for the length of columns rather than forcing 1. Moved nowhere. 12154. August 10, 2006 TV load functions Eric The LG, NG function used was very extreme (LOG (1 + 6500*X)) with X between 0 and 1. A value of 9 is very pleasing and 99 is extreme enough. Changed ISCALE To use 9 with LG, NG or 99 with L2, N2 codes GSCALE To use 9 with LG, NG or 99 with L2, N2 codes TVALUE To use 9 with LG/NG, 99 with L2/N2 codes PCNTR Added L2 and N2 to allowed function types. KNTR Added L2 and N2 to allowed function types. TVRGB Added menu items to switch from linear, to square root, to the new LG to L2 to linear in each color. Changed output routine to honor the functions as well. TVHUI Modified LOG/SQ transfer menu item to change order and add stronger log function. Changed the log functions from 65000 (!!) to 9 and 99. RGBMP Added L2 function and changed LG to 9 instead of 6500. Help files: TVRGB mention new options HLPTVRGB run-time help for each TVHUI change the description of transfer functions HLPTVHUI run-time help reworded and new option added RGBMP changed mathematical descriptions, added L2 FUNCTYPE added math and new L2 N2 codes TVLOD added L2/N2 - also helps for COTVLOD, TVALL, TVCUBE, TVMOVIE, and TVROAM. Moved nowhere. 12155. August 10, 2006 S/X split in procedures Lorant Changed VLBAARCH and VLBAUTIL (ie VLBAFQS and VLBAFIX) to check full tunable range of frequencies instead of nominal frequencies so projects observing outside the default band get recognized and split in separate S and X as usual. (90/50 was set fine). Moved nowhere. 12156. August 10, 2006 more help Eric Changed FUNCTYPE in GREYS, KNTR, PCNTR, PLAYR helps too. Moved nowhere. 12157. August 11, 2006 SDCLN Eric The SDI section was limited to images 4096 on a side. Changed arrays to MAXIMG. moved nowhere. 12158. August 14, 2006 QUACK, AN10RS =>UVFLG,UVCOP Leonia Add BASELINE to the QUACK inputs. The subroutine AN10RS is used to have ANTENNAS, BASELINES as inputs. A small error has been found in AN10RS when both ANTENNAS and BASELINES .GT.0. It fixed and UVCOP, UVFLG compiled with the corrected AN10RS. Moved nowhere. 12159. August 15, 2006 plotting Eric Changed: WINDOW Changed error message to have more digits (2 lines) GCOMNT The computation of needed space was wrong which could damage things badly. GREYS Changed the image alignment code. It did WINDOW on all before recomputing the windows anyway and so could fail when the recomputed ones would be fine. Made a work buffer MAPBFS rarther than 2048 in size. KNTR Same sub-image trouble as GREYS. PCNTR was fine however. Needed extra work to handle plotting a cube on top of a plane. Moved nowhere. 12160. August 16, 2006 AN10RS=>QUACK, UVFLG, UVCOP Leonia A mistake is found in yesterday AN10RS correction. The QUACK, UVFLG, and UVCOP COMLNKed with the corrected AN10RS. Moved nowhere. 12161. August 16, 2006 more plot Eric All the routines that used GCOMNT used MSGTXT as the call argument. That is in COMMON and is set = ' ' by ZFIO so it cannot be used in plot tasks that might call ZFIO when the buffer is full. Changed STARPL, ANBPL, VPLOT, CCNTR, CNTR, GREYS, IMEAN, KNTR, PCNTR, and SL2PL. Moved nowhere. 12162. August 19, 2006 POSSM yet again Eric The axis labeling in velocity when asking for a reversal of the axis was off by 1 channel. The printed numbers were fine in the output text file but not on the plot. Added option FACTOR to scale the plus signs used for phase and added plus signs to the connected lines for other types. FACTOR < 0 turns off the plus signs in those cases. Added option XYRATIO to control the XY ratio which did not look as I wanted in one of my plots. Changed AU8A to recognize new nparms of possm. Moved nowhere. 12163. August 21, 2006 VBMRG.HLP Amy Added basic recipe for use (from Jon Romney) to end. I got a phone call asking about merging multiple passes and it would have been useful to just point the person to EXPLAIN VBMRG. Most of the information was present in the HELP file but the recipe just makes it clearer. Moved nowhere. 12164. August 22, 2006 KNTR, RGBLEVS Eric KNTR: Fixed positioning of RGBLEVS line when there is a step wedge at the top. Added comments at each contour level so that the user can find each level at the places (plural usually but not too many) where it occurs. KNTR: Added comment about contour level comments to the help file. RGBLEVS: created a help file. Moved nowhere. 12165. August 23, 2006 CCEDT Eric Changed the automatic boxing code a bit more - removing an infinite loop situation occasioned by the old code combined with the newer code which reduced the amount that DX and DY are adjusted in each cycle. Dropped error messages for things that are really not errors. Note however that CCEDT automatic code gets different answers than it used to. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12166. August 23, 2006 IBLED Eric Found an error in which the window limits were computed as if IYP(1) < IYP(2) immediately following a computation that had them opposite! This caused error bars on phase plots to produce TV errors killing the task. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12167. August 25, 2006 VLBATECR & VLATECR Amy Added additional file in case jplg version does not exist. Moved nowhere. 12168. September 1, 2006 FILLM Eric History for CPARM(2) was not specific. Added code that should be more explicit. Moved nowhere. 12169. September 25, 2006 VLACALIB, CLIPM.HLP Amy VLACAL - Changed the FLAGVER so that 0 means highest flag table version. CLIPM - Fixed typo on help file. Moved nowhere. 12170. September 26, 2006 QUACK.FOR .HLP Leonia Change default of FLGVER from 1 to the highest existed FG. Moved nowhere. 12171. September 27, 2006 CVEL Eric Called CALINI with 4 arguments being the same variable. this caused the first of them (# polarizations) to be set to other values than the correct one, leading to error messages from CALINI that should not occur. Moved nowhere. 12172. September 29, 2006 UVCON Leonia UVCON has never simulated correction for the primary beam, if the model is given by SMODEL or clean components are not point. Taking into account, that such a simulation does not have a big sense or happens very rare and requires a very serious modification of UVCON, I came to the decision to add couple of warnings only. Moved nowhere. 12173. September 29, 2006 APCAL Amy APCAL didn't handle the situation where there was only one LL stokes correctly. Fixed. Moved nowhere. 12174. October 2, 2006 SETFC Lorant Changed it to leave BPARM(1) < 0 rather than resetting it to zero. This leaves an inner region to be ignored for NVSS sources and the like. Moved nowhere. 12175. October 2, 2006 FILLM Bryan Add flag to include tipping scans in the data written to disk. Moved nowhere. 12176. October 2, 2006 GAL Eric Cleaned up the typing of GAL.HLP and changed code and help to use logicals in standard ways (> 0 -> true). Added SYMBOL adverb, changed DPARM(9) to CPARM(7) to simply plot rather than fit. Made DPARM(9) the min velocity to plot. Moved nowhere. 12177. October 3, 2006 INSEQ limits Amy Changed INSEQ limits to 0-9999.0 (as in most other programs) in PUTVALUE, IMVAL, QIMVAL, TABGET and TABPUT. Moved nowhere. 12178. October 5. 2006 UVPLT Leonia Antenna parameters have not picked up if Azimuth vs Time is desired to plot. As a result UVPLT has crashed at this case. Fixed. Additionally, the plot header is changed. The antenna=refant is printed out at the plot header, if the plot type (14,15,16,18) is related to antenna (not base line) Moved nowhere. 12179. October 5, 2006 SPIXR Eric Bad error branch 2 places brached to a different loop. (Code was copied and not corrected.) Moved nowhere, should go to patch. 12180. October 6, 2006 SNSMO Eric Confused RA2KOL > 0 with NUMPOL > 1 so thatrates would smooth and clip with invalid data from wromg locations. The change to SNINI in TST caused this to happen at IF 2 rather than starting with IF 3. Moved nowhere, should go to patch. 12181. October 7, 2006 INTERPOL Eric Adverb help file was out of date. Moved nowhere. 12182. October 7, 2006 POSSM Eric Pickup of station ID for VLA out of date (used 5:7 not 6:8). Moved nowhere. 12183. October 9, 2006 FLAGUP, UVFLG Amy UVFLG (in external subroutine FLAGUP) did not check on TIMERANGE or BIF/EIF while unflagging. Fixed. Moved nowhere. 12184. October 12, 2006 VLBAUTIL, VLBAFIX Amy Added calibration table merging to VLBAFIX. VLBAFIX now tests to see if merging is necessary and then runs VLBAMCAL if necessary. Added comments about this change to the VLBAFIX and VLBAMCAL help files. Also simplified how VLBALOAD did table merging. Moved nowhere. 12185. October 12, 2006 ZZPROCEDUR.HLP, ZZVLBI.HLP Amy Fixed the entries describing what VLBAUTIL procedures did. Moved nowhere. 12186. October 23, 2006 SPFLG Eric The SPFLGR file sequence number was set to 1+previous max if the requested one did not exist. The user's requested value will now be used to create the work file. Fix length of list of antennas being gridded. Moved nowhere. 12187. October 23, 2006 UVGET, SETFC Eric Changed UVGET to complain about FRQSEL=-1 with a real FQ random parameter only if the is more than one FQ. Otherwise set it to 1. SETFC had a branch to stop without the resume aips call. Moved nowhere. 12188. October 23, 2006 IMAGR Eric Added optional display of an inscribed circle using adverb GRCHAN. Changed IMAGR Fortran and help, QCLEAN, and TVUTIL. Moved nowhere. 12189. October 23, 2006 VPLOT Eric Changed it to allow 2 separate polarizations in the same plot with or without separate coloring. Moved nowhere. 12190. October 24, 2006 VPLOT Eric Above change was missing a call sequence argument for doing time averaged data. Moved nowhere. 12191. October 24, 2006 IMAGR Eric The details in IMAGR.FOR and TVUTIL.FOR did not actually make it to the official areas somehow. Corrected that oversight. Moved nowhere. 12192. October 24, 2006 LOCIT Eric Changed code that locates columns in the SN table to allow for only one polarization. The later code did that okay, but not the startup code. Moved nowhere. 12193. October 29, 2006 MCAT Eric Changed CATLST to alphabetize up to 5120 catalog entries per pass. The old limit was 1536 - too small for those cases really needing the option. Moved nowhere. 12194. October 30, 2006 UVAVG Leonia The new OPCODE='SUBT' has been added. This is a special option requested by Frazer Owen (August 2006) to remove a coherent output happening as a result of mutual coupling between an array antennas (GMRT at the Frazer's case) or of other unknown effects. If OPCODE='SUBT', then the task works with two passes. At the first pass the averaging through the time of the whole experiment for each baseline, IF, Stokes, channel is carried out. At the second pass the task subtracts the found average values and then the standard averaging of the difference for the average time given at the input parameter YINC is carried out. Moved nowhere. 12195. November 3, 2006 UPDCONFIG Eric Removed the CFA site so that they can control their own destiny. Moved nowhere. 12196. November 4, 2006 WTMOD Eric If the SUBARRAY being modified did not occur in the first record of the UV data set, then the parameters did not get initialized and the reulting weights became 0. Moved nowhere. 12197. November 6, 2006 IMAGR Eric Changed IMAGR.FOR to survive (and blank) missing channels. Moved nowhere. 12198. November 6, 2006 IMEAN Eric Changed help file in an attempt to clear up confusion between the plotted histogram and that one used to find the noise. Moved nowhere. 12199. November 6, 2006 BPASS Eric The dynamic memory allocation and the usage of buffers was different when only some IFs were processed. This led to errors in normalization and an abort at some point due to dynamic memory being compromised. Also changed it to start with a (1,0) guess every time the channel number is 1. Otherwise it carried over the solution for the last channel, IF, polarization to start the next spectrum. Moved nowhere. 12200. November 7, 2006 IMAGR Eric Added a close that was forgotten when there are no points in a channel. OOP bypassed the open LUN but a table routine when merging the next channel's CCs did not. Moved nowhere. 12201. November 7, 2006 LTESS Eric Changed APLPBI To do straight copy (no beam) when BMSIZE < 0 or BEAMTP <= 0 also. It is only called by VTESS etc in the mode where BMSIZE and BEAMTP > 0 so this should not matter to it. LTESS To use APLPBI in all cases. this will mean that magic blanks in the input image will be honored and will not count in the sum of weights images. Previously every pixel counted in the sum of weights and blanked pixels used that magic numeric value as real on non-interferometric images. Moved nowhere. 12202. November 7, 2006 BPASS Eric Messed up the logic for when channel 1 is flagged. Should reset INIT to TRUE if channel is 1 but do not set it FALSE except in GCALC if it gets an answer. Moved nowhere. 12203. November 11, 2006 PCAL Eric PCAL set INSEQ of 0 to 1, contrary to the help file. It is now left at 0 which TABINI will set to the max. The routine that called FACSET was not told the IF and so always set the factor for IF 1. Added that to the call sequence and the call to FACSET. For some reason the source-table fluxes were set to zero when using a model although no place later summed the model back into VFLUX. The resulting VFLUX was put into the source table, reoving the Ipol flux that was there. Commented out the setting to zero. Moved nowhere. 12204. November 11, 2006 FACSET Eric Changed final message to be abs value of factor. The sign has to do with subtract vs add and is confusing to the user. Moved nowhere. 12205. November 13, 2006 FLATN Eric Changed it to allow the users to specify no primary beam correction in the PBPARMs. Added keywords to specify the coordinate type, increment, and reference pixel of the output image. Sensible defaults which should be used most of the time unless one knows well what one is doing. Changed POPSDAT and NEWPARMS to define new adverb COOTYPE and did a help file for it. Moved nowhere. 12206. November 21, 2006 help files Eric Updated the tab completion file HLPIT.LIS and the ABOUT files for ADVERB, ANALYSIS, CALIBRATION, PLOT, PROCEDURE, RUN, UTILITY, VERB, VLA, VLBI. Moved nowhere. 12207. November 22, 2006 VBGLU Amy Added merging of GC and IM tables. Changed the handling of the CT table so that it does not get copied if the CT tables on the files to be merged are not identical. Added paragraph to the help file about this. Moved nowhere. 12208. November 27, 2006 FLATN Eric Changed INTWIN in INTERPLATE.FOR to check to see if any of the test points are in the output image and return -1 if not. It now returns -2 if there are no good pixel coordinates. Changed FLATN to warn if the 3D option is selected and to tell about the inclusion of each image in the output. The output window is given when included and the reason (outside or geometry) when not included. Moved nowhere. 12209. November 27, 2006 VLBATECR Amy The DAYNUM subroutine was not calculating the correct day number for experiments that took place in August. Fixed. Moved nowhere. 12210. November 28, 2006 plots Eric Changed GVEC, GDVEC, G3VEC to do a roundoff when truncating vectors at the boundaries of the TV and to fill the buffer further checking the edges. LWPLA needed this one place too. Changed PCNTR and KNTR to do the correct Jy per arc sec display - it was a factor 2 too large for per pixel displays which KNTR would do on LTYPE not 3. Dropped the test on axis type since the polarization lines cannot be drawn with the wrong type. And the type is affected by LTYPE in just this spot in KNTR. Corrected FACTOR description in the 2 help files. Moved nowhere. 12211. November 28, 2006 FLATN Eric Found a naming problem for facets > 999 and an overwriting of memory for NMAPS > 64. Moved nowhere. 12212. November 29, 2006 FLATN Eric Add ROTATE adverb to set the image rotation rather than depend on the rotation in the first image. Note that this has no default; zero is meaningful. Moved nowhere. 12213. November 30, 2006 UVHGM Eric LUN 16 is now used by the calibration routines and so is a poor choice for the main data set. Changed it to something better. Moved nowhere. 12214. November 30, 2006 TVRGB Eric Messed up the enhancement functions when I added the 3 transfer function operations. It was depending on array indices which changed when the new ops were added. Moved nowhere. 12215. November 30, 2006 gridded subtraction Eric Changed QMTYP, GRDSET, and GRDAT to interpolate grids that are up to 4096. Previously the limit was 2048. Changed FORCED to recommend DFT for larger images and issue a message (once) in so doing. Moved nowhere. 12216. November 30, 2006 VLBAUTIL help files Amy Fixed the description and keyword lines in the help files for VLBACPOL, VLBACRPL, VLBAEOPS, VLBAFIX, VLBAFPOL, VLBAFQS, VLBAFRGP, VLBAFRNG, VLBAKRGP, VLBAKRNG, VLBALOAD, VLBAMCAL, VLBAMPCL, VLBAPANG, VLBAPCOR, VLBASNPL, VLBASRT, VLBASUBS, VLBASUMM and VLBATECR Moved nowhere. 12217. November 30, 2006 PEELR Eric Changed run file to default APARM(7) to 2 and APARM(9) to 1. If they are zero when read, they are set to these values. Otherwise they keep the values set by the user. Changed the help file to point out this difference from the usual CALIB behavior and to add the adverb WEIGHTIT. All of these changes are to aid in getting solutions without deleting serious quantities of the UV data. Moved nowhere. 12218. November 30, 2006 ABOUT files Eric Ran SHOPH for about files for Amy's changes. Updated ADVERB AP CALIBRAT CATALOG FITS PLOT POLARIZA PROCEDUR TABLE TASK UTILITY VLBI Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC06 is NEW *********************************************************************** *********************************************************************** 12224. December 5, 2006 IMAGR Eric Changed formats in PBUTIL - it was using I2 for field number. This stopped the frequency-dependent model subtraction if there were more than 99 fields. Moved from TST this date. 12225. December 8, 2006 3C138_L.MODEL Amy/Eric Put back Amy's latest calibrator source model. Moves from TST etc automatically. 12226. December 13, 2006 VLATECR Amy Fixed the bug in DAYNUM subroutine that VLATECR uses, same ask the fix to VLBATECR makde on Nov. 27, 2006. Moved from 31DEC07. 12227. December 13, 2006 install.pl Eric Added rsync of the $HIST area which was overlooked. Moved correction from 31DEC07. 12228. December 14-15, 2006 Flagged CCs Eric Changed GRDCRM to ignore flagged CC lines rather than die on them and changed FACSET to ignore them rather than use them. Moved correction from 31DEC07. 12230. December 16, 2006 SPLIT Eric Changed default of CHINC to NCHAVG. Moved from 31DEC07 this date. 12231. December 18, 2006 CLCAL Eric The phases of IFs 2-N were being changed for smoothing done on delays of those IFs as Delta(phase) = -2pi * (F(ifn)-F(if1)) * Delta (delay of ifn). This code was commented out as well as a non-trivial amount of supporting code. Moved from 31DEC07 this date. 12232. December 19, 2006 FILLM Eric FILLM was changed to parse the IF control bits word to see if the data have been scaled by the nominal sensitivity or not. All data for format numbers 24 and less are assumed to have been scaled and all recent data should be scaled. However, new data from the E/VLA starting in 2007 will not be scaled and so will need scaling by FILLM. FILLM checks the bit and makes sure that it is the same for all IFs and antennas and scales any data that have not been scaled. Moved to 31DEC06 to support 2007 VLA data. 12233. December 21, 2006 FLAGR Eric Fix it to flag singletons when it is supposed to do so. Moved from 31DEC07 this date. 12234. December 21, 2006 AVSPC Eric Correct alternate reference pixel location for frequency averaging. Moved from 31DEC07 this date. 12235. December 21, 2006 UVGLU Eric The test on correctness of frequencies was (F1-F2) / (DeltaF) > 0.1 This has to be absolute value > 0.1. Moved from 31DEC07 this date. ========================================================================= PATCHES ========================================================================= 12241. January 2, 2007 CALIB Eric Discovered an error in which the NX table times were extended by nearly 5 seconds in each direction to compute the integration start and stop times. This does not matter when doing scan averages, but does when a true SOLINT is set. An extension of 0.1 sec if probably okay. Moved to 31DEC06 patch this date. 12245. January 10, 2007 UPDCONTROL Eric Amazingly, this proc still used the old sort syntax. We finally hit a machine that refused it visibly. Since SUN4 is no longer a current OS, simply commented out the old syntax and replace the lines with the new. I think the operation that it was to do is not significant. Moved to 31DEC06 patch this date. 12265. February 9, 2007 LPSEC, UVFIX Leonia The leap second subroutine LPSEC is updated until the last correction at Jan 2006, and the task UVFIX (only one using LPSEC) is comlnked. Moved nowhere, to patch 2/27. 12268. February 20, 2007 CLCOR Leonia OPCODE='SUND' has not worked probably since the time when OPCODE='EOP' had been installed. The GO TO 999 was missed Moved nowhere, to patch 2/27. 12272. February 26, 2007 VBGLU Amy Fixed merging of the STATE count column of in the PC table. It used to not expand to the correct number of rows because it had twice the number of row as all the other 2D columns. Moved nowhere, to patch 2/27. 12274. February 27, 2007 CVEL Eric The final call to DSHIFT had the wrong number of arguments causing the task to abort (with the GNU compiler anyway) at the close down stage (before extension file copies). Moved nowhere, moved to patch this datep. 12283. March 14, 2007 SNSMO Amy There was an error introduced last year in the call sequence to the subroutine SMORAT. This caused odd smoothing of the rates for the first IF, including flagging solutions. Fixed. Moved nowhere, went to patch April 26. 12289. March 21, 2007 UVFIX Leonia UVFIX has not calculated REFCHN for uncompressed data. It has provided mistake at the frequency calculation when source shift was involved. Moved nowhere, went to patch April 26. 12291. March 21, 2007 CLCOR Eric The source position correction updates the SU table first so that the source positions taken as the "input" positions were really the "output" positions. This makes a second-order error in the correction. the SOUMOD routine had a 3rd order error in that it used the output apparent declination to scale the DX term when it should be the input apparent declination. Moved nowhere, went to patch April 26. 12321. April 26, 2007 BLCAL, UVFND Eric The cal system has an integration time parameter DXTIME which is the integration time in DAYS. If it is not set, then the random parameter is consulted or 1 sec taken. It controls the gain correction made for averaging when there is a delay and rate. A 1 day integration makes for a large correction! Two tasks, BLCAL and UVFND, set this parameter to 1.0. Changed to set it 0.0. Moved DXTIME correction to patch this date. 12331. May 4, 2007 Man pages Eric The man pages for aips got lost. The Mac systems did not allow us to have both aips.l and AIPS.l in the same directory, so I deleted AIPS.l. Unfortunately AIPS.l had all the text and aips.l was just a link to AIPS.L. Fortunately, we have an antique version of aips still on disk and I was able to recover to file, now stored as aips.l. man AIPS will not work but man aips does. Moved to patch this date. 12342. May 24, 2007 FILLM Eric There was an extra call to MRECO (read routine) when an EOF occurred that caused it to skip the first data record in the next file. This led to real problems when the next file was incompatible with the previous, but also caused it to lose a record when the two were compatible. Moved this fix (not other FILLM changes) to patch this date. 12349. May 30, 2007 SNPLT Eric The Hour angle computation had a mysterious -PI/2 causing all hour angles plotted to be too low by 6 hours. Moved to patch June 14, nowhere else. 12357. June 10, 2007 SPLIT, SPLAT Eric Errors occurred when TIMERANG limited the times and SOURCES included more sources than were in the TIMERANGe. Changed: SPLIT If UVGET ('INIT' returned an error code, DOCAL et al were left in the temporary (false) state for fetching the new source. Moved SPLIT fix simply to patch, nowhere else. 12364. June 14, 2007 SNPLT Eric The inner plotting routine changed the phas code for PC tables to a special value and then did not test that special value when checking for phase wraps. So lots of points potentially did not get plotted. Moved to patch this date. 12365. June 16, 2007 VLA station names Eric There has been a protest over the use of a blank in the VLA station names so I have changed it to an _. Let us hope that this does not mess up labeling in AIPS where _ is changed to an unshrinkable blank on occasion in string handling for display. Changed FILLM to do the _ and changed PRTAN, VLANT, LOCIT, FRPLT, and POSSM to support all possible VLA station naming conventions. FRPLT especially but also POSSM did not handle the previous conventions entirely correctly. FILLM did not test the possibilities of names properly a couple of other places. Moved to patch (except VLANT) this date. 12373. June 22, 2007 POSSM, FRPLT Eric There were typos in the display string handling due to the VLA station name change. The string length was too short to handle non VLA names as well. Moved the display fix to patch. 12436. September 4, 2007 BLAVG Amy In weighted average the visabilities were being multiplied by the weight twice, so when the weights were calibrated and became very small the average vis became nearly zero. Fixed. Moved to patch Sept 5. 12445. September 21, 2007 FILLM Eric Changed response to nominal sensitivites of 0.0. Previously, 0.333 was used for weights but 1.0 was used for data scaling. That should have been sqrt (0.333) to be consistent and so leads to weights wrong by factors of 2 for baselines having one such antenna and 4 for baselines with both antennas affected. The ModComps would not have allowed nominal sensitivities of 0.0, so this is a recent problem. Moved to patch Sept 28. 12467. October 15, 2007 SPLIT, SPLAT Eric The computation of the IF frequency of averaged channels was incorrect because it used FINC(BIF) and FINC(BIF) was already scaled by the degree of averaging. This affects all arrangements of IFs, not just the mix of USB and LSB for which the error was initially entered into the code. SPLIT was dufferent in that it scaled the FINC(IF) first and then used it with the scaled FINC(BIF). For IFs with the same FINC, no error was produced (unlike SPLAT), but a worse error was made when the FINCs had different values. Moved to patch this date. 12474. October 23, 2007 SPLIT Eric SPLIT still did not get things right. There was a very long-standing error in the header reference pixel computation when averaging channels but still making a multi-channel output file. SPLIT leaves the ref value alone for frequency and so the offset frequencies in the FQ table should not be changed - a more recent error. SPLAT does things differently, puting the ref pixel to 1.0 and changing the ref values. Moved to patch this date. 12489. November 5, 2007 SPLIT, SPLAT Eric I do not believe this - both tasks had the number of words/correlators to copy wrong (3 times too large). This can cross array boundaries messing up needed data and can certainly get wrong results for the scaling of compressed data. Moved to patch this date. 12491. November 6, 2007 SPLIT, SPLAT Eric I further don't believe how I messed yesterday's correction to copy too few for compressed data. Moved to patch this date.