;----------------------------------------------------------------------- ; Copyright (C) 2015-2017 ; 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 31DEC16 is TST *********************************************************************** *********************************************************************** 14741. November 26, 2015 PCAL Eric LUN 28 over used - changed all 27 and 28 to other values. LUN 28 already open was reported. Moved nowhere. 14742. November 30, 2015 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 31DEC15 instead of 31DEC14 - also fix dates in install.pl. Moved nowhere. 14743. November 30, 2015 install.pl Eric Corrected grammar in my recent if $ARCH statements. Moved to 31DEC15 this date. 14744. December 1, 2015 FDEFAULT Eric Changed the options for gfortran to those I would recommend. The -Wall makes a message for every integer = float statement for no good reason, so dropped it among others. Moved to 31DEC15 this date. 14745. December 1, 2015 install.pl Eric Add warning message to changed section re laptop. Moved warning to 31DEC15 Dec 6. 14746. December 6-7, 2015 fix_aips_elcap Eric The Xwindows libraries had an extra close parentheses in their names for no good reason. The nice loop that Josh wrote gave a file name of *.dylib for one of the links so changed to make it do the 3 files by the names they are known to have. Dec 7: Josh wrote a proper test on Xwindows rather than linking the files bu force. His loop looks okay but gives odd results on El Cap so left my explicit file linking. Moved to 31DEC15 this date. 14747. December 8, 2015 EDIT class Eric Changed QEDIUTIL at NEXT ANTENNA. It failed when there was only one antenna other than the edited one and that antenna was the next antenna. Now it will start upping the second antenna number to one more than the edit one. Moved to 31DEC15 this date. 14748. December 11, 2015 CLVLB Eric It treated the fact that one CL record was below the horizon as a fatal error. Changed to count the below horizon indicators and report that number plus the total at the end. Moved to 31DEC15 this date. 14749. December 15, 2015 install.pl Eric Add copies of fix_aips_alcap.sh to $SYSLOCAL and $AIPS_ROOT. Moved to 31DEC15 this date. 14750. December 16, 2015 UVFLG Eric The text-file reading got confused about times which it set to minus the day number. Then it tested if both were zero. Added test for both = minus the day number and changed the complaint to allow time1=time2 which should not occur. INTEXT reading did not re-initialize all adverb values between commands leading to later commands inheriting parameters from the previous ones. Simplified the re-initialization to be all inclusive. Moved to 31DEC15 this date. 14751. December 16, 2015 DAYNUMBR Eric NEW VERB: prints the day number (in the year) of the observation date. Changed POPSDAT.HLP, NEWPARMS.001, AU3A, and made a new help file. Moved DAYNUM from $APLNOT to $APLSUB and corrected it to add the 400 year and 100 year provisos although neither matter. Moved nowhere. 14752. December 17, 2015 ABOUT lists Eric Updated HLPIT.LIS and the about files for AP, CALIBRAT, CATALOG, PROCEDUR, SPECTRAL, TASK, UTILITY, VERB, and VLBI. Moved some to 31DEC15.. 14753. December 18, 2015 PCAL Eric Found out that data points with some but not all polarizations flagged would persist in the data selection and cause problems including the infamous 'POLARIZATION MODEL NOT PROVIDED' message which caused the task to stop. Moved to 31DEC15 this date. 14754. December 18, 2015 PCAL Eric Changed to do linear polarization model only when the data are linear and CPARM(5) > 0. That model does not seem to work and circular (APPR) does when the source model has zero for polarized flux. Moved to 31DEC15 this date. 14755. December 18, 2015 UVHOL Eric A choice of text LUN=89 was unfortunate - changed to 91. Fixed it to do I only with linear polarizations in data set. Moved to 31DEC15 this date. 14756. December 21, 2015 CookBook Eric/Amy Updated chapters: 3 : PRTMSG (OUTPRINT, line count check), re-word tape section 4: TVFLG LOAD LAST PIECE, SOUSP fix SNs, EDITA DO3COLR. de-emphasize tape, drop UVRANGE table 5: antique wording, OBOXFILE temp fille, DO3COL in EDITR/EDITA, reduce priminence of mag tape 6 : Memo 119 (TVSAD), Memo 118 revised, TVHLD, TVMOVI (DOALL), UVIMG, XG2PL, RM2PL, XGAUS (baseline setting) Downplay magnetic tapes at the start 7 Memo 119, TVHLD, MCUBE forgiving, STACK, Memo 118 revised, XG2PL, RM2PL 8. MCUBE forgiving, TVMOVIE DOALL, Memo 118 revised, UVFRE 9. ACSCL, DOVLAMP, DIGICOR in FITLD, dispersion in MBDLY and FRING, RLDLY, TYCOP, DELZN, EDITx CROWDED with color Overhauled for VLBARUN/VLBAUTIL new cal method and to modernize 13. Update C Overhauled for VLBARUN/VLBAUTIL new cal method and to modernize E. FRING dispersion, RLDLY SN tables, fix antique wording places, FRING divide in > 2 parts, TVFLG groups, SOUSP SN tables, UVRABGE -> break IFs into separate files, SYCOP renamed TYCOP Moved nowhere. 14757. December 22, 2015 VPFLG Eric Added calls to make NX (index) table on the fly. Moved to 31DEC15 this date. 14758. December 22, 2015 IRING Eric The geometry was questioned by a user and was reviewed in detail. The radii were increased erroneously by a division by cos(inclination) but the rest of the geometry was found to be correct. Changed the task to have OUTTEXT='NONE' suppress both the text and message file outputs. Also changed it to honor the Y axis range in CPARM(9,10) when CPARM(8)>0. Previously it honored it only when the values did not exceed the stated range. Moved to 31DEC15 this date. 14759. December 28, 2015 PBEAM Eric Changed the task the handle data in any order. Previously it demanded that the data lie in columns on the sky, although there was some confusion with a row-ordered plan also. Changed the plots to be contour plots and to write out the data. model. and residual images. Previous plots were row traces. Corrected a minor issue in EVAUV's contouring. Moved to 31DEC15 this date. 14760. December 28, 2015 AIPSLetter Eric Put in LET15B TEX, PS, and PDF files in $AIPSPUBL. Moved everywhere. 14761. December 29, 2015 PBEAM Eric Found a variable that was not initialized which caused the task to fail to read the data the second time in which the normalization was supposed to have been done. Added APARM(9) to control the width of the convolving function which was too narrow before. Re-named the axes and made the X axis right handed. Made less drastic contour intervals for the residual image and made the data sample markets slightly bigger. Moved to 31DEC15 this date. 14762. December 29, 2015 HOLOG Eric Correct the amplitude image by the pixel-dependent sum of weights making the image be an interpolated one rather than a convolved one. Also normalize that image and make unsampled pixels be magic blanks. Moved to 31DEC15 this date. 14763. December 31, 1015 holography Eric Further changes: UVHOL Was limited to documenting only 14 averaged antennas. Changed to write more than one line (like ref ants) HOLOG Changed to subtract the phase at the pixel of peak amplitude from the phase image, thereby "normalizing" phases as well as amplitudes. PBEAM Changed parsing of text file to free format and added parsing of the Averaged antennas and Averaged RefAnts input lines. Changed display of antennas and refants to support the full (possible) lists. Moved to 31DEC15 30 Dec. 14764. December 31, 2015 PBEAM Eric Made changes not carried over to 31DEC15: 1) Dropped lots of unused COMON variables 2) Changed most of the computation to double precision since the dynamic range of is so great. 3) Dropped unused PIXRANGE adverb and added DOPLOT which must be set to get plots. 4) Added IRING-like plot for the beam and model images, plotting them on the same plot for comparison. A second plot of the cumulative sum inside the radius was also added. 5) On the 3rd DOPLOT bit, optional plots of the difference of the two IRING plots and the difference divided by the model were also added. 6) Scaled the Fit parameters by FITPAR(1) not just for the printout but also for computing the model. Otherwise the model did not have a peak of 1.0. 7) Scaled the data and model interpolated images to have peak 1.0 and changed the residual image to be the difference of the two normalized images. Previously the difference image was the interpolated image of the data minus model and so lacked all of the normalization. 8) Fixed it to turn off the graphics planes between plots on DOTV true. Otherwise a plane used in one plot would still be present in the next plot. Moved nowhere. 14765. December 31, 2015 year-end Eric Copied the EVLA.ant.2015 file from the VLA/baselines data area and put it into $AIPSTARS. This has the antenna corrections made during 2015 for the EVLA. Also added dummy VLA files for 2015 and 2016. Moved everywhere (TEXT area available to all). 14766. January 3, 2016 PBEAM Eric Fixed up some MAX and MIN statements that mixed real*4 with real*8. Moved nowhere - bug in 31DEC16 only. 14767. January 5, 2016 FRING Eric Changed the fitting of dispersion to allow more than one group of IFs each with its own delay but sharing the same dispersion. Moved nowhere. 14768. January 7, 2016 GC tables Eric Changed PGCV.INC back to allow up to 200 values (per polarization, per IF, per antenna) for the Y and gain. Some correlators write out 200 values (almost all magic blanks) even though they have no use for them. Moved to 31DEC15 (patch) this date. 14769. January 7, 2016 tables Eric IMLOD should be able to read a FITS file containing only a table. Trying this on a source catalog table, the following changes were required: FITLD Limited the number of columns to 60, should be 128. MAKTAB Limited the number of columns to 60, should be 128. TABHDR Limited the column number of parsed cards to 99, should be 127. Doing these corrections, IMLOD loads the catalog as a UK (unknown) table without complaint attached to a dummy image. FITLD loads the table as well but complains loudly about the lack of an array geometry table. UVLOD dies because a required header card is not present. If we encounter tables with more than 127 columns, a special table reader will be required to load some but not all of the columns. Moved nowhere. 14770. January 12, 2016 PBEAM Eric Corrected a subscript error that made the addition of RR and LL to make Stokes I get scaled incorrectly. Also fixed a format issue telling the user the re-scaling factor. Moved fixes only to 31DEC15. 14771. January 13, 2016 RLDLY Eric The output SN table was written to "CLOVER" which could be set to some non-zero value if a new CL table was also written. Fixed to always write a new SN table and improved the history writing as well. Moved to 31DEC15 this date. 14772. January 15, 2016 FITLD Eric Two subroutines GNINI and TABGN were not compiled when PGCV was changed. They include an include that includes PGCV and so have to be compiled. Then FITLD needs to be relinked. Moved to 31DEC15 this date. 14773. January 19, 2016 CATALOG Eric Changed the format for catalog number and user numbeer and date in UCAT, MCAT. PCAT, CATALOG, etc. verbs. There should never be overflows now. Moved nowhere. 14774. January 19, 2016 EDITA, EDITR Eric Changed QEDIUTIL to make the all-source status false to begin with. Corrected the expansion of times in flags to avoid going a long way in one direction due to a gap in the selected data in the work space. Moved to 31DEC15 (patch) this date. 14775. January 20, 2016 TVSPC Eric NEW TASK: Plots spectra from one or two transposed cubes at coordinates selected by cursor on a grey-scale image (e.g. moment-0). The menu allows the last plotted spectrum to be fit with up to 4 Gaussians and saved as a slice (SL) file. Works with e.g. an I image and Q and U spectral cubes or with a moment 0 image and I and V spectral cubes. Includes help with an explain section and real-time help. TVLAB was changed to avoid errors on empty strings and to honor NTEXT/TEXT even with LABEL >= 7. Moved nowhere. 14776. January 20, 2016 HOLOG Eric Added NORMALIZ adverb to control normalization of gridded amp and phase images. Moved nowhere. 14777. January 21, 2016 VLBAUTIL, VLBABPSS Amy Fixed BPASS inputs, main one is use set ICHANSEL to entire band as recommended by Craig. Moved nowhere. 14778. January 21, 2016 DOVLAMP Amy Added UVCOP to DOVLAMP for the case where not all IFs are needed for VLBI calibration. Only works if IFs are contiguous. Moved nowhere. 14779. January 22, 2016 TVSPC Eric Changed APARM(5) to be fraction of split area for spectra, APARM(7) for smoothing support/FWHM. SET APARM gets the new question and is rearranged. Bottom line on spectra plots gets the radius over which averaging was done. Changed HELP and run-time help. Moved nowhere. 14780. January 22, 2015 AIPS Memo 120 Eric New memo detailing TVSPC with 3 example data sets (HI, OH Zeeman, and polarization). Moved nowhere. 14781. January 28, 2016 DBCON Eric DBCON reinstated flagged table rows. Changed to drop them completely. Number of output records had to be fixed in output table header. Moved tp 31DEC15 (patch) this date. 14782. February 1, 2016 TRANS Eric Changed it to allow blank and comma separators in TTRANSCOD. Also changed it to refuse to put the same axis out twice and to require that all ``real'' (> 1 pixel) axes be specified. Updated the help file and the TRANSCOD help file. Moved nowhere. 14783. February 1, 2016 IMLOD et al. Eric Changed the FITS table header reading routines to limi ITNCOL to 127 but to report excessively large tables without quitting. Changed TABHDR, TABHDK, and PRTTP. Moved nowhere. 14784. February 1, 2016 UVHOL Eric Copied DATPOL into this task and commented out the parallactic angle correction, which should not be done to holography data. It rotates things from telescope to sky but in holography we must stay with the telescope frame. Moved nowhere. 14785. February 2, 2016 documentation Eric Updated the TAB completion file HLPIT.LIS and the ABOUT files on ANALYSIS, IMAGE-UT, INTERACT, TASK, TV, AP, CALIBRAT, OOP, PROCEDUR, and UV.. Moved nowhere. 14786. February 3, 2016 MNJ Eric Changed MNJ to use rsync for binary installations rather than CVS. Getting CVS in Macs is a pain. MAKE.BMNJ Add comment about rsync, say cvs useful. Ask for cvs path only once, do cvs priming only if CVSLOC is an actual file. AIPSUPD Add comment about using rsync, add code to rsync the UPDATE area mostly to get LASTCVS.DAT to set the dates. Special work only if neither UpdCVS nor UpdRsync are true. UPDCONFIG Make UpdRsync and UpdCVS have opposite initial values based on whether or not it is a binary installation. UPDCONTROL Special work only if neither UpdCVS nor UpdRsync are true. UPDRSYNC Reposition the directory to which it cd's. Change existing binary file rsyncs to match and add rsyncs for code (AIPS, APL, INC, Q, QY, Y) and control/doc files (DOC, HELP, HIST, RUN, SYSTEM). Moved nowhere. 14787. February 5, 2016 install.pl Eric Changed error messages around the make of XAS to point out that fixing Makefile and doing a make in $YSERV/XAS is all that is needed. Also added $CCOMOPT in the rewrite of the Makefile before the make. Moved nowhere. 14788. February 5, 2016 web pages Eric Did some work on the MNJ web page (mnj.shtml), a serious overhaul on the Mac web page (Mac.shtml), and went though the installation guide (install.shtml) carefully. All should now reflect modern reality. Also install_changelog.shtml. Moved to the official aips web site. 14789. February 9, 2016 TVFLG, SPFLG, FTFLG Eric Changed all 3 tasks to allow STOKES = 'FULL' to mean all 4 polarizations at one time. Fixed also the 3 help files and the 3 run-time help files. Moved nowhere. 14790. February 9, 2016 DTSUM Eric Typo meant that it did not interpret the new UV format correctly and assigned all baselines to the same autocorrelation. Moved to 31DEC15 (patch) this date. 14791. February 12, 2016 PBEAM Eric Changed Fortran to have APARM(10) control whether the fit parameters are normalized (G0 = 1.0) or not and to have the residual image contoured at intervals of 0.01 rather than 0.02. Changed help for these and to change the explanation of APARM(7) to reflect what it currently means (switch X and Y if the first column is vertical offset). Moved nowhere. 14792. February 15, 2016 FITLD Eric Apparently the WR table (IDI weather) has added a WIND_GUST column. Fixed Fortran so that this does not cause an abort. Did not change the AIPS weather table to include this column. Moved nowhere. 14793. February 15, 2016 FRING, RLDLY Eric Changed the second test on frequencies being all in the same order so that it generates a warning but does not die. This means that two IFs can overlap but we cannot have IFs going in one order with channels in the opposite order. RLDLY somehow did not get these tests last year even though it was supposed to. Added them now with the second being only a warning. Moved nowhere. 14794. February 16, 2016 WX table Eric Added two columns now supported by DiFX in IDI: WIND_GUST in m/s and PRECIPTATION (cm since midnight). Changed DWXV.INC To add 2 columns and pointer parameters WXINI Add 2 columns TABWX Changed call sequence, use pointers carefully TABSEL Change TABWX call sequence TABLEUTIL Change call sequence of OTABWX and call to TABWX. CLCOR Change TABWX call sequence FILLM Change TABWX call sequence (2 more 0's) FITLD Change TABWX call sequence, change TABWR, WRINI, and WR local include INDXR Change OTABWX call sequence APCAL Change TABWX call sequence WETHR Changed to support all possible columns plus the computed ones. Corrected bad subscript in flagging routine that caused it not to take absolute value when it was supposed to. Fixed one variable which needed a SAVE between plots. WETHR Help - added new OPTYPEs. Moved nowhere. 14795. February 16, 2016 OOCAL Eric NEW PROCEDURE IMSCAL (image self-cal) to split a large CASA image into facets with CC files and then do OOCAL (divide by model in UVSUB, then CALIB the divided data set, copy SN back to original UV data set). Then zap the facet images. New HELP file for IMSCAL. Moved nowhere. 14796. February 16, 2016 PBUTIL Eric Changed the primary-beam utility OOP package to read the full spectral-index and curvature images into dynamic memory. Then I/O is not needed to find each correction parameter. This vastly improved the performance of the step which recomputes CC tables (in one test 6-8 minutes went to < 2 seconds!). Moved nowhere. 14797. February 16, 2016 AIPS Memos 114, 117 Eric Updated the memos for changes to weather table. Moved everywhere (in $AIPSPUBL). 14798. February 16, 2016 PBEAM Eric Added a data-model column to the optional print out and corrected the BESSEL computation (needed real*4 not *8). Also added a print out of the average residual. Compute the residual image from grdding the actual data samples - model rather than difference the two smoothed and separately normalized data and model images. Scale the residual image by the average scaling of the two images (only not a normalization). This gets rid of a negative bias apparently due to the different normalization scalings. Moved nowhere. 14799. February 16, 2016 EDITA, SNEDT, EDITR Eric The DO3COL routine did not take BIF into account when allocating colors and so left colors zero for high numbered IFs. Also changed the coloring when showing only 1 IF but 2 polarizations so that 2 distinct colors will be used. Moved nowhere. 14800. February 19, 2016 UVFLG Eric Added BFREQ/EFREQ keywords to the text file input. UVFLG will translate these into zero or more sets of BCHAN/ECHAN/BIF/EIF to match the current data set. This will allow text files to specify the standard RFI signals at the VLA for example to be used on whatever data set you observed. Moved nowhere. 14801. February 19, 2016 UVHOL Eric Add DOCAL, DOPOL, ICHANSEL to precusor comment cards written with 'HOLG' mode. Moved nowhere. 14802. February 22, 2016 LISTR Eric When the last record in a "gain" file (SN, CL, TY, SY) had a time greater than the records just before it, the task's table record logic caused that record not to be printed. Changed the logic to deal with this situation without messing up the normal case where the last record is the same time as at least one preceding record. Moved nowhere. 14803. February 23, 2016 SPLIT, SPLAT Eric Added the ANTENNAS adverb (only) to both. SPLAT is often used to do a few once and for all things to a data set and dropping bad antenna(s) should be included in that. Moved nowhere. 14804. February 24, 2016 TGET Eric the test in AU2 on disk file size was for the old format (305 max tasks, 4 records each) while the new format (508 tasks, 8 records each) needs much more. Fixed the test. Moved nowhere. 14805. February 25, 2016 APCAL Eric TABGC only fills in YVALGC and GAINGC to the number of terms expected. The routine that tried to avoid being upset over duplicate GC records however checked the full amount allowed in the GC table. This led to apparent differences that were not real. Changed to check only to the number of terms for the particular antenna, IF, and polarization. Moved nowhere. 14806. February 27, 2016 FTFLG, SPFLG, TVFLG Eric When changing between Stokes to be displayed, the selection of string between NOXX and NOYY was reversed. The actual flags 1011 and 0111 were correct. It was fine for the other polarizations, just not linear. Error arose when 4 polarizations at the same time was added (2/9/2016). Moved nowhere. 14807. March 1, 2016 PBEAM Eric Changed the IRING difference plots to use the IRING found from the real difference image. Added plot of the IRING interior sum of the difference image and changed the difference/model to use the real difference rather than the difference of the data and model IRINGs. Fixed an issue with string placement in these plots. MOved nowhere. 14808. March 7, 2016 RFLAG Eric Added YINC adverb to pre-average data in time before applying the algorithms. This will allow the excision of RFI at somewhat lower levels. Also fixed the default XYRATIO to 1.414 for plot files. Moved nowhere. 14809. March 9, 2016 TYAPL Eric Changed to have a local version of COPTAB, stolen from UVCOP, applying flags to SY, TY, and SN tables when FLAGVER > 0. Added DOFLAG adverb to suppress this action if desired. TYAPL, like UVCOP, allows 10 times as many flags per time as the normal task. Moved nowhere. 14810. March 14 & 15, 2016 TYAPL ERic Unlike UVCOP, TYAPL uses UVGET and so closed the flag table before the final routines. Added an open of the flag table to COPTAB. The FQ table was not copied by the previous, so allowed it to be copied simply since there is no IF selection here. Moved nowhere. 14811. March 16, 2016 UVPRT Eric The setup for STOKES left out a section for XXYY pols. Moved nowhere. 14812. March 25, 2016 UVGIT Eric NEW TASK version of UVFIT with IMFIT's math method instead of JMFIT's. It converges correctly in some cases where UVFIT returns the initial guess. Moved nowhere. 14813. March 31, 2016 CURVALUE Eric Changed the AIPS-task version of TVALUE and the local versions in XGAUS, RMFIT, ZEMAN, and TVSAD to handle pixel values > 9999. It does so by dropping the X= and Y= parts of the pixel position string. Moved nowhere. 14814. March 31, 2016 ANFIND Eric Added ASKAP as having the same XYZ signs as ATCA. Moved nowhere. 14815. April 15, 2016 PCAL Eric The baseline pickup for one of two data copies had the handling of the new antenna format wrong! IA2 was not set so I do not see how it could work. Moved to 31DEC15 this date. 14816. April 16, 2016 PCAL Eric Changed the handling in RPCALC and LPCALC of times for which the I model has value 0. It reports it but keeps going so long as the weight is also zero. This was caused by the ionospheric Faraday rotation being blanked for reasons I cannot discover - probably relating to flagging being applied to the CL table. I added DPARM(1) option to tell the task not to flag data when the IFR is blanked. The 2 antennas affected for some reason get different (similar) solutions with and without this option but the others are essentially unchanged. Moved nowhere - I am patching something I do not understand. Moved to 31DEC15 29 April without DPARM. 14817. April 27, 2016 UVMOD Eric The FLUX adverb has always been confusing. It is a scaling factor times the individual sample's rms - but that is meaningless if one is ignoring the input data (FACTOR=0). So changed it to use FLUX as noise in Jy when FACTOR is zero and added explanations to the help file. Moved nowhere. 14818. April 28, 2016 DTSIM Eric DTPARM had an illegal read format containing '/' and it converted individual VLBA or VLA antennas as if latitude, longitude, and elevation were given for them rather than the XYZ position. GEOPOS returned a left-handed coordinate - changed to right handed (Y = -Y). Moved nowhere. 14819. April 28, 2016 CookBook Eric Added TVSPC to chapters 6, 7, 8, and 13, Added IMSCAL to 13. Redid Appendix L to modernize and be more explanatory and to do index. Updated the numbers in chapter 1 since I ran them for the AIPS FAQ web page which was updated. Update chapter 0 (T of C) and the index files LaCookI and J and the Recipe file. Moved nowhere and everywhere. 14820. April 29, 2016 PCAL Eric Found what was wrong - the Faraday rotation values were lost entirely when an IF was copied to a scratch file before pulling out individual channels. Changed it to: (1) Not delete the ION file when killing other scratch files except between IFs. Zero scratch pointers when deleting the scratch file. (2) Create a new ION scratch file when needed before the copy of the individual IF to its scratch file and write the ION as that copy is done (3) Do the ION file with the data read only when not doing spectral solutions as opposed to always doing it. Values are not available if the IF was moved to a scratch. Subroutine $QNOT/GRDSUB has to defend itself when pointers to SCRGRD and SCRWRK point at improper values (outside NSCR or disk or cno <= 0). Moved to 31DEC15 this date. 14821. April 29, 2016 MNJ Eric The new rsync-only MNJ for binary installations failed to rsync the TEXT directories (PUBL, OFM, STARS, IONS, etc.) Changed UPDRSYNC to correct this oversight. Moved nowhere. 14822. May 5, 2016 SLPRT Eric NEW TASK: prints the contents of a slice file, optionally with one of the models fit in SLFIT or TVSPC. Moved nowhere. 14823. May 6, 2016 GREYS Eric Corrected help file to mention STVERS not INVERS and changed Fortran to call the variable STVER not INVER and to turn off ST plotting if STVERS < 0. Moved nowhere. 14824. May 9, 2016 SLPRT Eric Changed the "simple" print routine to check for blanks in the slice. Added an rms computation and display when a model is printed. Moved nowhere. 14825. May 10, 2016 SNFLG Eric Changed it to loop over subarray when SUBARRAY=0. Moved nowhere. 14826. May 10, 2016 IMAGR Eric Changed QCLEAN - it had some wrong subscripts when testing islands for making automatic boxes. If IMSIZE(2) < IMSIZE(1) then islands would be found only for X values less than IMSIZE(2) - i.e.the left half, quarter, etc of the image. Moved to 31DEC15 this date. 14827. May 13, 2016 PRTAB Eric Rick printed of order 800 pages without meaning to. Added to the code a routine that counts the number of lines when DOCRT <= 0 and OUTPRINT=' '. If > 500, it forces the user to answer Y or y on his terminal before making the large printout. Put a note in the help file about this. Lots of other tasks should get this protection too. Moved nowhere. 14828. May 17-18, 2016 Print tasks Eric Print tasks need to check the lines of print when the job will go directly to the printer. This is to prevent the > 1000000 line job that PRTAB, LISTR, PRTUV, UVPRT, and others can generate. Changed ANBPL, PRTAB, PRTCC, PRTIM, PRTUV, UVHOL, and UVPRT to have separate routines to count lines of print and require the user to confirm that he/she wants any job > 500 lines. (PRTAB was done before but I messed up the parameter controlling when to return to AIPS quickly or not.) Moved nowhere. 14829. May 19, 2016 Edit Class Eric The routines used by SNEDT and EDITA to read in the table data allowed source number column values <= 0. But the flagging routines with "ONE SOURCE" would not write a flag record if no source number > 0 was found. Fixed EDGT SN, TY, SY in QEDIUTIL to force the source number to be 1 or greater. Moved to 31DEC15 this date. 14830. May 19, 2016 OOP code Eric In ZOINTD.C changed the maximum number of objects from 10240 to 102400. Moved nowhere. 14831. May 25, 2016 calibration IO routines Eric Corrected DGINIT to search only the correct number of formal-I polarization codes. It searched more but probably over addresses containing non-matching stuff. Changed UVGET which tested the number of polarization times the number of channels times the number of IFs against MAXCIF asserting the buffers will be too small. Changed the test to check this product times 3 against the buffer size. A later test will check that the user has requested less than MAXCIF for output. Moved nowhere. 14832. May 27, 2016 VLBAUTIL, VLBACRPL Amy Adjusted APARMs in VLBACRPL (simplified POSSM) to work better for VLBA tutorial. test will check that the user has requested less than MAXCIF for output. Moved nowhere. 14833. May 27, 2016 line printer outputs Eric Changed more tasks to check the number of lines going directly to a line printer and querying the user if there are > 500. Tasks changed are SLPRT, ALVPR, MFPRT, PRTSD, PRTSY, UVDIF, DTSUM, ISPEC, and RSPEC. Both Fortran and Help files were changed. Moved nowhere. 14834. May 27, 2016 DOHIST Eric A number of tasks use DOHIST = -2 and -3 for reasonable things but several used DOHIST < -1.5 to mean do not write any history at all. Changed these tasks to take DOHIST <= -9.5 to have that meaning. Tasks involved EDITA, EDITR, FTFLG, IBLED, SPFLG, and TVFLG. QEDIUTIL was changed for the first two, the remaining 4 had their Fortran changed. Help files also updated. Moved nowhere. 14835. June 3, 2016 STARS, MFPRT Eric Changed STARS to accept RASHIFT DECSHIFT coordinates when DECIMAL > 1.5. These must be in the format for RASHIFT and DECSHIFT (RA = RA_0 + Value1 / cos(DEC_0) and DEC = DEC_0 + Value2) rather than the form in columns 4 and 5 of the MF table Changed MFPRT to honor the user's DOHMS adverb value and to let DOHMS=-2 mean offsets as described above in arc seconds. Changed the determination of the component pixel from using the recorded pixel to using the linear offsets which should be immune to changes in cell size and reference pixel. Moved nowhere. 14836. June 6, 2016 IMSCAL, IM2CC Eric Changed (1) RUN file OOCAL.001 to clrtemp to avoid too many strings blowing core. Changed it to handle the unknown number of images from IM2CC. (2) IM2CC to keep only those images with actual components in the CC table. (3) PBUTIL to be more careful about trying to use more models than the arrays allow. Moved nowhere. 14837. June 6, 2016 print tasks Eric Fixed all print tasks that count lines to allow 1000 lines in batch without trying anything interactive. Tasks are $APLPGM/PRTAB $APLPGM/PRTCC $APLPGM/PRTIM $APLPGM/SLPRT $APLPGM/UVDIF $APGNOT/ALVPR $APGNOT/LISTR $APGNOT/MFPRT $APGNOT/PRTSD $APGNOT/PRTSY $APGNOT/PRTUV $APGNOT/SHOUV $APGNOT/UVHOL $APGNOT/UVPRT $APGOOP/DTSUM $YPGM/ISPEC $YPGM/RSPEC $YPGNOT/ANBPL Moved nowhere. 14837. June 10, 2016 beams Eric Updated PBCALC subroutine to support the EVLA which becomes the default if ARRAY is ' '. There are a lot more numbers since there are up to 48 frequencies within a band having separate beam parameterizations. The beam factors are found at the two closest frequencies and then interpolated to the desired frequency. Average values are used fo K, Ka, and Q bands for which the apparent frequncy differences were judged (by Rick Perley) not to be significant. No beam values are available for 4-band and the P-band ones are thought to be unreliable. Moved nowhere. 14838. June 10, 2016 PATGN Eric Added OPTYPE to specify the array type (VLA, EVLA, ATCA) rather than use OPCODE for operation and array. Moved nowhere. 14839. June 10, 2016 antenna files Eric The buffers in TABAN were too small for a significant number of IFs. Changed 200 to a parameter value which is currently 500. Moved nowhere. 14840. June 13, 2016 TYAPL Eric The new option to flag SY, TY, and SN tables had a residual error - the FG table needs to be re-opened by the table copy routine. That re-opening was done even when no flags were to be applied so FGVER was changed from <= 0 to something positive and no FG tables were copied. The task knew not to apply flags to the table data anyway. The help file lacked info about DOFLAG. Moved nowhere. 14841. June 14, 2016 POLPLOT Eric Subroutine POLVIS had a sign error in the computation of the imaginary part. This affected all ratio plots in FRPLT, POSSM, and VPLOT since 1993 when the option was first made available. Moved nowhere. 14842. June 16, 2016 UVFIX Eric UVFIX goes to a lot of work to find the actual frequency of each data sample. However, the definition of U,V,W is wavelengths at the header frequency not wavelengths at the real frequency. Changed scaling of U, V, W accordingly. Normally, AIPS keeps the offset of the first IF zero so this would not matter normally. But CASA/ALMA has no such scruples. Moved to 31DEC15 this date. 14843. June 16, 2016 OOCAL, IMSCAL Eric Changed: POPSDAT New adverb ONEFREQ NEWPARMS New adverb ONEFREQ OOCAL Help: reorder adverbs, add ONEFREQ, INVERS 0 -> 1 OOSUB Help: reorder adverbs, add ONEFREQ, INVERS 0 -> 1 IMSCAL Help: add OUTNAME OOSUB Fortran: reorder adverbs, add ONEFREQ, pass ONEFREQ as 'MODSFREQ' in object UVIN. UVMOD SAVE two frequency variables used in scaling phases (affect GFORTRAN only) OOCAL Run: add OUTNAME to IMSCAL proc, add progress messages, set INVERS and ONEFREQ before OOCAL invocation. PBUTIL Add MODSFREQ flag to subtraction and division methods, get model frequency in both, pass these down to lower level routines, ultimately to PBFACT. Changed PBFACT to compute frequency corrections to model based either on the model frequency alone or all frequencies depending on ONEFREQ. (IMAGR will have ONEFREQ false always.) Spectral index computation now uses the spectral index image frequency for reference. This will have to be the new standard for images anyway. Corrected error - the flux is 10**F not exp(F) where F is the flux correction factor. Moved nowhere. 14844. June 16, 2016 SETJY Eric Fixed typo in output message string. Moved nowhere. 14845. June 21, 2016 FITLD Eric FITLD sets up antenna files with 2 polarization calibration values per IF. It then allowed the AT file to change the number of PCALs to zero. In general this does not matter but VBGLU comes unglued when finding lots of data but no second dimension. Discovered that the AT2AN routine did not pick up any calibration that might be inthe AT file except for the first IF (it appears someone though NOPCAL was the actual NOPCAL times the number of IFs). Corrected to loop over IF should there be incoming values, to retain the presumption of 2 polarization terms, and to make sure that they are zero initially. Moved nowhere. 14846. June 22, 2016 CLPLT Eric For some reason the error bars were also converted to be betwwen -180 and 180 degress. It makes sense to do the closure phase in that way, but not the error bars. Moved nowhere. 14847. June 23, 2016 install.pl Eric I wanted to show the user $CDTST but perl insists on translating that so changed the execution of fix_aips_elcap.sh from $SYSMACIN to $AIPS_ROOT which will translate properly. Moved to 31DEC15 this date. 14848. June 24, 2016 ALGMEM Eric If the last pass through this routine started with a model number greater than one but only included a single range of frequency channels, then the array of scaling numbers to step through the frequencies was not set properly. Moved to 31DEC15 this date. 14849. June 24, 2016 IMSCAL Eric Changed OOCAL.001 RUN file to add OUTNAME for all temporary files, namely the facet images and the output of OOSUB in the OOCAL step. This allows multiple AIPS to run multiple copies of IMSCAL without name conflict. Changed HELP files for OOCAL and IMSCAL. Moved nowhere. 14850. June 27-28, 2016 BPEDT Eric NEW TASK: interactive graphical editor much like EDITA but using bandpass tables to edit UV data. Files include $APGOOP/BPEDT.FOR, $APLOOP/BPEDIT.FOR (large library of routines), $HLPFIL/BPEDT.HLP, and $HLPFIL/HLPEDIBP.HLP (run-time help). Moved nowhere. 14851. June 29, 2016 TABED Eric Changed code to copy all rows of a table to the output table even when the output table is new unless APARM(10) > 0. Then only the selected rows are copied. Before this change, a DELE operation would copy only the deleted rows to a new output table and a UFLG would copy only the previously flagged rows to a new output table. This change allows one to be more cautious and not write over the input file but still keep all of the input file. Moved nowhere. 14852. June 30, 2016 AIPSLetter Eric LET16A in TEX, PS, and PDF put into system in $AIPSPUBL and on the web pages. (TABED, BPEDT not included) 14853. July 1, 2016 VLBAUTIL Amy Just for safety sake, changes all "sour" to "sources". Moved nowhere. 14854. July 2, 2016 BPEDT Eric Changed BPEDIT to (1) improve formats, (2) to find next sample in the area being flagged (area, above, below) and order that one to be flagged, (3) do flags as ordered ignoring sample's flux. The last matters when flagging more than one antenna or time based on the values of one antenna and one time. Moved nowhere. 14855. July 4, 2016 PC Tones Eric Changed PPCV.INC to allow up to 256 tones per IF per polarization. Requires a some COMRPLs and a bunch of COMLNKs. Moved nowhere. 14856. July 4, 2016 POSSM Eric Changed POSSM to plot PC tones as spectra on APARM(8) = 9. Scan averaging is based on the NX table and the tones axis is always labeled in channels. Moved nowhere. 14857. July 5, 2016 PCFLG Eric NEW TASK: Based on SPFLG, PCFLG edits pulse-cal tables writing a new version with blanked out channels. Also a help file, a run-time help file. Fixed run-time help files for SPFLG and FTFLG from things found while editing the one for PCFLG. Also fixed a bug in the PC-plotting of POSSM. Moved nowhere. 14858. July 6, 2016 PCPLT, PCFLG Eric NEW TASK: PCPLT is like BPLOT except for PC tables. Lets you look at pulse-cal spectra as function of antenna, one time slice per plot page or as a function of antenna, one antenna per plot page. PCFLG was found to ignore magic blanks when binning up the input pulse-cal table. Corrected it to put magic blanks for real, imaginary, and flag number in the grid for such samples. Moved nowhere. 14859. July 8, 2016 VLBAUTIL/VLBAFRNG Amy Added some CLRTEMPs to VLBAFRNG after a report of a BLEW CORE when running. Moved nowhere. 14860. July 18, 2016 ACCOR, ACSCL Eric Changed ACSCL to add the SUBARRAY adverb and to allow it to loop over subarray on SUBARRAY=0. Discovered that the code and ACCOR did not handle the last scan correctly if it contained auto-correlation data. That last scan was dropped. Also changed help file to change undefined SVAN to "scan length" and changed the cutuff to 0.1 * min (SOLINT, scan-length) in both code and help. Changed accumulation of integration times in case integration time is not one of the random parameters. Moved nowhere. 14861. July 19, 2016 XG2PL Eric XG2PL needed to set some parameters in order to work in the absence of an XG and/or ZE table. It also had an error in computing the Zeeman models done without Gaussians. Moved nowhere. 14862. July 19, 2016 TVSPC Eric Added APARM(8) option > 0 -> label spectra in channels, else label with the first axis units. Moved nowhere. 14863. July 21, 2016 XGAUS, XG2PL, ZEMAN Eric XGINI and ZEINI: to add keywords to the call sequence to define the frequency reference value, pixel, increment, and type. On read of an old file, the current header provides default values if necessary (thereby handling the translation problem). XGAUS: (a) to add option DOMODEL to plot the individual Gaussian components when there is more than one, (b) call sequence to XGINI including header values when creating a new XG table, (c) testing old versus new coordinate values with warning about using old ones, and (d) using old reference values rather than current header values throughout. This is particularly important since the component locations are in pixels wrt the reference pixel. ZEMAN: (a) to new call sequence for XGINI and ZEINI including providing header values for a new ZE table, (b) testing old reference values against current header with warning, (c) testing ZE reference values against XG reference values to warn if they are not the same, and (d) using old XG refernce pixel for Gaussian models and old ZE header reference values throughout. XG2PL: (a) to new call sequence for XGINI and ZEINI, (b) comparing reference information from XG and ZE tables with a warning if they do not match, (c) using old XG reference pixel for frequency reference, and (d) copying and displaying enough parameter fit values when doing non-Gaussian fits. The task already uses the current header intelligently to plot the X-axis labels. RM2PL: correct units to "Jy/beam" rather than "Jansky". Moved nowhere. 14864. July 21, 2016 PCFIT Eric NEW TASK: Fits delays and phases to multi-pulse cal data and writes an SN table usually with the difference in phase and delay between the start and the current time. Moved nowhere. 14865. July 26, 2016 BPEDT Eric The Explain section of the help file had not been edited for this task rather than EDITA. Moved nowhere. 14866. July 26, 2016 PCEDT Eric NEW TASK: graphical editor for pulse-cal tables. Plots each PC spectrum and allows editing functions sort-of like SNEDT, EDITA, and EDITR. It writes a new flagged table if any flags are generated. Changed BPEDIT adding EDITPC (main control routine), PCGTPC (read PC table into ram), PCFCAP (apply flag command table to PC table values). The routine to find the "nearest" pixel was changed, in the case where the pixels are spread distinct along the X axis, to take the nearest one always ignoring any directive about above or below. Moved nowhere. 14867. July 27, 2016 IMSTAT Eric AU6D retained case sensitivity about Jy/beam (JY/BEAM) and so did not always show total flux depending on the brightness units case. Made it case insensitive. Odd that i missed this one when I went through lot of routines fixing this issue (CASA uses Jy/beam). Moved nowhere. 14868. July 27, 2016 XG2PL Eric Changed to print to OUTTEXT the same X axis info as requested in the plot - frequency, velocity, or whatever is in the header. Moved nowhere. 14869. July 28, 2016 PCFIT Eric I subtracted Cable-Cal from the delays, but forgot to check for magic blanks. Added tests for that, setting the cable-cal to zero. Moved nowhere. 14870. July 28, 2016 WIPER Eric Changed to keep track of up to 25 baselines at each pixel, to have large enough buffers to load really wide TVs, and to position the wiper display correctly and more sensibly on WIPER SET. Moved nowhere. 14871. July 29, 2016 CLCAL Eric Changed: SNAPP To return a logical if any phases, delays, rates are valid and non-zero SN2CL Poor logic allowed the SN table record number to be counted beyond 2 billion and so to warp to negative. At that point it thought it could read more SN table data and died badly. CLCAL Fixed an array declaration to use MAXANT rather than a hard-coded 90. Changed call to SNAPP and set it up to call CALREF only when phases actually need to be re-referenced. This will save alarming messages when processing APCAL SN tables. Moved nowhere. 14872. July 30, 2016 DFTPL Eric Added almost all of the calibration adverbs to this task. The original and perhaps best usage is still for a uv data set having all calibration and flagging applied and interfering sources removed. However, there is no reason not to allow a more complete range of access to the UV data. Flagging is especially important. Extracted TBTIME from TVFLG to use in this task and changed TVFLG, SPFLG, and FTFLG to use the library version of this routine henceforth. Moved nowhere. 14873. August 12, 2016 subarrays Eric In the new format, the maximum number of subarrays is not limited to 99 so changed DRNX.INC Changed MAXSUB to 512 from 50 FITAB Added test on maximum subarray to decide output random parameter form. FITTP As FITAB Moved nowhere. 14874. August 12, 2016 RFLAG Eric A couple of variables in the FLAGIT routine required SAVEing. Moved nowhere. 14875. August 16, 2016 UFLAG Eric NEW INTERACTIVE TASK to display data gridded onto a U-V plane and enable various kinds of editing. The samples are gridded with a one-cell pillbox convolution and the image is divided by the sum of the weights - equivalent to pure uniform weighting. The images maintained in ram include the scalar average amplitude, the vector average amplitude, and their difference. Also maintained at each pixel are the lists of baselines contributing to the pixel and the absolute visibility record numbers contributing to the pixel (up to 50 of each). The image is of positive V only, using the Hermitian property, and the scaling is set by CELLSIZE and IMSIZE to match the pixel sizes in IMAGR. Automatic flagging options are planned but not yet present. Moved nowhere. 14876. August 17, 2016 WIPER Eric Changes suggested by UFLAG: 1. Dropped adverb LTYPE, changed code to use -7. 2. Changed default image size to allow axis labels to fit. 3. Use pixel replication if (sub)image is small enough. 4. Take axis labeling into account when positioning image. 5. Interactive wiper size settings take pixel replication into account. 6. CURVALUE-like display durring editing now at top of screen instead of overlapping the data explicitly. 7. Changed flag writing to try to combine adjacent spectral channels when possible. 8. Interactive flagging reports the number of pixels and vis samples flagged. (FAT only when done, rest on each flag) 9. Added PRTLEV adverb to allow users a choice about the detailed flagging report. Moved nowhere. 14877. August 17, 2016 DFTPL Eric Added forgotten setting of BCHAN, ECHAN, BIF, and EIF before calling the UVGET routine. The CHNDAT call had numerous issues including LUN, CNO, and wrong header causing it to try to find the antique CH table. Moved nowhere. 14878. August 17, 2016 CookBook Eric/Amy Changed LaCook9 to have DOCAL true for BPASS - one does not want the delay to also be part of the bandpass solution. Moved nowhere. 14879. August 20, 2016 FGTAB Eric NEW TASK: writes a text file with BFREQ and EFREQ values for those flags that have no time range, source, subarray, or frequency ID specified. In addition, no antenna shouold be specified, although one may be if OPCODE is 'ANTE'. In that case the ANTENNAS keyword is also written. This allows flags developed for common RFI to be translated into text files which may then be read by UVFLG for data sets with different frequency structures. Moved nowhere. 14880. August 21, 2016 FGCNT Eric NEW TASK: compares two flag tables by counting the visibility samples that get through them separated by source, polarization, and IF. All standard calibration adverbs are available for use. There appears to be a real interest in knowing how many samples are flagged by various interactive or automated flagging routines. This task performs a reliable count - simply counting in the FG table cannot account for significant time ranges in some flags and all of the other peculiar situations which AIPS' general flag tables make possible. Moved nowhere. 14881. August 21, 2016 FUDGE Eric Added adverbs ANTENNAS and BASELINE which had for some reason been omitted. FGCNT was based on FUDGE and needed them. Moved nowhere. 14882. August 21, 2016 PBCOR Eric The history file was not telling truth for EVLA images (or for any other for that matter). The real parameters should be displayed and, if doing the new EVLA interpolation, both sets should appear. So made an include DPBCALC.INC and changed PBCALC to use it and fill in some extra parameters. Changed PBCOR history to print the actual parameters used and, if EVLA, to say so and print the second set of parameters used. Moved nowhere. 14883. August 22, 2016 FGCNT Eric Added options to list iver antennas in a couple of ways. Cleaned up issues when only one source is selected from a multi-source file and when using a single-source data set. Moved nowhere. 14884. August 22, 2016 UVHIM Eric Added option to do both Hermitian points when doing a u-v grid. Moved nowhere. 14885. August 23, 2016 VLAMP Eric/Amy Changed to count valid correlators through the full CL/SY table reading rather than depending on the SY table's last time. Changed to average the gain over the IFs and to do gains for both polarizations. At present the values are the same in all IFs and polarizations but there is no reason to assume that that will always be the case. Changed to write only the one or two averages, rather than a value for each IF which ANTAB will not swallow. Moved nowhere. 14886. August 23, 2016 HINOTE Eric The adverb PRTASK should be 8 characters like TASK, not 5. Changed POPSDAT To make PRTASK 8 AU1 To handle old and new lengths silently. AU7 To handle old and new lengths silently. AU7B To handle old and new lengths silently. Changed HINOTE to use PRTASK instead of "HINOTE" when DOHIST <= 0 and doing either the COMMENT form or the interactive input form of the verb. PRTASK HELP: Changed the length, mentioned the change of length and updated the list of verbs using this adverb. HINOTE HELP: Added PRTASK and changed the descriptions of the pre-pending function. Moved nowhere. 14887. August 23, 2016 IMSCAL, OOCAL Eric Changed OOCAL.001 RUN file to add two procedures to write history records to the input file. OOCAL and IMSCAL invoke ordinary tasks which write history records to their output files, but those files end up deleted most of the time. The SN table is all that is saved and that goes on the input file eventually. IMSCAL will write a full history including OOCAL and OOCAL when not used by IMSCAL will write the OOCAL part of its history. All this depends on the changes made to HINOTE and the length of PRTASK made today. Moved nowhere. 14888. August 24, 2016 MULIF Eric/Walter Alef There was a wrong argument to UVDFND hard coded as 4, when only 1 or 2 are allowed. Changed to 2, thanks Walter. Moved nowhere. 14889. August 25, 2016 PRTAB Eric The printer line count routine did not have the width of the printer set and so counted lines as if one column at a time was printed. Fixed this. Moved nowhere. 14890. August 25, 2016 INDXR Eric This task had a local limit of 20 for maximum subarray. Changed to 512 which is expensive. Moved nowhere. 14891. August 25, 2016 UVLOD, FITLD Eric The conversion of BASELINE to SUBARRAY was inaccurate and failed in some cases. Adding roundup of 0.0001 was not enough, changed to 0.05. Moved to 31DEC15 this date. 14892. August 26, 2016 Subarrays Eric The subroutine GETNAN is widely used and customarily had an output array dimension of 51 for number of subarrays and number of antennas in subarray(i-1). Changed this to 513. Gridding routines in TVFLG and friends double this and use the upper half for cumulative number of baselines. Changed: APLNOT: GETNAN, GNSMO, GRIDTB, AN10RS APGNOT: ACFIT, ACLIP, BASRM, CLIP, CORER, FGDIF, FLGIT GETJY, INDXH, MULTI, OFLAG, PHSRF, QUACK, REFLG, SPCAL, TRUEP, TYAPL, TYSMO, UVAVG, UVCOP, UVDI1 UVFLG, UVIMG, UVLIN, UVMLN, UVMTH, UVRFI, VBGLU QPGNOT: BLCAL, BLCHN, DEFLG, LPCAL, PCAL , RLCAL YPGNOT: FTFLG, LOCIT, SPFLG, TVFLG, UFLAG, UVHGM, WIPER The following tasks called GETNAN but did nothing with the output, so dropped the call: $YPGNOT: CAPLT, CLPLT, FRPLT, POSSM Moved nowhere. 14893. August 26, 2016 SPLIT Eric When averaging IFs as well as channels, SPLIT got the wrong output frequency. UVGET had been called so the header frequency was adjusted for BIF; the computation did not take this into account. Use OLDFRQ instead. Moved to 31DEC15 this date. 14894. August 26, 2016 UFLAG Eric 1. Changed it to make the list of visibilities a linked list so that every pixel can account for its visibility records. 2. Added ISCALIB adverb to support automatic flagging more completely when looking at RR, LL, XX, YY, or I polarization. 3. Fixed it to handle switching polarizations when no new FG table is created. 4. Added USER and AUTO FLAG VIS functions. Looks at some of the grid pixels, flags visibilities outside specified ranges in the FC table, updating the grid. 5. Computes and displays the median, mean, and rms of the current data type. Uses them in the AUTO FLAG VIS. 6. Corrected handling of pixel-replicated screen values. 7. Display phase as well as amplitude when examining vis. 8. Fix history when looped throuugh Stokes. Moved nowhere. 14895. August 29, 2016 UVHGM Eric changed: 1. Allow 32768 total boxes and fix Gaussian fitting to allow that many. It was limited to 1024 but did not check this limit. 2. Changed plotting of histogram with Gaussian fit to plot just the outline (DOALL > 1) or the full vertical boxes. 3. Changed setting of default plot range for real, imaginary, and amplitude to find the actual min and max of each. Moved nowhere. 14896. August 30, 2016 UVLOD, FITLD Eric Subroutine RUVTAB did not handle compressed data properly when DOKEEP was false. Added a check that went over the buffer properly. Moved to 31DEC15 31 August. 14897. September 3, 2016 UVIMG Eric Added adverb DPARM(8) to signal that the convolution adverbs are actually wanted. XTYPE and YTYPE are usually set to 5 for IMAGR but do stuff here that usualy is not wanted. Added a variety of comments about this and about DPARM(5) and (6) which almost always have to be set. Changed Fortran to implement DPARM(8) and to keep the convolver from going off the edges of the image. Moved nowhere. 14898. September 6, 2016 AIPS Memos Eric Added info about related AIPS Memos to help files for XGAUS, RMFIT, ZEMAN, XG2PL, RM2PL, TVSAD, and TVSPC. Moved nowhere. 14899. September 7, 2016 UFLAG Eric Changed: Help file: correct inputs description of DOCENTER, mention new memo, add phase image, change VIEW ONLY and auto flag descriptions in Explain Realtime help: Change SET PIXRANGE (now type dependent), add phase image type, change AUTO description, change VIEW ONLY and FLAG ALL (now type dependent) Fortran: 1. VIEW ONLY and FLAG ALLs now done in image type dependent ways 2. SET PIXRANGE now image type dependent 3. Phase is now image type 3, difference is 4 4. Change plot positioning, reposition if current window corners have changed, corrected Y limit 5. Allow VIEW ONLY range to go outside actual range (which happens for some reason) 6. Allow negatives in image (for phase) 7. Improve curvalue-like display and other displays for units 8. Auto-editing progress messages now every 2 minutes or so 9. Correct some catalog status issues Moved nowhere. 14900. September 12, 2016 UFLAG, Memo 121 Eric Changed UFLAG to have APARM control the multipliers times the rms in the AUTO FLAG VIS function. Wrote AIPS Memo 121 to describe this task. Moved nowhere. 14901. September 23, 2016 DOVLAMP Amy Made a couple of improvments to make DOVLAMP more bulletproof: 1) Now if INCLASS='TYAPL' the proceedure does not try to load the data even if ASDMFIL is set to something. I.e., if INCLASS='TYAPL' then always assume that the data is already loaded (and TYAPL has been run). 2) If primary flux calibrator that is present in the data should not be used, CALSOUR(1) can be set to 'NOCAL' and the SU table will not be searched for a flux calibrator. Moved nowhere. 14902. September 26, 2016 PBCALC Eric Updated all help files with PBPARM primary beam correction options. This included BOXES, FACES, IMFIT, JMFIT, LTESS, MOMNT, PATGN, SETFC, STESS, FLATN, PBCOR, SAD, SPCOR, SPIXR, XMOM, UTESS, VTESS, and TVSAD. Moved nowhere. 14903. September 27, 2016 VLOG Amy Took out the half channel shift in the calculation of the freq of the edge of IF. Also added a default of 0.1 Hz for the FQTOL (frequency comparison tolerance). Closed the WRIT status of input UV data file. Moved nowhere. 14904. September 27, 2016 CONVL Eric Added OPCDE='GMOD' to convolve with the specified Gaussian ignoring all Clean beam information. Moved nowhere. 14905. September 27, 2016 CALIB Eric Added more tricky normalization options which average over subarray but separate IF or IF and polarization. Moved nowhere. 14906. September 28. 2016 PBCALC Eric Added a message to say what array parameters are being used by PBCALC. Added web address for EVLA Memos to all the PBCALC task help files. Moved nowhere. 14907. September 29, 2016 PCINI Eric The PC table initialization routine regarded a mismatch in revision number as an error. Changed it to a warning. Moved nowhere. 14908. September 29, 2016 BPEDT Eric Fixed 3 bugs: 1. The flags written by BPEDT were too "exact" in their times and needed to be rounded outwards (by 0.095 sec) to get all the data contributing to the bandpass solution. 2. The automatic choice of a fully positive plot range for phase failed to change the phase values to all positive. It still plotted correctly, but the flagging failed. Two loop counters were not properly incremented. 3. A parameter that kept track of whether an IF was flagged for EDITA, EDITR, et al. was left in the code. This is needed if all IFs are not in RAM at the same time. BPEDT really does not support this, so the parameter is not needed. It was being filled in badly, writing over other data making a mess of things. Moved nowhere. 14909. October 4, 2016 PLGET, EXTLIST Eric Went over all plot tasks to make sure that PLGET and EXTLIST are up to date. Added RM2PL, XG2PL, ALVAR, PCPLT, and PLROW to AU2A (PLGET) and AU8A (EXTLIST). Also udated UVPLT, VPLOT, CLPLT, CAPLT, DFTPL, KNTR. RFLAG, EVAUV. DELZN, and SOUSP in these routines. Modified XG2PL, RM2PL, DFTPL, PLROW, PCPLT, and ALVAR to fill in extra defaults into the adverbs to support EXTLIST. Changed PLGET and EXTLIST help files. Moved nowhere. 14910. October 6, 2010 SLICE Eric The operation to drill down through a cube with a single pixel in X and Y and the slice running along Z had a problem if the X and/or Y pixel was not exactly an integer. Bad slices resulted. Corrected the code to interpolate in X and Y correctly during this operation. Moved to 31DEC15 this date. 14911. October 7, 2016 TVSPC Eric Changed to add SET CHANNELS and RESET CHANS operations to set the channel range interactively and reset that range. The plots and fits then use only the requested channel range. Any slice written out however has the full channel range. Moved nowhere. 14912. October 6, 2016 CookBook Eric Corrected a bad underline which came across as a subscript in a Memo reference. Changed LaCook9.tex, COOK9.PS and COOKBOOK.PDF. Moved everywhere (AIPSPUBL goes everywhere). 14913. October 9, 2016 New Mac Eric The new Mac to compute binaries from a 10.11 OS (El Cap) but designed to run on 10.7 and later requires some changes to the compiler options and to the libraries used. Changed $SYSMACIN versions of FDEFAULT.SH. LDOPTS.SH, CCOPTS.SH, and fix_aips_elcap.sh and MACINT options in the Makefile contained in $YSERV/XAS.SHR. Moved nowhere. 14914. October 12, 2016 DFTPL Eric Fix up Stokes handling some to allow linear polarizations. Moved nowhere. 14915. October 12, 2016 TVSPC Eric The new SET CHANNELS option could set BCHAN > ECHAN which would then cause the task to die. Fixed it to defend itself so that BCHAN is always < ECHAN. Moved nowhere. 14916. October 12, 2016 BPASS Eric New adverb MINANTEN for minimum number of antennas - in POPSDAT, NEWPARMS, and a help file. Added this adverb to BPASS help and code. MINANTEN was 2 inside the code previously. Moved nowhere. 14917. October 14, 2016 MNJ Eric Discovered that the MNJ was not working properly - most obvious symptom was POPSGN was not being run for new POPSDAT.HLP file. Changed: AIPSUPD Binary versions need to rsync the transaction files to 2 places. The cvs version has the appropriate copies. UPDCONTROL Dropped use of UPDOBSLT program which is not needed since the chkout system already removes any "removed" files from PUTBCK, COMRPL, COMLNK lists. UPDOBSLT was not compiling properly on the new Mac. Changed the sort command grammar a bit to produce a more correctly sorted PUTBCK.UPD file. MAKE.MNJ Dropped the compilation of UPDOBSLT and fixed reference to 31DEC03. MAKE.BMNJ Changed reference to 31DEC03 (!) Unfotunately, these two issues affect the ability of MNJ sites to update file automatically so these corrections will have to be done manually. Moved nowhere. 14918. October 18, 2016 XAS Eric Mac OS 10.10 (or earlier) made a change to the file descriptor lengths that caused them to be too large for the basic limited size in the standard socket routine "select". This is a bug which they admit probably brought out by the switch to a 64-bit version. Added the compile option -D_DARWIN_UNLIMIT_SELECT to the Makefile for MACINT. Also fixed some setting of the TVLIMG parameter which had errors (clearly of little or no significance). Moved nowhere. 14919. October 21, 2016 TABED Eric The relatively new controls to always copy had a bug in which the records before the first requested record were not copied properly but instead a zero-filled record was appended at the end (if the input and output tables are the same). Moved nowhere (not in NEW). 14920. October 21, 2016 FTFLG, SPFLG Eric There was an addressing issue when flagging a time or time range over a range of IFs. It left a few channels on some end unflagged and potentially could have flagged channels other than those desired. This only arose when a sub-image of the whole was loaded to the TV. Moved nowhere. 14921. October 21, 2016 PBCALC Eric The Pband EVLA parameters were not correct. The values for the fourth term were tabulated for the third term and the existence of a fourthe term was ignored. Fixed the third term tabulation and added code for the fourth term. No ther frequency uses 4 terms. Moved nowhere. 14922. October 24, 2016 FTFLG, SPFLG Eric When flagging IFs N1 - N2, the interactive flag options want to keep the cursor in that region. If the image is zoomed in a way that that region is not visible, confusing things happen. Changed code to detect this and exit gracefully. Moved nowhere. 14923. October 24, 2016 VLBARUN Amy Corrected the tests on the results from BSCAN. Moved nowhere. 14924. October 25, 2016 MULIF Eric Fixed the task to copy autocorrelation data (probably was just an oversight) and added (non-OOP style) code to create and fill an index table. Moved nowhere. 14925. October 26, 2016 ZMKTMP Eric $APLUNIX/ZMKTMP.C did not work on the new 64-bit Macs. The format %010d into a 10-character string was no longer acceptable. Changed to use %9d which is more than enough. This affects all attempts to make a temporary file name - primarily in Cleans (IMAGR for boxes) and verbs FILEBOX, DFILEBOX, and the GRIPE verbs. The calling routine ZFULLN is called a million places but I think we need only link imaging tasks (IMAGR, WFCLN, SCMAP, SCIMG) and AIPS*. Moved nowhere. 14926. October 27, 2016 Slice fitting Eric Changed POPSDAT.HLP Add verbs TVCOMPS, TVACOMPS to plot individual components NEWPARMS.001 Add verbs TVCOMPS, TVACOMPS to plot individual components AIPS Change verb number range for AU6E AIPSB Change verb number range for AU6E AIPSC Change verb number range for AU6E AU6E change branch number values, add TVCOMPS, TVACOMPS TVSLIN change branch numbers SLFIT.HLP add ORDER adverb to specify baseline fit order SLFIT add adverb, add solving for baseline up to quadratic MV2C06SL Document added 6 words AU8A Change EXTLIST for slice models to know about baselines TVSLPL Add component plotting and baseline fit TKGMPL Add baseline fit TKRSPL Add baseline fit SLPRT Add baseline fit TVSPC.HLP Add APARM(9) = ORDER+1 TVSPC Add APARM(9) to SET APARM, add baseline fitting to FIT SPECTRA and SAVE SPECTRA. HLPTVSPC.HLP Correct SET APARM and include order setting. SL2PL.HLP DOMODEL up to 3 for comps, sum model SL2PL Change plot to do comps and or sum of model including baseline. Moved nowhere. 14927. October 27, 2016 VLBARUN Amy VLBARUN often assumed that the input data file's INSEQ was 1, fixed. Also added a check on number of stokes before setting stokes='half' for UVPLT. Moved nowhere. 14928. October 28, 2016 MNJ Eric The revised script still updated some files in special areas (POPSDAT and MNJ scripts mainly) **before** they were copied from the master when using rsync. Changed AIPSUPD to do the update at the right 2 places. Moved nowhere. 14929. October 28, 2016 slices Eric Fixed one more reference to verb number in AU6E (TVSET and TV1SET) did not work. Added help files for the new verbs TVCOMPS and TVACOMPS. Moved nowhere. 14930. October 28, 2016 AIPS Memo 120 Eric Upgraded AIPS Memo 120 to describe SET CHANNELS, RESET CHANS, changes to APARM and SET APARM. Required new figures in most cases for the new menu and pagination did not turns out so well. Moved everywhere ($AIPSPUBL goes everywhere). 14931. November 8, 2016 RMSD Eric Added OPCODEs MEDI for median and MAD for median absolute deviation from the median. The latter is thought to be a more stable estimator of noise than regular rms. RMSD uses a robust rms so that may not be important. Moved nowhere. 14932. November 8, 2016 general lists Eric Updated: 1. HLPIT.LIS the TAB completion file 2. ABOUT lists: ADVERB, ANALYSIS, CALIBRAT, EDITING, IMAGING, INTERACT, MODELING, ONED, OOP, PLOT, SPECTRAL, TABLE, TASK, TV-APPL, TV, UV, VERB, VLBI Moved nowhere. 14933. November 11, 2016 CookBook Eric Updated chapters dropping references to 31DEC11 and older releases 4. BPEDT 5. UFLAG, FGCNT 6. SLPRT, TVCOMPS, UFLAG, BPEDT, PCFLG, PCEDT 7. TVSPC, SLPRT, PBCOR, PATGN, TVACOMPS, slice baseline, SL2PL 8. minor rewording 9. BPEDT, new PC tasks 13. Update 18 ABOUT lists E. BPEDT, UFLAG, FGCNT 2, 3, 10, 11, 12, D, F also changed for old references 0, I, J brought up to match above All COOK*.PS updated, COOKBOOK.PDF, and the html web version. Moved nowhere. 14934. November 16, 2016 PCLOD Eric Added capability to read the new DiFX format text files. Other bugs found: BPEDIT The setting of min/max went over all times - appropriate for QEDIUTIL where all times are in the plot, but not for BPEDT and PCEDT which have only 1 time per plot. PCFLG Changed the setting of min/max - extending it some. Phases of about -180 would stay very black. Also made same change to FTFLG, SPFLG, TVFLG. Moved nowhere. 14935. November 21, 2016 RMSD Eric Scale the MAD output to be equivalent to an rms (* 1.4826). Moved nowhere. 14936. November 21, 2016 PC tables Eric Pulled the PC fitting code from PCFIT, making a subroutine set to do the fitting, returning the fit delays, phases, errors, weights, and residual PC spectra. The new routine is called $APLNOT/PCFITR. Changed also PCFIT to use PCFITR and to write a new residual PC table as well as an SN table. PCFIT HELP: add info about new PC table PCPLT to use DOCALIB to decide whether to plot residual pulse-cal spectra or input pulse-cal spectra. PCPLT HELP: to change "bandpass" to "pulse-cal spectrum" throughout and to add DOCALIB BPEDIT to store residual phase and PC spectra frequencies in the mammoth data array for PC tables, to display and edit with residual phase if requested, and to re-compute all residual phases when requested. PCEDT HELP: rewording of what it does and fix up the explain file's menus. HLPEDIPC HELP: new on-line help reworded some form HLPEDIBP with new options added PCLOD changed to be forgiving about mismatches between the UV data set and the PC table. A UV data set must be used but it does not have to match in times or frequencies. It does help if the antennas match or completely do not match. The task wants to convert times with an NX table to source and FQID numbers and an AN table to convert antenna names to numbers. PCLOD to explain why a data set must exist but that the task is very forgiving. Minor typos fixed in BPEDT.HLP and HLPEDIBP.HLP as well. Moved nowhere. 14937. November 23, 2016 PC again Eric Changed BPEDIT to get the PC application routine corrected to the new addressing required for the residual display. It died when trying to write out the flagged PC table. PCFLG Changed to do residual computation and display as well as displaying alteratively the input values. Also PCFLG and HLPPCFLG help files for new options and the availablility of residual as well as input data. Moved nowhere. 14938. November 28, 2016 Roll-over prep Eric 1. Changed all HLP*.HLP files to say run-time help and fixed the wording of several. 2. TAB completion file (for HLPEDIPC) 3. Significant number of ABOUT files for the changes to HLP*.HLP 1-liners. 4. Chapter 13 of the CookBook for the ABOUT lists Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC16 is NEW *********************************************************************** *********************************************************************** 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. 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 16, 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 27 Dec. 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. 14958. January 12, 2016 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. 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. 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. 14985. February 28, 2017 DFTPL Eric Corrected addressing of visibility data - pointers need to be set after UVGET ('INIT'..) not before. Led to IFs > BIF being ignored. Moved pointer fix to 31DEC16 this date. 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. 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. Changed 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. SPIXR Fortran - changed output image frequency to 1 GHz since that is what it is. 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 above to 31DEC16 this date. 15029. June 6, 2017 PCAL Eric The evaluation of the flux at 1 GHz was wrong for low frequencies leading to bad solutions. Made a similar fix to 31DEC17 this date. 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. 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. 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. 15043. 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.