;----------------------------------------------------------------------- ; Copyright (C) 2016-2018 ; 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 31DEC17 is TST *********************************************************************** *********************************************************************** 14939. November 30, 2016 Version update Eric Changed $APLSUB/GETRLS.FOR. $INSUNIX/INSTEP1, $SYSPERL/install.pl, $SYSPERL/update.pl, $SYSPERL/ZXHLP2.PL, $SYSPERL/LaCook.PL, $UPDUNIX/UPDCONTROL, $SYSUNIX/AIPSPATH.CSH, $SYSUNIX/AIPSPATH.SH to make 31DEC17 instead of 31DEC16 - also fix dates in install.pl. Moved nowhere. 14940. December 1, 2016 PCASS Eric NEW TASK: Uses pulse-cal amplitudes and, optionally, residual phases, to compute bandpass records for each calibrator scan. They look like proper bandpass things but remain to be tested against real data. Moved to 31DEC16 this date. 14941. December 2, 2106 PCASS Eric The handling of scans and PC records outside of any scan needed significant improvement. Also the BP table requires the max antenna number in its header or DOBAN > 0 will abort with a nasty set of dynamic memory failure messages. Added option to treat the PC amplitudes as power or (default) as voltage. Quick tests suggest that I do not know which is right. Moved to 31DEC16 this date. 14942. December 5, 2016 UVFLG Eric The shadowing function used variable DEC for computing the W term. Unfortunately, DEC is the header declination in degrees which is zero for multi-source and which needed to be in radians for the computation. Changed it to use DECEPO which is correct in all cases. Changed the inputs section of the help file to clarify the meaning of APARM(5) = diameter of the antennas to be flagged in meters. Moved to 31DEC16 this date. 14943. December 5, 2016 IMAGR Eric Added info to help file re ANTENNAS and BASELINE. Some users are apparently confused by the wide-spread usage of these two adverbs in AIPS. Moved nowhere. 14944. December 12, 2016 VLANT Eric The VLA has had an issue in the on-line atmospheric delay. It was wrong from IAT August 9 to Nov 15, 2016. I have added routines to correct for this following a formula from Barry Clark to VLANT which users should use first on their data and which writes a new CL table anyway. The routines check the observation dates and only do the correction if it is needed. I also added a little-used adverb to let the user turn off either this elevation correction or the antenna position correction with the default to do both when appropriate. Note that the error is a delay error but it is a function of time (elevation and azimuth) and of source so the full phase correction enters at each IF. The delay correction itself is small at elevations > 20 degrees or so but the phase correction matters even at rather higher elevations. Moved to 31DEC16 this date. 14945. December 15, 2016 VLANT Eric Vivek has determined that there was in fact only one time interval and that the approximate best value for the correction is -1.0E-15 sec/m. Moved to 31DEC16 this date. 14946. December 15, 2016 RFLAG, REWAY Eric Added some general advice (from Fraser) to the help files for these tasks. There are very many knobs and most can be ignored. Moved to 31DEC16 this date. 14947. December 16, 2016 PCNTR Eric Changed PCNTR adding DOFRACT adverb to make the polarization lines fractional polarization rather than total. New adverb in POPSDAT and NEWPARMS with new help file. Changed AU2A (PLGET) and AU8A (EXTLIST) for change in adverbs. Changed help to allow DOVECT to go up to 4. Task only cares that it is > 0, but sets it to 1, 2, or 3 depending on where the vectors come from. This lets PLGET; GO work. Moved nowhere. 14948. December 21, 2016 AIPSLetter Eric The last 31DEC16 aipsletter was released: LET16B .TEX, .PS, .PDF files. Moved everywhere since in $AIPSPUBL. 14949. December 22, 2016 UPDCONFIG Eric VALEN uses rsync so does not need cvs. Moved to 31DEC16 this date. 14950. December 27, 2016 TVSPC, Memo 120 Eric Changed TVSPC to display optionally a "spatial" (XYf) cube in grey-scale in a quadrant of the screen. This lets the user look at spectra at a pixel and images at a channel with a surprisingly fast interactivity. Updated the AIPS Memo 120 describing TVSPC as well. Moved nowhere. 14951. December 28, 2016 SETFC Eric Changed it so that BOXFILE=' ' will direct output to the message screen/file. That way one does not have to create a disk file if one does not want to. Field messages are at level 4, box messages at level 2. One can then PRTMSG only the field ones if one wants. Moved nowhere. 14952. December 29, 2016 UVFRE Eric Changed help file header so that it will be listed under spectral. Moved nowhere. 14953. December 30, 2016 REWAY Eric Added statistical print out to help find reasonable ranges of weights and pin-point odd baselines/IFs/Stokes. Moved nowhere. 14954. January 3, 2017 CALIB Eric Added adverb DOHIST. If DOHIST >= 0, a full history is written to the input file even if an output file is written. If DOHIST < 0, the task writes a full HI file to the input if no output is written or a short history if one is written. The latter was the standard previously. Moved nowhere. 14955. January 4, 2017 POSSM Eric The help claimed that correlation functions could be labeled in seconds (actually milli seconds or less) but the code went to trouble to turn off the option. Changed to allow it using Delta_t = 2 * Nchan / Delta_f. Moved to 31DEC16 patch this date. 14956. January 6, 2017 ATLOD Eric My error setting up TAPIO - worked for real tapes which I apparently had some time ago, but failed for disk files. The LUN tells TAPIO whether it is a real tape or a disk file. Moved to 31DEC16 this date. 14957. January 10, 2017 EVAUV Eric Added all of the calibration and data selection adverbs - forcing a sort-of CALCOP to a scratch file. Also added CELLSIZE which is used when SMODEL is set in order to make a usable histogram versus radius plot. Added more tests to skip plots that are not available (e.g. image histogram when SMODEL is set). Moved nowhere. 14958. January 12, 2017 CCRES Eric CCRES ignored blanking and so could add a component to the magic value making it no longer magic, just very large. It would also rescale magic blanks even if no CCs touched the blanked pixel. Also changed the default OUTCLASS and reworded some in the help file. Moved to 31DEC16 this date. 14959. January 13, 2017 NANS Eric Changed it to do both UV and MA files. In the latter, it does not look for magic blank values as they are "normal" and handled correctly (we hope). Moved nowhere. 14960. January 17, 2017 UVIMG Eric Added DPARM(9) to let empty pixels be zero rather than magic blanks. Added stuff from DPARM to history file. Moved nowhere. 14961. January 18, 2017 EVAUV Eric When I changed EVAUV to apply calibration and selection to copy to a scratch file, the header needed by most of the steps was the header of the scratch file. But the plot files needed the initial uv data set header. Added code to keep and use that header in making plot files. This bug destroyed the header info of the input file. Also dropped checking on Stokes values - leaving that to the calibration/selection code. Moved nowhere. 14962. January 19, 2017 UVIMG Eric The Hermitian property should always be used except when the two samples will fall on top of each other (normal baseline and/or projected baseline length and/or time as the 2 axes). Changed code to do this and count the samples in the grid where both Hermitians fall on the grid so as to make the reports more understandable. Moved nowhere. 14963. January 26, 2017 UVPLT Eric Changed the binning option to use double precision - there are too many numbers being summed for single precision to be accurate. Changed the task to make TV plots always as an array in memory and then loaded to the TV. Also changed it to make plot file plots as an array in memory unless the user says not to do so. (Small plots or ones requiring great accuracy might be better done the old way which remains available.) Note that the labeling and the plot of the binned results are done the old way always and are not part of the "image". That means that one can still modify the PostScript file from LWPLA fairly easily. The new option requires that I add the DOACOR adverb to encompass the old usage of BPARM(10). Moved nowhere. 14964. January 26, 2017 EXTLIST, PLGET Eric Changed AU8A and AU2A for the added adverb in UVPLT. A really old (2003?) version of UVPLT has fallen off of the EXTLIST support. Moved nowhere. 14965. January 27, 2017 DISKU Eric Changed the disp space counters from integer to double precision. The integer could overflow at 2 Terabytes. Moved nowhere. 14966. January 27, 2017 MOVE Eric Changed to test OPCODE 'MOVE' for the in and out disks being the same. It switches to copy with a message if not. Added OPCODE='DUPL' which allows one to write a file with a duplicate name although not on the same disk. This will allow copying a data area while preserving file names. It is potentially dangerous if the old data area remains in service include the DUPLed files. Moved nowhere. 14967. January 30, 2017 UVPLT Eric Removed debug statement that caused failures for TVs wider than 2000 pixels. Changed to allow the slow methood for the TV if BPARM(10) > 1. It can be instructive. Moved nowhere. 14968. January 30-31, 2017 DFTPL Eric Did not get the frequencies correct when BIF > 1. Also messed up the top label if BIF=EIF or BCHAN=ECHAN. Moved to 31DEC16 this date. 14969. February 2, 2017 FRPLT Eric Reworded and added to the help file introduction. Added BPARM(5) and (6) to set the X range optionally and only if APARM(2) > 0. In the Fortran, moved the adjusting of SOLINT until after the start time is adjusted to match the data. Changed the number of points plotted up by 1 and round up rather than down to make the number even. Save/restore the input values of APARM and BPARM before fixing them to match a plot in the plot file header. Set XMIN and XMAX from the plot or from the user and do the plotting using them. Moved nowhere. 14970. February 2, 2017 TVALL Eric TVALL used J as a temporrary variable. Using GG will mess up scripts less. Moved nowhere. 14971. February 3, 2017 ZEMAN Eric Debug arrays left in code. Moved nowhere. 14972. February 4, 2017 CATINI Eric The catalog header init routine set the epoch to 1950 - changed to 2000. Fixed lower case letter too. Moved nowhere. 14973. February 9, 2017 EVAUV Eric CELLSIZE was left out of help file. Moved nowhere. 14974. February 13, 2017 XGAUS, ZEMAN Eric Changed: XGINI To add a keyword to indicate if the model is emission or absorption. Changed amplitude units if absorption. ZEINI To add a keyword to indicate if the model is emission or absorption. Fixed Gaussian amplitude units in absorption. Added column for XGAUS baseline parameters which should have been there all along. Changed it to be forgiving about the old format file. TABZE Changed call sequence to include XGAUS baseline parameters. Read/write them if the column exists. XGAUS Test pre-existing XG for emission mode. Changed call sequence to XGINI. Dropped commented-out lines. ZEMAN Carry XGAUS baseline in a common along with the Gaussian parameters. Change call sequence to XGINI, ZEINI, and TABZE. Test pre-existing XG and ZE files to make sure they are in emission mode. Include the baseline parameters in the Ipol model plot. Dropped DOMAX which did not work when false. Moved nowhere. 14975. February 13, 2017 FRPLT Eric Changed formats to allow for more spectral channels. Moved nowhere. 14976. February 15, 2017 POPS Eric When certain verbs are used in FOR loops, WHILE loops, or IF-THEN-ELSE constructs the "optional" immediate numeric arguments are no longer optional but required. Else those verbs take a necessary control value from the stack and destroy the control. Affected help files include POPSDAT (TVALL proc), FOR, TO, BY, EHEX, GROFF, GRON, HUEWEDGE, IMWEDGE, REHEX, TVOFF, TVON, TVWEDGE, WHILE, IF, THEN, ELSE. Moved nowhere. 14977. February 16, 2017 PLOTR Eric Added digits to the model fit rms display. Moved nowhere. 14978. February 16, 2017 XG2PL Eric Changed to new call sequences for XGINI, ZEINI, and TABZE. Changed to handle absorption spectra and Zeeman fits. Added option to display absorption I polarization as optical depth instead of flux density. Moved nowhere. 14979. February 23, 2017 XGAUS, ZEMAN Eric Changed: ZEINI Pre-sursor comments about absorption keyword - new value => gain method TABZE Corrected error handling Gaussian baseline parameters XGAUS Made special versions of GRPOLY to handle images made with pixel replication - otherwise the polygon drawing was done with narrow line which were almost invisible and then polygon fill did not always come out right. ZEMAN Changed to display one-pixel fitted images. Moved nowhere. 14980. February 23, 2017 AGAUS, ZAMAN Eric Made absorption-line versions of XGAUS and ZEMAN. The math differs in so many places that it was best not to combine the two versions in a single task even for the Zeeman splitting. AGAUS fits Gaussians to the optical depths but fitting the observed lines. This should handle relatively high optical depths. The math for the Zeeman spliting also changes. A special mode for the leakage gain when the Vpol cube was made from continuum subtracted visibilities is also offerred. Moved nowhere. 14981. February 23, 2017 XG2PL Eric Changed to handle outputs from all 4 fitting tasks including the special gain mode added to ZAMAN (for Vpol from continuum subtracted visibilities). An option to plot optical depth spectra rather than the observed Ipol is also available. Moved nowhere. 14982. February 23, 2017 AIPS Memos Eric A new AIPS Memo 122 has been written to describe AGAUS and ZAMAN, plus changes to XG2PL. A section on systematic errors in Gaussian widths and Zeeman splitting is also presented. AIPS Memo 118 was modified to mention 122 and to remove adverb DOMAX which was found not to work. 14983. February 24, 2017 DBCON Eric A BAD ERROR was made May 28, 2014 in DBCON. If the 2 data sets differed in frequency (after shifting of the reference pixel), the U-V-W of the second data set was scaled by the ratio of the frequencies. This is not correct - the u-v-w with no scaling are in wavelengths at the observation and should not be changed. (A shift of reference pixel does require a modest re-scaling based only on the frequencies in the particular data set.) I have restored the correct code in DBCON and added a loud warning message to the Fortran and the help file. The output of mixed frequencies should NEVER EVER be processed by a task that needs to change U-V-W since it cannoy fully know the frequencies to use. The most obvious of these is UVFIX. Moved to 31DEC16 this date. 14984. February 24, 2017 PRTUV Eric Corrected it to take BIF into account for the frequency it prints. It did the channel but not the IF. Moved nowhere. 14985. March 1, 2017 DFTPL Eric Corrected addressing of visibility data - pointers need to be set after UVGET ('INIT'..) not before. Led to IFs > BIF being ignored. Added STOKES to adverbs in help and Fortran and got rid of the messy selection of correlator to plot. Trapped actual plotted RA and DEC and added that info to plot header. Moved pointer fix to 31DEC16 28 February.. 14986. March 1, 2017 DFTIM Eric NEW TASK: makes an image of the DFT at a selected coordinate from visibility data as functions of frequency and time. It can also make an error image. The data selection, flagging, and calibration adverbs are offerred as well as regular frequency averaging (NCHAV) and time averaging per pixel. Moved nowhere. 14987. March 1, 2017 TYAPL, TYSMO Eric Added paragraph to TYSMO and TYAPL help files emphasizing the need to have valid post-detector gains as well as Pdif and Psum for use by TYAPL. TYAPL in most places defended itself against blanked post-detector gains. Added that defense to the few places missing it. Moved nowhere. 14988. March 1, 2017 VBGLU Eric changed it to glue together multiple versions of any extension tables it finds (except FG tables). All input versions n go to output version n. This is essential for ANtenna tables since they connote subarrays. It may be useful for other tables but also offers opportunities for mischief. Reworded help file introduction to modernize and to mention the tabls issue. Moved nowhere. 14989. March 3, 2017 ELFIT Eric NEW TASK: Fits polynomials to table data as functions of elevation, zenith angle, hour angle, parallactic angle, and azimuth. Rick is using it for Psys vs zenith angle. It is a simplified SNPLT with the added fitting. Moved nowhere. 14990. March 6, 2017 ELFIT Eric Some arrays were only 1000 - made much larger and added tests to avoid overflow. Moved nowhere. 14991. March 7, 2017 Maximum disk number Eric Changed the maximum disk number to 71: Changed ZDCHIN $APLGEN: use characters 3 and 4 of disk logical ZPHFIL $APLGEN: use characters 3 and 4 of disk logical volume limit test changed ZPHOLV $APLGEN: use characters 3 and 4 of disk logical DSKASS $AIPSUB: use characters 3 and 4 of disk logical REBYTE $AIPNOT: use characters 3 and 4 of disk logical and use DA1Y and DA1Z for scratch areas DADEVS.SH $SYSUNIX: Use $DA0$ehex for < 36, $DA$ehex for greater and set limit to 71. DADEVS.PL $SYSUNIX: Use $DA0$ehex for < 36, $DA$ehex for greater and set limit to 71. DDCH,INC $INC: several arrays (35) -> (99) DDRC.INC $INC: (36) -> (99) DISK.INC $INC: (36) -> (99) ZFREE $APLGEN: several larger arrays ZFRE2.C $APLUNIX: DA0n needs if statement with a new DA1n ZFRE2.C $APLMACIN: DA0n needs if statement with a new DA1n ZCREAT $APLGEN: bigger buffers to call ZFRE2. DDBGR $AIPNOT: corrected to use ZHEX10 to parse DAnn OPTIMIZE.LIS $SYSUNIX: restore APCLN at level 1 opt only Major COMRPL required, relink everything Moved nowhere. 14992. March 7, 2017 PARANG Eric PARANG Did not recognize EVLA as having the same attributes as the old VLA (same parallactic angle for all stations). Depends on the array name being set in DANT.INC and uses the FILLM final array center values. Moved nowhere. 14993. March 7, 2017 AIPS Memo 122 Eric Fixed some typos and corrected the Zeeman absorption math. The answers were correct but the development was all messed up. Moved nowhere. 14994. March 8, 2017 BPASS Eric The normalization routine for BPASSPRM(10)=1 had an error affecting linearly polarized data. The normalization was likely to be very wrong for some IFs and polarizations. Moved to 31DEC16 this date. 14995. March 9, 2017 ELFIT Eric Added APARM(2) and (3) to allow user control of fiducial points when printing model values after fit. Moved nowhere. 14996. March 9, 2017 AGAUS, ZAMAN Eric Created run-time help files HLPAGAUS.HLP and HLPZAMAN.HLP and fixed a couple of references in the two tasks to specify the new (overlooked) run-time helps. Moved nowhere. 14997. March 14, 2017 DBCON Eric The routine that merged AN files did not allow for the second data set to have antennas not in the first. DBCON cannot renumber antennas so the two must be compatible but antenna n can occur in the second file even if it does not occur in the first. This required keeping track of the used antenna records from data set 2 and what antenna numbers have already been used and then adding a loop to read any unused records from the 2nd AN file and potentially writing them to the output AN table. Note - the task keeps going even if antennas do not match up - but the results amy not be desirable. The task MATCH is now recommended on various disagreement conditions. Moved nowhere - this is a bit messy for a patch. 14998. March 14, 2017 FITLD Eric Corrected ANTNAME usage so that it is not ignored when concatenating. Instead, it is tested and the pre-existing file will override the users' ANTNAME specification. This can be very bad so warnings were added to the help file on ANTNAME and DOCONCAT. The special ANTNAME='VLITE','' will generate 28 VLITE-like names in order 'V0','V1',...'V27'. Moved nowhere. 14999. March 17, 2017 FITLD Eric The VLITE names will use EHEX rather than decimal - changed the current implementation: V0...V9, VA, VB, ... Moved nowhere. 15000. March 17, 2017 Holography Eric The date of observation and the antenna number did not appear in the final PANEL outputs. Changed UVHOL to add DATE-OBS to the output text file in the first line. Changed HOLOG to pick up the antenna and date-obs from that line and to use them to set TELESCOP and DATE-OBS in the output image headers. Changed PANEL to pick these parms up from the header and put them ina new first line in the text outout. Moved nowhere. 15001. March 21, 2017 Spectral Index Eric I insisted that spectral index was based on 1 GHz and then became inconsistent in its application. Those tasks that use Rick Perley's calibrator lists (CLIP, SETJY, UVFND, BLCHN, BPASS, CPASS, FTFLG, SPFLG, UVPLT) are obliged to stay on 1GHz as a reference to match the lists including when SPECINDX and SPECURVE are specified by the user. But other tasks can be more flexible. I note that IMAGR and OOCAL already read the reference frequency for spectral index from the spectral index images and use that in PBUTIL. That leaves: SETJY Help - fixed wording POPSDAT Help - make new adverbs SPECFLUX, REFREQ NEWPARMS RUNSYS - make new adverbs SPECFLUX, REFREQ UVMOD Help - clarify the meaning of the flux, spectral index, and curvature to be the header reference frequency. UVMOD Fortran - changed a bad kludge that sort of used 1GHz for the curvature, but set the fluxes at the header reference frequency. Now it is the header frequency for all. EVAUV Help - add SPECINDX, SPECURVE to be used when SMODEL is specified. Also add automtic fitting when SMODEL < 0. EVAUV Fortran - add adverbs, pick up spectral index and fit it from FNDSPX if requested, display adverbs used, in EVADFT (local VISDFT) add SMODEL with spectral index code to do a channel at a time in both subtraction and division. SPIXR Help - add REFREQ to specify the desired frequency to shich the output images are computed. 0 -> 1 GHz the old value. < 0 -> header frequency SPIXR Fortran - moved parameter setting so that output header frequency will be REFREQ. Compute frequency array relative to REFREQ rather than 1 GHz for doing the fits. SOUSP Help - added REFREQ adverb as SPIXR SOUSP Fortran - added REFREQ and do all scaling wrt REFREQ as opposed to 1 GHz. Got tricky in plotting the X axis but worked it out. FNDSPX Fortran routine made the mistake of leaving a loop counting valid FLUX values on the first invalid. This can happen due to flagging etc - the counting loop must go over all IFs Moved UVMOD, FNDSPX, some of SPIXR, 15002. March 27, 2017 OOCAL, OOFRING Eric Added NEW PROCEDURE OOFRING in the OOCAL.001 RUN file. It is like OOCAL but does FRING rather than CALIB. Changed OOCAL help file dropping a bunch of adverbs that calibrate and select the data to be processed - since OOSUB cannot do that. Added paragraph to help file pointing this out and requiring that the data be fully selected in advance. Also added INDXR to procedure and actions to preserve the input values of CALIB and FRING and INDXR because the procedures must change them. Changed OOSUB since on multi-source files the SU table is copied in advance of the later extension file copies - like the FQ and AN tables. Changed low-level routine TABCOP since it did not make messages about a bunch of its many possible errors. Moved nowhere. 15003. March 28, 2017 FITLD IMLOD UVLOD Eric The number parser caused these tasks to quit on bad numbers when the keyword is not structurally significant. Changed them to report bad numbers but set the value to 0 except when reading number points on an axis, PCOUNT, GCOUNT, and other required and structural keywords. Moved nowhere. 15004. March 29, 2017 HOLOG Eric Swap X axis left to right so that we are looking at the antenna from the front (as we thought we were doing all along). Moved nowhere. 15005. March 29, 2017 OOCAL IMFRING Eric NEW PROCEDURE: IMFRING to take a CASA model image and break it into facets with Clean components to feed to OOSUB. Changed OOFRING help - drop SNVER adverb OOCAL help - drop SNVER adverb, restore ICHANSEL adverb, remove adverbs no longer used from help section IMSCAL help - remove all adverbs no longer "used" by OOCAL including CALSOUR, QUAL, CALCODE, SELBAND, SELFREQ, FREQID, TIMERANG, SUBARRAY, DOCALIB, GAINUSE, DOPOL, PDVER, BLVER, FLAGVER, DOBAND, BPVER, SMOOTH, and CMODEL. IMFRING help - new proc help has adverbs for IM2CC, OOFRING. OOCAL.001 run file - add IMFRHIST proc to do IM2CC history to input data set and enable call to OOFRHIST from OOFRING. Add proc IMFRING to do IM2CC, count output images, invoke IMFRING, clean up images. OOCAL and OOFRING now clean out SN tables from temp file. Moved nowhere. 15006. March 31, 2017 PANEL Eric Added more explanatory text to the output to make sure that all conventions are fully described. Moved nowhere. 15007. March 31, 2017 Perley-Butler 2017 Eric Rick and Brian have measured fluxes again and have now fits that cover the entire frequency range for standard sources and other fits for less standard sources which may be of use. Changed BPASS, CPASS, BLCHN, and PCAL to use the new values if the observation date is after Jan 1 2014. They do only the 5 standard calibrators (3C48, 3C138, 3C147, 3C286, 3C295) automatically but will use the fluxes in the SU table for other sources. SETJY help - incremented the values of APARM(2) so that 0 points at Perley-Butler 2017. Document all the extra sources whose flux may be found even if they are not good calibrators (absent a decent model). SETJY Fortran - made new larger tables of spectral index and possible sources. The new sources (poor cals usually) have frequency ranges that are tested with a message for out of range. Changed so that errors from GETFLX do not stop the processing just fill zeros. Moved nowhere. 15008. April 5, 2017 UVIMG Eric Changed it to allow NCHAV >= (EIF-BIF+1)*(ECHAN-BCHAN+1) to mean make a single plane using all the channels. Moved nowhere. 15009. April 5, 2017 IRING Eric Added FUNCTYPE='LG' to plot Y axis in log base 10 for the plot of mean values. This lets us deal with severely centrally condensed things like UV distributions. Moved nowhere. 15010. April 7, 2017 IRING, UVIMG Eric Changed IRING to allow 10240 rings maximum. Fixed UVIMG addressing of channels - was wrong for new option. Moved nowhere. 15011. April 12, 2017 SUBIM Eric Added OPCODE 'SUM' for choices when XINC > 1 and/or YINC > 1. It did min, max, and average but sum should have been included. Moved nowhere. 15012. April 12, 2017 Slices Eric Changed: SLICE Corrected the buffer size parameter to match the actual size. The parameter was way small for big slices. AU6E The call to plot components had an incorrect argument. SL2PL Help file had small limits on BDROP and EROP. Moved nowhere. 15013. April 12, 2017 DTSIM Eric Bad addresses were used when there was no bandpass model. Changed DTSIM and DTBAND to check for index 0 and fill with zeros rather than whatever happened to be in RAM at the bad address. Fixed potential failure in history also. Moved nowhere. 15014. April 13, 2017 CLCOR Eric While trying to trace a mysterious installation error (START_AIPS with AIPS_ROOT defined as /home/AIPS which is correct for the AOC only) I found a hard-coded path name in CLCOR for the JPL ephemeris file. Corrected that code to use the normal AIPS file name - AIPSTARS:... and to translate that with ZPATH. Changed JPLEPH to allow the incoming file name to be up to 256 characters rather than 80. Moved nowhere. 15015. April 13, 2017 SLICE Eric WaWa IO tripped me up - the buffer size sent in is in full words not AIPS "bytes". Blew up badly when reading multiple rows. Moved nowhere. 15016. April 13, 2017 history Eric HISCOP took a failure of HICOPY too seriously - change to report failure at level 8 (not suppressed) but return 0 if new file okay. HICOPY did not set error to 0 when read 0 records. Moved nowhere. 15017. April 19, 2017 IMAGR Eric Added option of inverse taper with several statistical counters to display what was done. Changed IMAGR Pick up part of IM2PARM and pass on as INVERT enahnce history in beam files QUVUTIL Pick up INVERT and pass to UVWAIT as new argument UVWAIT New call sequence, implement inverse taper plus counting and average weight and sum of weight in inverse taper region. Moved nowhere. 15018. May 5, 2017 SETJY Eric A logic error caused the comparison of 2017 fluxes to 2013 fluxes to be erroneous. The actual fluxes stored in the SU table were fine. Moved nowhere. 15019. May 8, 2017 PCLOD Eric Corrected code to handle multiple concatenated files with multiple header lines each. Added check on antenna name in each data line matching the one from the previous header (and changing if necessary with a message). Moved nowhere. 15020. May 9, 2017 FITLD Eric VLITE antennas are named V1...V9, V10...V27 rather than the EHEX options previously suggested. Moved nowhere. 15021. May 10, 2017 SETJY Eric Another residual line of code changed type 0 to -1 for data below 500 MHz. We want to do that for types > 0 now, since Perley 2017 should work for low freq. Moved nowhere. 15022. May 12, 2017 PCLOD Eric Round out the scan boundaries - PC records were just missing when they should not. Moved nowhere. 15023. May 12, 2017 PCFIT Eric Added time averaging under control of SOLINT. It uses scan boundaries but also does not average over changes in antenna, source, and pulse-cal frequencies. Therefore it does handle PC tables with multiple frequencies even if they are not all supported in the UV data set. Moved nowhere. 15024. May 15, 2017 STUFFR Eric The procedure was missing a semi-colon and needs to force INTYPE to 'UV' so that the ZAPs will work. Moved nowhere. 15025. May 16, 2017 FRING Eric Corrected APARM(5) info in help file. Moved nowhere. 15026. May 26, 2017 FILLM Eric Changed the positions equal test from 1.0D-6 (3.6 milli-asec) to 6.0D-6 degrees (~21 milli-asec) since coordinates do seem to shift a bit in some of these old data sets. Moved nowhere. 15027. May 26, 2017 IMAGR Eric Changed QCLEAN to keep track of which field number set the current intensity range. If it changes, the code will use the full range of the new field. If it does not, the past behavior will be used (keep pix range unchanged unless the image range is too different). Moved nowhere. 15028. June 2, 2017 1-liners Eric Improved the 1-liners in the help files for CALRD, CALWR, and CALDIR. Moved nowhere. 15029. June 6, 2017 PCAL Eric The source assignment with Perley 2017 was confused and bad values could result. Led to bad solutions with DOSCALE set The evaluation of the flux at 1 GHz was wrong for low frequencies leading to bad solutions. Moved a similar fix to 31DEC16 this date. 15030. June 7, 2017 RLDLY, CLCAL Eric Discovered that CLCAL will re-reference the right-left delays all to zero if it does any re-referencing. Made remarks in the RLDLY help file. Changed CLCAL to detect if an SN table was created by RLDLY and if so to block re-referencing. If there is more than one SN table messages will appear noting any difference in creation task. If any one is RLDLY, no re-referncing is done. Moved nowhere. 15031. June 15, 2017 Solar cals Eric Changes to support Solar observations: The CD table gets two new columns giving the values of the Solar Tcals. The SY table gets a columns indicating which type of Tcal one uses with that antenna and time. SYINI Add column 3 = "CAL TYPE" an integet with 0 being normal, 1 Solar, 2 ??. All that column to be missing. TABSY Add CALTYP argument with default value out of 0. SYSEL TABSY call sequence CDINI Add 2 columns for SOLCAL1 and SOLCAL2 and allow those columns to be missing. TABCD Change TCAL from 2 per IF to 4 per IF always. Return solar if present, blanked if not. A solar "value" of exactly 1.0 also returned blanked. CDSEL Number columns, move TCAL values in IF change. GETCDS Number columns, dimension of TCAL, loop over all 4 in error message and limit the number of same QEDIUTIL Changed dimension of TCAL. Added one more "random" parameter to the SY data format and filled it with CALTYP (reading from file if columns exists only). Changed code to flag the time only when Psum, Pdif, or Pgain are bad. The Psys values are blanked on those and also when Tcal undefined. FIXRL TCAL dimension and swap Solar cals too, TABSY call sequence LISTR Pick up the type columns when needed and pass to the gain average routine - select TCAL based on the current value in that column. Done several places. PRTSY Change calls to TABSY, use CALTYP to slect TCAL value SPLAT Changed TABSY call and dimension of pointers SWPOL TCAL dimension and swap Solar cals too, TABSY call sequence SY2TY TCAL dimension, TABSY call sequence, use CALTYP to select which TCAL to use TACOP Changed TABSY call and dimension of pointers TYAPL Changed TCAL dimension, call sequence to TABSY. It will fail if solar data are given to it. TYCOP Dimension a proper TCAL and use it and the solar values. Call sequence of TABSY also. TYSMO Change column pointers for SY table, pick up type column. Use it as needed to select TCAL values UVCOP Changed TABSY call and dimension of pointers VLAMP TCAL dimension, TABSY call sequence - solar had better not be used ELFIT pick up CAL TYPE and use when TCAL is needed SNPLT pick up CAL TYPE and use when TCAL is needed SNIFS pick up CAL TYPE and use when TCAL is needed. Fixed error zeroing max/min arrays. Help files: TYAPL Do not use for solar data TYSMO Set DOBTWEEN false for solar data! VLAMP Does not work for data including the Sun. Moved nowhere. 15032. June 15, 2017 SYSOL Eric NEW TASK: A version of TYAPL for SY tables and solar data. For the Sun it averages the visibility and weight scales over those antennas having Solar Tcals and then applies the average to those antennas that do not have Solar Tcals. It does TYAPL exactly for regular sources. Moved nowhere. 15033. June 15, 2017 CVEL Eric Corrected a DO loop that assumed there were 10 IFs - used in display only. Changed warning for single source files to apply only to VLBA and EVN - they care about doing BP corrections on the fly. Moved nowhere. 15034. June 16, 2017 ABOUT files Eric Updated the tab completion file and about files on ADVERB ANALYSIS AP CALIBRAT CATALOG FITS IMAGE IMAGE-UT INTERACT MODELING ONED OOP PLOT POLARIZA PROCEDUR SPECTRAL TASK TV-APPL TV UTILITY UV VERB VLA VLBI Moved nowhere. 15035. June 20, 2017 MODAB Eric NEW task: makes a simple image model cube in I and V in either emission or absorption. Moved nowhere. 15036. June 21, 2017 SQASH Eric The task did not handle BLC and TRC correctly in the first two axes. It worked with output plane size equal input plane size but not otherwise. Moved to 31DEC16 this date. 15037. June 22, 2017 CookBook Eric Changed chapters: 2 For change from 35 to 51 maximum data areas 4 For SETJY, Solar calibration 6 For PCNTR, TVSPC, AGAUS, ZAMAN, downplay Tektronix usage 7 For DFTIM, TVSPC, AGAUS, ZAMAN, ELFIT, RMSD 8 For TVSPC, RMSD, AGAUS et al. 9 For PCASS, PCLOD 13 To update numerous ABOUT files E For REWAY, SETJY, IMSCAL, OOFRING, Solar cal Also chapters 0, I, J Recipe to match the changes Moved nowhere. 15038. June 30, 2017 UVLSF Eric The task for no good reason called COPTAB twice for the BASFIT (continuum channel) output file. This made a bad FQ table in particular and occasioned some error messages. Moved to 31DEC16 this date. 15039. June 30, 2017 AIPSLetter Eric Put back TEX, PS, PDF versions of June AIPSLetter. Moved nowhere. 15040. July 3, 2017 SCMAP, SCIMG Eric The call to the self-cal routine had variables in the wrong order, making the data array description messed up when there were not 2 IFs. Moved to 31DEC16 this date. 15041. July 3, 2017 SCMAP, SCIMG Eric Changed help files to emphaisize APARM(8) to allow negative CC components. Moved nowhere. 15042. July 4, 2017 IMAGR Eric Added OVRSWTCH and FGAUSS to the allowed SHOW/TELL adverbs. Changed QCLEAN to pick up these adverbs from TELL and also to change IMAGRPRM(11), the multi-scale bias. Changed IMAGR to prohibit any OVERLAP except >= 2 for multi-scale, blocking both OVERLAP < 2 and any OVRSWTCH. Changed help file to enable these TELLs and to describe the TELL possibilities. QCLEAN also blocks OVRSWTCH and had a bad-things-are-happening change to OVERLAP 1 altered for multi-scale. A bad format for "un-boxes" was also corrected. Moved nowhere. 15043. July 5, 2017 VLBARUN Amy Added CLRTEMP after VLBAPANG, to solve occational BLEW TEMP C problem. set BOXFILE and OBOXFILE to '' for IMAGR runs. Moved nowhere. 15044. July 11, 2017 PCAL Eric If the IF being copied to a scratch file had no data, the task left the ionosphere file open. When the next IF was to be copied this caused a failure. Moved this fix to 31DEC16 this date. 15045. July 12, 2017 AVSPC, NOIFS, UJOIN Eric AVSPC and UJOIN both copied CD and SY tables in the global copy and then attempted to change their contents. This would work if the format had not changed. But it has. Added CD and SY to the list of tables not to copy, leaving that to the change of contents step. Added CD table reformating to NOIFS. Moved nowhere. 15046. July 14, 2017 RSPEC Eric The previous usage in which BLC-TRC was an exclusion region was confusing. Added DOINVERS adverb with now normal meaning - do statistics in BLC-TRC if DOINVER <= 0, outside BLC-TRC if DOINV > 0. Also changed some logic that dropped BLC/TRC when the exclusion region covered nearly all of one axis. Now it drops them when both axes are nearly completely covered and only on DOINV > 0. Moved nowhere. 15047. July 15, 2017 TYAPL, PRTSY, VLAMP Eric All 3 of these had an SYFSEL subroutine locally to apply flags to the SY table but had not had the number of columns in the SY table updated in the declarations. Other tasks with SYFSEL had had this correction made earlier (UVCOP, SPLAT, TACOP). This led to copy errors in some of the architectures. MOved nowhere. 15048. July 15, 2017 DBCON Eric If the two frequency refernce pixels are in agreement, we should not change that unless requested to center REFPIX. Moved nowhere. 15049. July 15, 2017 ABORTASK Eric Added adverb NUMTELL to this pseudoverb. This now allows a second AIPS session to stop a task from the another AIPS session. This is useful when the task has locked up the input terminal that session. Changed HELP files for NUMTELL and BORTASK and subroutine AU2. Moved nowhere. 15050. July 20, 2017 TVSPC Eric Added LABEL IMAGES? and CURVALUE options and suppressed the VERBOSE option. Edited the help and run-time help files to add these options. Memo needs to be done. Moved nowhere. 15051. July 20, 2017 VLBAUTIL, VLBAEOPS.HLP, CLCOR.HLP Amy URL for the EOPs file changed. Fixed in the VLBAEOPS script and VLBAEOPS.HLP and CLCOR.HLP files. Moved nowhere. 15052. July 20, 2017 WETHR Eric Changed DWXV.INC which specief a record length one too short. There are 13 columns but time is double precision so length is 14. Changed WETHR which used a very bad way to decide whether time is double or single precision in the table. Could affect APCAL, FILLM, FITLD, INDXR as well as WETHR. Moved nowhere. 15053. July 21, 2017 ANREFM Eric The subroutine had ways to generate an error message and leave without closing open file(s). ANTINI wants to procede even if errors but the open files blocked that. Fixed ANREFM to close any open files. Moved nowhere. 15054. July 21, 2017 WETHR Eric Added option with FACTOR < 0 to connect the points with lines. Points can be arbitrarily small in that case. Moved nowhere. 15055. July 21, 2017 UVHOL Eric the task was not careful with antenna numbers and so managed to have them wrong when an unwanted baseline ended the group to be printed. Since autocorrelations are allowed for some reason, the amplitude ended negative (potentially) and the phase pure zero in parallel hands. Moved nowhere. 15056. July 22, 2017 AIPS Memo 120 Eric Updated AIPS Memo 120 to cover the LABEL IMAGES? and CURVALUE options. Fixed color scheme to make the changes obvious and updated all figures. Moved nowhere. 15057. July 24, 2017 IMAGR Eric Changed QUVUTIL to call UVWAIT when inverse taper is requested even if taper and uniform weight are not requested. Moved nowhere. 15058. July 24, 2017 UVHOL Eric For some reason, autocorrelations were allowed. Dropped that. MOved nowhere. 15059. July 28, 2017 UVMOD Eric Changed to allow comment lines starting with $ or #. Moved nowhere. 15060. July 31, 2017 UVMOD Eric Changed to allow negative components and to emphasize that NGAUSS limits how much is read from INLIST. Moved nowhere. 15061. August 1, 2017 HOLOG Eric Changed to add routines to compute the images from user-supplied parameters (VPARM) rather than fitting the data. This can either replace the phase with the model or subtract the model from the observed phase. It can also replace the amplitude (above some value) with the average of values above that value and set the rest of the amplitudes to 0. Also added a new OPTYPE = 'SURP' to simulate the effect of reference pointing on the model (it pulls the beam back to the center, changing the math for the feed offset X and Y parameters). Moved nowhere. 15062. August 2, 2017 LISTR Eric The gfortran optimized version did very strange things on the GAIN function. Changing the table IO record to be equivalenced to a buffer specified in LISTR rather than DSEL.INC seems to cure it. Moved nowhere. 15063. August 11, 2017 FGSPW Eric NEW TASK: designed to find spectral windows overflowed by RFI and to flag the full window and scan on a baseline basis. Moved nowhere. 15064. August 14, 2017 OPTIMIZE.LIS Eric testing on my computer with my intended compiler options revealed that APCLN requires fails even at optimization level 1 at least when variables are allowed to be dynamic. Moved nowhere. 15065. August 28, 2017 GREYS Eric It messed up LTYPES < 0. Copied the handling of this from PCNTR and KNTR. Moved nowhere. 15066. August 28, 2017 WIPER Eric WIPER did not position the CURVALUE display correctly in zoomed images when the display was on the far right. It also could not handle antenna numbers > 100 (not a problem for most people). Moved nowhere. 15067. August 29, 2017 RMFIT Eric I had forgotten to move the correction made to XGAUS Feb 23 into RMFIT. The plotting and handling of blotches in pixel-replicated images is a bit tricky and is used when swapping TV-selected regions. Moved nowhere. 15068. August 31, 2017 ANTAB Eric Changed the explain section on TSYS/TANT to expose the full allowed formats for times. KEYIN handles the sexagesimal indicator (a colon) nicely, simply adjusting a scaling factor for the next numbers it reads. Thus hh.hh, hh:mm.mm, and hh:mm:ss.ss are all okay for times in KEYIN. Moved nowhere. 15069. August 31, 2017 WAITTASK Eric Adde adverb NUMTELL to WAITTASK in subroutine AU2. Also fixed help files for NUMTELL and WAITTASK. Now one aips can wait on another. E.g. wait on IMAGR before running a KNTR in a 2nd window. Moved nowhere. 15070. September 7, 2017 ANREFM Eric CASA is now writing AN tables with one of the new columns and not the other. Changed ANREFM to avoid being upset by that stupidity. Moved nowhere. 15071. September 8, 2017 REFRMT Eric Changed subroutine REFRMT to drop ALL leading blanks as well as its previous job of reducing consecutive blanks to a single one. Moved nowhere. 15072. September 8, 2017 TVFLG, SPFLG, FTFLG Eric Work on PCFLG revealed an issue with time labeling. Changed it to show days when the day number > 0 even when the day number does not change within the plot. Moved nowhere. 15073. September 11, 2017 TVFLG, SPFLG, FTFLG Eric More work on PCFLG has revealed that enough parameters are changed while finding that the requested image (i.e. next part) lacks usable data that the labeling of the image will change. Therefore, it is necessary to fill the image with zeros rather than leave the old image with new labeling. Moved nowhere. 15074. September 11, 2017 Pulse Cals Eric PCLOD 1. Corrected to set the start day only once and to do that from the data set's reference date. Later concatenated files could have a later start day but we must keep the one appropriate to the data. 2. Added DOKEEP adverb to choose whether records not inside scans are kept and added counters to report the number of records written and the number not inside scans. Added these to the history as well. PCFLG 1. Raised the maiximum number of times allowed and added code to test for it being exceeded 2. Changed to copy the PC table if it is already correctly sorted. 3. If a start time is specified make sure that the scan information updates. Scans before the start time (in TIMERANG) were creaping in partly. 4. The gridding routine assumed that all desired planes were present when the memory allocation put a cap on things. Dropped the cap, but also fixed the assumption so that the cap could be restored. 5. Fixed discord between eror messages and their FORMAT statement. 6. Fix time axis display to include day numbers > 0. 7. When no valid pixels are found, still load the all-zero image. PRTAB 1. Changed it to always add at least a 1-digit day number to times in sexagesimal format. PCPLT 1. Added message about sorting the PC table Moved nowhere. 15075. September 14, 2017 PC stuff Eric PCAVG NEW TASK: time averages a PC table either freely or in strict intervals within each scan. The latter helps get al antennas to have the same times. PCFITR Subroutine assumed that frequencies increased with tone number (channel). The opposite sign is also possible. PCFLG Help file - fixed typo, added paragraph to point out what I was missing PCFIT Added CUTOFF to reset the delay zero points if the delay changes by too much. Moved nowhere. 15076. September 15, 2017 LISTR Eric 1. The LIST option divided all autocorrelation values by 100 - changed to do that for amplitude only. 2. Added type 19 for post detector gain in the GAIN listings (all 4 subroutines). Moved nowhere. 15077. September 28, 2017 KEYIN Eric Changed KEYIN to read text-file lines up to 2560 characters in length rather than 256. Link edits a bunch of stuff but main task this is aimed at is ANTEB. Edited line-length remarks in help files for ANTAB, ANCAL, and LDGPS. Moved nowhere. 15078. September 28, 2019 APCAL Eric Raised the buffer that holds all the data for an antenna by a factor more than 10 (260000 -> 3000000). This will allow 45 IFs with 2 polarizations for the 32768 maximum number of times allowed. Let's hope that is enough. Raised the load module from 39 to 160 Mbytes. Moved nowhere. 15079. October 9, 2017 RMFIT Eric Output units were messed up by things done for XGAUS rather than RM images. Affected the Q0 and U0 images. Moved nowhere. 15080. October 12, 2017 RFARS Eric It did not handle FQID axes correctly - with bad catalog header references. Moved nowhere. 15081. October 21, 2017 PCPLT Eric Corrected: 1. Number of parameters - XYRATIO was being ignored. 2. Pick up of ANTENNAS when specified. It failed causing the plot to find no data. 3. Set TRC with XYRATIO and use BLC and TRC for coordinates rather than 0 and 1000. 4. Allow horizontal axis labeling for any number of panels. Moved nowhere. 15082. October 21, 2017 PCFIT Eric Changed the sign of the delay and phase written to the SN table. Tests comparing the answers with FRING require this. Moved nowhere. 15083. October 22, 2017 PCEDT Eric Changed BPEDIT to make a bigger scratch variable, to correct units, to remove BP references in PC places, and to add progress messages. Moved nowhere. 15084. October 22, 2017 SNPLT Eric Added XYRATIO. Moived nowhere. 15085. October 24, 2017 Pulse-cal Eric Changed: (1) PCAVG, PCFIT, and PCFLG to insure that the output PC table sort order is properly indicated in the table. This has to be done explicitly - TABINI clears the sort order. (2) PCASS to add sort and progress messages, to add BPARM(4) which can cause all scans to be averaged together. (3) POSSM fixed to honor XYRATIO. Moved nowhere. 15086. October 24, 2017 new pulse-cal tasks Eric Two new tasks: PCRMS computes PC table statistics, can write an output file with PC IF-average data, and can write a new flagged PC table. This is an "auto-editor" meant for use in pipelines. PCHIS is a generalized histogram plotter written to handle PCRMS output, but usable for other sorts of text files. Moved nowhere. 15087. October 26,27, 2017 Compiler finds Eric Installed a new Intel compiler on a new, virtual "rishi" and tested AIPS under it. This version (18.0.0 20170811) is an attempt to get away from the NaNs that have appeared recently with the old (2009) version of the compiler we were using. Y2K performance results are only marginally improved over the old version and both old versions beat out gfortran by only a little. The new Intel compiler for LNX64 does not like overstating EQUIVALENCEs such as repeating the statement or giving an unnecesary one that replicates what the COMMON statement does. Changed: BLOAT, CANDY, CUBIT, DBCON, FETCH, MODVF, TAPPE, and TARPL for this. The new compiler also points out all EW.d FORMAT codes where W is less than d+7. Subroutines with this were GR3DUV GRDMEM Tasks include SAD TVSAD CLCOR CUBIT FARAD FIT2A FITLD IMLOD LISTR M3TAR MK3IN POLCO VLOG APGPS BSPRT HFPRT DELZN IBLED UVPLT ELINT had many, many format issues and did not put a 0P following 1PE formats so the regular F formats were scaled inadvertently. CHANGED $SYSLNX64 versions of FDEFAULT.SH, LDOPTS.SH, and CCOPTS.SH to match the lesser requirements of the new compiler. Changed several compiler options in XAS.SHR's Makefile section. Moved nowhere. 15088. October 27, 2017 USUBA Eric Raised MAXSOU from 300 to 400 and MAXENT from 100000 to 750000. This is as big as the compilers and linkers will allow without rewriting this code. And it is complicated code that I have never been able to comprehend fully. Moved nowhere. 15089. October 28, 2017 KRING Eric KRING set multiband delay and dispersion to magic blanks when a polarization was missing. This apparently causes the calibration routines to flag all polarizations rather than just the missing one. Changed KRING to write zeros for these (like FRING). Changed solution weights from -1.0 to zero although this does not change the behavior of CLCAL or calibration application. Moved nowhere should be a patch. 15090. October 29, 2017 GNU compiler Eric Changed $SYSLNX64 versions of FDEFAULT.SH, LDOPTS.SH, and CCOPTS.SH to use the appropriate gcc and gfortran compilers here in Socorro. Also changed the Makefile portion of XAS.SHR. NOTE: these are NOT the /usr/bin versions of RedHat 6 which are back at 4.4. I am using version 6.4 for the local LNX64 binaries which are also those distributed for the binary installation. Moved nowhere. 15091. October 30, 2017 AIPS Memo 123 Eric New memo about the suite of tasks that may be used on the new large DiFX pulse-cal tables. Moved everywhere. 15092. November 2, 2017 DATPOL Eric gfortran needs this to be NOOPT apparently. Changed OPTIMIZE.LIS. Moved nowhere. 15093. November 7, 2017 AIPS Memo 123 Eric The PostScript file contained the $Id string associated with the pstricks macro package. The MNJ will change these evey night and so copy the PostScript file as changed infinitely. Moved everywhere. 15094. November 8, 2017 RMFIT Eric RMFIT: 1. Added the option to compute and plot P=sqrt(Q*Q+U*U) and angle=atan2(U,Q) / 2 alternatively to Q and U. Model plotting requires computing arrays of Q and U and then plotting P first, then A. 2. Added P&A to PIXRANGE function and manage P&A scaling separately from Q and U. 3. Made an alternative initial guess being the last fit RM and thickness with the amp and phase added from the current spectrum. RMFIT.HLP: Corrected remarks about BLC and TRC and added option S (SHOW Q&A or SHOW A&P on the TV) to the menu seen during fitting. HLPRMFIT.HLP: Reworded the PIXRANGE explanation and added explanations for SHOW Q&U and SHOW A&P. AIPS Memo 118: updated 2 figures, reworded, changed coloring, and added a figure of P&A. Moved nowhere. 15095. November 13, 2017 FARS Eric Added routine to write slice files of the real part, imaginary part, amplitude, and phase of the RMTF (beam) attached to each output image. Dropped the TABINI copyright section in the special FARS version of TABINI. Moved nowhere. 15096. November 16, 2017 FQUBE Eric FQUBE did not like having FQ numbers that did not match the current image (e.g. one out of SUBIM with a restricted window inthe FQ axis or an increment in that z=axis > 1). Fixed it rather simply to handle such things correctly. Moved nowhere. 15097. November 20, 2017 AIPS Memo 118 Eric Fixed various typos and poor wordings. Moved everywhere. 15098. November 20, 2017 doc files Eric Update the tab completion file HLPIT.LIS, the apropo file LSAPROPO.HLP, and about lists (ZZ*.HLP) for ADVERB CALIBRAT EDITING INTERACT OOP PLOT POPS SPECTRAL TASK TV-APPL UV VLBI Moved nowhere. 15099. November 21, 2017 RM2PL Eric Clarify in the help file what appears in OUTTEXT. Add option to plot and print total polarization and angle instead of Q and U. Moved nowhere. 15100. November 22, 2017 CookBook Eric Changed chapters and updated html and COOKBOOK.PDF 3 for ABORTASK, WAITTASK 4 for FGSPW 6 for PCHIS, PLOTR 7 for FARS, RM2PL 9 for additional pulse-cal tasks and memo 13 for updated ABOUT lists E for FGSPW L to include some spectral-line considerations 0,I,J to match the above Moved everywhere ($AIPSPUBL) 15101. December 10, 2017 polarization Eric Changed - keep in directory ~/AIPS/polar DPDC.INC to carry PANGLE plus subarray and xy->qu translation indicator (TRXYQU) SELINI to init parameters in DPDC DATPOL to put PANGLE in common FNDPOL to test XXYY coversion to true stokes in a manner similar to RRLL DGINIT Filled in dummy JADR and SELFAC coded plus TRXYQU=1 for XY -> I, Q, U, V. Note all 4 linear are required for a correct translation. One can get an approximation of I from XX and YY only and that remains supported (TRXYQU <= 0). DATGET Call DATPAR when DOPOL false and TRXYQU true to make sure parallactic angles are current. DATPOL does this on DOPOL true. Call DXYGET rather than DGGET on TRXYQU>0. DATPAR New routine to update antenna info if subarray changes and update source and parallactic angle info if source changes or time changes by > 1 sec. DXYGET New routine to do a full translation of XX/YY/XY/YX to needed parts of I/Q/U/V. Change LXYPOL, ORIPOL, VLBPOL, UVHOL to use PANGLE in the common rather than locally. Call sequences of the first 3 changed, dropping PANGLE. Moved nowhere. 15102. December 11, 2017 LUNTMP Eric Add argument 3 to do a new text file LUN not using any of the standard ones. Changed HELPS for RUN to use this - the RUNfile may contain GO or DEFAULT or ?? that will use the standard text LUNs so we must not use them for the nested RUNs. Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC17 is NEW *********************************************************************** *********************************************************************** 15104. December 14, 2017 UVHOL Eric Corrected DATPOL version in UVHOL for correct call sequences to LXPOL and friends. Added APARM(10) option to apply the normal parallactic angle correction, default is not to do it. Changed OPTIMIZE.LIS to not optimize this task. Moved to 31DEC17 this date. 15105. December 15, 2017 HOLOG Eric Added adverb REFREQ to give the reference frequency for the image spacing and DOVECT to choose between amplitude/phase and real/imaginary. In Fortran added arrays for real and imaginary. Changed U/V coordinate labels to L/M. Changed GRID to fill in real and imaginary and blank them. Changed to honor DOVECT for real/imaginary when requested. Try to pick up the Stokes from the INFILE name and if found add a Stokes axis to the output images. Use REFREQ to scale images but save the actual frequency to put in the headers. 15106. December 15, 2017 ANINI Eric AN tables may need NIF keywords plus the 22 or so that occur always - PCAL requires this. Changed ANINI to ask for this many keywords - others IF 30 led to PCAL failing. Moved to 31DEC18 December 18. 15107. December 18, 2017 PCNTR Eric Changed it to draw vectors even when the axes are not a position pair - with a warning. Moved to 31DEC17 this date. 15108. December 18, 2017 COMB Eric Changed to set Stokes to Q, U, or V when doing ratio with I image. Moved to 31DEC17 this date. 15109. December 20, 2017 TVFLG, SPFLG Eric Changed TVFLG, SPFLG, FTFLG, GRIDTB.FOR to support STOKES = 'IQU' and 'IQUV' Moved to 31DEC17 this date. 15110. December 21, 2017 PCNTR Eric For the special case of axes = 'L' and 'M' with a positive increment for L, change the sign of the plotted polarization angle. Moved to 31DEC17 this date. 15111. December 21, 2017 HOLOG Eric changed to get the Stokes from the data file - using the file name depends too much on the user. Moved to 31DEC17 this date. 15112. January 2, 2018 VLANT files Eric Created the $AIPSTARS file EVLA.ant.2017 from the web location maintained for the EVLA and the dummy file VLA.ant.2018 to handle very old data. Moved everywhere. 15113. January 2, 2018 DXYGET Eric The subroutine to change linear polarization into true Stokes had addressing errors that caused the output to compress all flagged channels. Moved to 31DEC17 this date. 15115. January 9, 2018 linear polarization Eric The parallactic angle correction done for circulars must not be done for linears. We do it later in DXYGET when converting to Stokes. Changed: DGHEAD To add 10 to any DOPOL > 0 if the input file is linear. POLSET To do MOD (DOPOL,10) when testing DOPOL. VLBPOL To do MOD (DOPOL,10) when testing DOPOL. DATPOL To use MOD (DOPOL,10) and to omit the parallactic angle change if DOPOL > 10 (set by DGHEAD) or if the output polarization is linear. Moved to 31DEC17 (patch) this date. 15118. January 16, 2018 Batch Eric Both QMNGR and AIPSB had the wrong call sequence to subroutine BATQ - they lacked the host parameter. I know I tested this to allow remote machine batch jobs. Somehow, the changed programs did not get putbck. Moved to 31DEC17 this date. 15119. January 16, 2018 CALIB Eric The self-protective code in CALIB did not allow a multi-source file to use SMODEL even when properly specified (some SMODEL not zero, no (2nd) model image, only one cal source). Added code to recognize the proper specification and allow it. Moved to 31DEC17 this date. 15123. January 19, 2018 UVSUB Eric The gridded version of OPCODE 'MODL' only included the last field. This was due to a scaling factor which applies to the input visibilities in field 1 but then applies to the sum of the model for fields > 1. Fixed ALGMEM and ALGSTB to set this factor to 1 for fields > 1. Moved to 31DEC17 this date. 15124. January 20, 2018 UVSUB Eric The fix I did yesterday does not work fully. Changed ALGSTB and ALGMEM to zero the input data when MODL is requested before those data are written to the pseudo-AP. The Q routine then adds the scaled model to the data always. Moved to 31DEC17 this date. 15133. February 21, 2018 UVPLT Eric A counter which is tested to see if the data may be self-scaled was a mere INTEGER which therefore could roll over to negative a 2^31-1. Changed it to double precision. Also changed the max color to 0.97 rather than 1 to get a more visible deep blue. Moved to 31DEC17 this date. 15138. February 28, 2018 SETJY Eric It was quoting errors bars for no valid reason for P band at least with the PB 2017 fluxes. Moved the setting of error to 0 to before the 2017 computation. Moved to 31DEC17 1 March. 15151. March 14, 2018 gridded models Eric The correction made in January to ALGMEM and ALGSTB to compute models correctly for multiple facets caused the data weights to be zero. Corrected this by putting an intelligent zeroing of the input data (leaving the weights alone). Discovered that the zeroing of the input array when data fall off the grid was done with bad lengths and addresses. Corrected these too. UVSUB made an error in the HI file. This error affects OPCODE='MODL' only. Moved to 31DEC17 this date. 15145. March 8, 2018 gridded models Eric Changed CCSMEM to check the AP memory and ask for enough - somehow it survived when the images were moderately small. But it needs in a larger case a lot for the grid plus what was already allocated for the UV data. Moved to 31DEC17 on 4 April. 15146. March 8, 2018 ANINI Eric When over-writing an antenna table, as in VLANT, the number of keywords allowed would be <= 51 unless PCAL had been run. The code assumed that PCAL had been run (or was running) and so could demand more if the number of IFs was > 26. Changed to check if the file is pre-existing and accept the number of keywords in same. 15148. March 12, 2018 ANINI Eric The "fix" of March 8 made a requested version of 0 (i.e. make a new file) into version 1 and then managed to destroy it. This almost certainly would happen only in inadequate AN tables that looked as if they needed reformatting. Moved to 31DEC17 on 4 April. 15159. April 2, 2018 FITLD Eric The program tested for duplicate FQIDs but used bad logic to substitute a new FQID number for the old. Corrected the logic and improved the messaging. 15160. April 3, 2018 FITLD Eric Yesterday's change fixed the UV data, but did not address the tables. Easy changes in the writing of PC, TY, IM and GC tables were implemented. The AT table needed to have FQIDs dropped if they were no longer used. The MC table required the FQIDs to be replaced with the corrected ones. Two new subroutines more or less cribbed from DFQID were added to FITLD to perform these chores. Moved to 31DEC17 4 April.. 15162. April 9. 2018 DBAPP Eric First serious test in a while. Fixed 1. Warn that UT1UTC differs in the antenna headers but do not quit. 2. When testing the times, include any difference in day number. 3. The DUVH common changed some time ago and the parallel versions in DBAPP did not keep pace. The data got trashed as a result. Moved to 31DEC17 this date. 15163. April 11, 2018 DBAPP Eric One more issue: the NX table needed the addition of any day offset to the times of the input NX table and the start vis number for the NX table had to point at the last of the input version of the output table rather than 0. Moved to 31DEC17 this date. 15167. April 16, 2018 FQUBE Eric FQUBE was writing N CG tables for N input images when there was no input CG table. It should only write one version, filling it with N lines. Moved TO 31DEC17 23 April. 15169. April 24, 2018 FITLD Eric The change made in March 2017 to have ANTNAME honored when concatenating did not work in complicated circumstances. I have returned FITLD to initializing ANTNAME to blank when concatenating. It sets antenna names/numbers from the existing AN table and can add new ones if encountered. Moved changes only to 31DEC17 this date. 15170. April 27, 2018 DOBAND 3 Eric The interpolation of bandpasses attempted to use the one good one when the other was bad at the upper levels of the code. But down at the bottom BPGET blanked the bandpass to be applied whenever either of the 2 was blanked. Changed it to use the good one if there is a good one. Moved to 31DEC17 this date. 15175. May 8, 2018 model subtraction Eric When VISDFT attempted to run with data having all 4 polarizations, it put the pointer telling whether to compute the model or to subtract the model on top of the factor for the first cross-hand polarization. This often led to the model being computed rather than the data-model. Changed the addressing in VISDFT, QXXPTS, QGASUB, and QSPSUB to do this right. Moved to 31DEC17 20 June. 15176. May 9, 2018 model computation Eric The zeroing of the input data neglected the possibility of compressed uv data. Corrected ALGMEM and ALGSTB to handle compressed data as well. Moved to 31DEC17 20 June. 15193. June 26, 2018 DOBAND 3 Eric The gfortran LNX64 compilation of BPGET in optimized mode produced wrong answers. Changed OPTIMIZE.LIS to compile this routine no-opt and also changed another oddity having to do with interpolation in the presence of blanked solutions. Both issues caused loss of data. Moved to 31dEC17 this date. 15206. July 16, 2018 DOPOL Eric Changed POLSET to look for magic blanks in the solution and pass them on in POLCAL properly. Changed DATPOL to look for these blanks and to flag appropriate visibilities if they appear. Also changed the specialized routines LXYPOL, VLBPOL, and ORIPOL to handle blanked D-term solutions. They are more complicated. Moved to 31DEC17 this date. 15222. August 10, 2018 OOSUB Eric OOSUB attempted to read multi-source files to do its thing. The standard OOP code however caused the output file to lose the FREQSEL random parameter, so the file being written was larger than the file that was created and had the wrong structure. Changed the type of open done on the input and moved the close to after the output file creation. This fixes the problem. Added note to help file informing the user that multi-source files will be allowed, but that bad things will happen if more than one source is actually in the file. Moved to 31DEC17 this date. 15230. August 17, 2018 AIPS Eric In AU2, there was an error when the number of adverb values exactly fill N disk records for the TD file used to communicate with tasks. In general, the TPUT/VPUT file would not also have this condition, but when the fTD condition was met the last of the TPUT or TGET file was not written. RLDLY hit this case when I aadded BPARM. Moved to 31DEC17 this date. 15258. October 3, 2018 RM Eric Despite the help file claiming that the task could do 65000 frequencies, it could only do 5 due to the size of the spectral row buffer! Made that buffer suitably large. Moved to 31DEC17 this date.