;----------------------------------------------------------------------- ; Copyright (C) 2017-2018 ; 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 31DEC18 is TST *********************************************************************** *********************************************************************** 15103. December 11, 2017 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 31DEC18 instead of 31DEC17 - also fix dates in install.pl. Moved nowhere. 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. Moved to 31DEC17 this date. 15106. December 18, 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 31DEC17 December 15. 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. 15114. January 3, 2018 RLDIF Eric Changed the matrix print routine to make a nicer output on DOPRINT < -2.5. Moved nowhere. 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. 15116. January 12, 2018 DO3COLOR Eric Changed the spelling of DO3COL to DO3COLOR in POPSDAT.HLP. Changed help files for: ANBPL Add to help file description of DO3COLOR. BPLOT Add a bit to help file description of DO3COLOR. ELFIT Add a little to DO3COLOR help file description. GCPLT Add detail about color usage. PCPLT Add a bit to help file description of DO3COLOR. SNIFS Add a little to the DO3COLOR description SNPLT Add a lot to help file description of DO3COLOR. TARPL Add a bit to help file description of DO3COLOR. UVPLT Add description of DO3COLOR to Explain section VPLOT Change description of DO3COLOR to be more detailed. Upgrade STOKES amd POLPLOT for linears and added options. FRPLT Improve the POLPLOT description to include linears. POSSM Improve the POLPLOT description to include linears. Changed Fortran: POLCHK Allow linear polarization strings (affects FRPLT, POSSM, VPLOT) ANBPL Turn off 3-color when only 1 thing plotted, go to color 0.97 BPLOT Colors 0 to 0.97 (1 too dark) ELFIT change max color from 0.95 to 0.97. FRPLT Corrected the Stokes labeling on POLPLOT PCPLT Colors 0 to 0.97 (1 too dark) POSSM Corrected the Stokes labeling on POLPLOT SNIFS Prevented an unlikely failure in the coloring logic, colors 0 - 0.97. SNPLT Colors go to 0.97 (pure blue is too dark). Drop usage of plot types for plots with just 2 polarizations and 1 IF. GRCHAN > 0 made that not show color. UVPLT Colors go to 0.97 (pure blue is too dark) VPLOT Allow more STOKES values including plotting parallel and more than 2. Fix POLPLOT label and stop color plotting when there is only one thing plotted. reversed the order of colors plotted; now red to blue. Moved nowhere. 15117. January 12, 2018 PBEAM Eric Changed APARM to VPARM and added VPARM(11) to indicate that the data are power rather than voltage. This case arises when the data are from two moving antennas (and hence real visibilities). Moved nowhere. 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. 15120. January 17, 2018 Compiling on Macs Eric Changed $SYSMACIN versions of FDEFAULT.SH, LDOPTS.SH to add *gfortran* as a possible compiler and changed CCOPTS.SH to allow gcc to have characteers both before and after (as with gfortran above). Moved nowhere. 15121. January 17, 2018 SOUSP Eric Before plotting, put the fit rms in as the noise for those IFs for which the input adverb NOISE was zero. Moved nowhere. 15122. January 19, 2018 SETJY Eric Overhauled the EXPLAIN section of the help file. It had not been updated when the coefficients chosen by APARM(2) were changed. I hope that is more correct and understandable now - the coefficients are given in great detail for the more recent flux calibrations. Fixed precursor comments in the Fortran as well. Moved nowhere. 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. 15125. January 22, 2018 RLDIF Eric The error bars in the SPECTRAL solution were too large. One wants to give something like the expected error - i.e. divide by the sqrt of the number of independent samples. As a proxy for this unknown, I changed it to divide by the sqrt of the average number of baselines contributing at each time to the computation. The results are reasonable anyway. Moved nowhere. 15126. January 22, 2018 XYDIF Eric For linear polarization the X-Y phase difference is found from source(s) that have U flux but zero V flux. Applying all calibration plus DOPOL but not applying parallactic angle in the conversion to Stokes, one gets apparent U and V. The ATAN(-Y/U) is then the phase difference. XYDIF is much like RLDIF except for how it determines the phase difference. Moved nowhere. 15127. February 7, 2018 MORIF Eric Changed error handling when trying to update the NO_BAND keyword in CT tables. The CTINI call can have an error that was not tested. Changed handling of the output frequency reference pixel. It was set to 1 for simplicity, but the center of the band is preferable. Moved nowhere. 15128. February 9, 2018 VLATECR Amy Changed VLATECR adding APARM which allows dispersive delay correction to be added. No change neccesary in VLAPROCS, new inputs will just be used. Moved nowhere. 15129. February 12, 2018 ICHANSEL Eric Added some examples to help files for ICHANSEL and to the ICHANSEL part in help files for BPASS, PCAL, REWAY, RFLAG. Moved nowhere. 15130. February 14, 2018 EDITA Eric Changed QEDIUTIL to display the solution weights in SN and CL tables along with amplitude, phase, delay, and rate. Changed the display of flagged data in 3 colors to white - the red graphics channel was too much like a real color assigned to some IF or polarization. Flagging in 3 colors required some change in logic as to the order of erasure and drawing. The weights were already in the data, but were not able to be displayed previously. This caused changes in the help files for EDITA, HLPEDICL, and HLPEDISN. Moved nowhere. 15131. February 16, 2018 SNPLT Eric Added OPTYPEs 'RPHS' and 'RDLY' and REFANT to plot the phase and delay re-referenced to REFANT. Moved nowhere. 15132. February 16, 2018 QEDIUTIL Eric Changed EDITA and other edit-class tasks to (1) correct some too small Formats and (2) to compute the mean and rms of the values in the edit display and then to test them in FLAG ABOVE and FLAG BELOW. If the former starts less than avg+rms or the latter starts more than avg-rms, then the user is asked if he wants to continue. Moved nowhere. 15133. February 20, 2018 XYDIF Eric Added DOINVERS to allow debugging by adding 180 to the solutions. Moved nowhere. 15134. 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. Moved to 31DEC17 this date. 15135. February 22, 2018 CookBook Eric Changes: 4 Rewrote the chapter to describe the modern EVLA calibration habits based on old appendix E, upgraded polarization in particular 9 Extracted all references to MK III and IV. O New appendix for older systems - contains much of the calibration of the old VLA plus sections for Mk III and IV. E Deleted the EVLA appendix - some was obsolete, the rest moved to 4 1 Corrected obsolete wording, changed appendix and chapter descriptoins 5 Updated references 8 Updated references A Updated references B Updated references L Updated reference, added item on polarization primarily referencing the section in Chapter 4 0 Update table of contents, figures, recipes I Update index J Update index Moved nowhere. 15136. February 23, 2018 FREESPAC Eric Changed ZFREE to omit the display of the TIMEDEST limits. They are not relevant since TIMEDEST was eliminated. This leaves 5 extra characters for the data directory names, which are frequently needed. Moved nowhere. 15137. February 23, 2018 IMFIT, JMFIT Eric Added adverb NDIG to these two tasks. It will allow you to print the map with fewer digits than the default (which is the most between 3 and 6 that will fits on the page/screen). Moved nowhere. 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. 15139. March 1, 2018 EDITA Eric Changed QEDIUTIL to correct the multi-band delay option. Both menus had a line they should not have had and the 3-color option had not been turned off. Moved nowhere. 15140. March 1, 2018 help file clarity Eric Changed help files to improve clarity: RMFIT Add missing close parentheses, enhanced wording on DOSPIX to make it clear that it can be non-zero for some components and zero for others. SNFLG Rearranged wording to make separate OPCODEs more visible. CLCOR enhance opening paragraphs, emphasize CL version handling. FRING Clarify APARM(5)=2 (MB delay) a bit EDITA Clarify DODELAY=2 (MB delay) quite a bit. Moved nowhere. 15141. March 2, 2018 FRING Eric Changed to avoid the flag table sort on the first UVGET call when creating an output file. This may let any failures happen faster. Moved nowhere. 15142. March 2, 2018 linear polarization Eric The warning about linear polarization not being calibrated should only come when producing Stokes Q, U, and/or V and not applying DOPOL. Changed DGHEAD to be much more selective on when to issue this warning. MOved nowhere. 15143. March 5, 2018 TABINI Eric Added an error message telling file type, disk, cno, version for the error condition. This will let people track down issues better. Moved nowhere, should link everything. 15144. March 6, 2018 EDITA Eric Changed: (1) The minimum range down another factor of 10 - 1/10000 of the basic unit. (2) Changed the average and rms computation to use the median and the scaled MAD (median absolute deviation) (3) Fixed display of questionable flag below and flag above to use the correct units. Moved nowhere. 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 2 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. Moved to 31DEC17 on 2 April. 15147. March 9, 2018 SN table weights Eric Since SN table weights may now be used to edit UV data in EDITA, added comments to the help files of CALIB and FRING to describe their special values. A weight of SNRmin+1 implies a suspect solution in CALIB but implies the (good) reference antenna in FRING. Moved nowhere. 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 nowhere. 15149. March 12, 2018 DTSIM Eric Fix the AN file creation so that the files do not look inadequate. Moved nowhere. 15150. March 14, 2018 DTSIM Eric Changed to omit fully flagged data records and report the count. Corrected the noise routine which did not SAVE an important variable. Corrected the SU table to make epoch 2000 when not set. Moved nowhere. 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. Moved to 31DEC17 this date. 15152. March 23, 2018 POSSM Eric The CP plot routine did not test for magic blank in I when plotting Q and U and did not test for 0 in I when dividing by I. Both cases arise and, in the former, Q and U were nuts. Moved nowhere. 15153. March 28, 2018 IMAGR Eric Changed QCLEAN to (1) Not re-image all facets on REMAKE IMAGES if they are all already current. (2) Offer an IMSTAT option to do robust and normal statistics on the image currently displayed. All pixels are used, not just the current display window. Also changed IMAGR and HLPCLEAN help files. Moved nowhere. 15154. March 28, 2018 REWAY Eric Added more tests to make sure that data have been found before continuing with the various operations. Moved nowhere. 15155. March 29, 2018 EDITA Eric Wrote new general purpose routine ANTOUT to find antenna numbers called OUT or with 0,0,0 coordinates in the antenna file. Added code in EDITA to use this to set the ANTENNAS adverb when the user has not done so. BDF2AIPS makes a CL table with all 28 antennas no matter which ones are valid. Moved nowhere. 15156. March 29, 2018 SNFLG Eric Added flagging on delay/rate rather than amp/phase as an option and also an option to flag failed solutions. Moved nowhere. 15157. March 29, 2018 FGSPW Eric Added option SOLINT to control the averaging interval and an option to declare the cutoff levels to be in Jy/flux (and so are multiplied by the SU table flux as function of source and IF before the data comparison). Moved nowhere. 15158. March 29, 2018 FRING Eric FRING set the multi-band delay to magic blank when doing the dispersion option (APARM(10)>0). It failed to give it a more meaningful value even when one was found. Improved the wording a bit in the FRING help file too. Moved nowhere. Moved nowhere: needs patch? 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. Moved to 31DEC17 this date. 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.. 15161. April 4, 2018 Misc Eric DBCON mention DBAPP in help file FRING count frequency increment errors in CHKFRQ and tone down warning messages. FRNSEL will find the really egregious issues. Moved nowhere. 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 data. 15164. April 11, 2018 CONVL Eric On OPCODE 'GAUS' the task will now find the largest clean beam in the CG table (or the header) to use as a default for BMAJ, BMIN, BPA. Leaving them at zero is still an error for some of the OPCODEs. Moved nowhere. 15165. April 13, 2018 dispersion Eric The CL and SN tables have 2 columns for dispersion for the two polarizations and both FRING and KRING solve for dispersion separately in the two polarizations. However, the calibration routines applied only the first one to all polarizations. Changed DSEL.INC Made the dispersion parameters an extra dimension of 2. GAININ Find the second polarization dispersion rather than the unused rate of change of dispersion CSLGET Pick up and use 2 polarizations worth of dispersion CGASET Interpolate 2 polarizations of DDTAB into DDELAY DATCAL Do 4 different dispersion corrections for the 4 polarizations. Moved nowhere. 15166. April 13, 2018 SNEDT, EDITA Eric Changed QEDIUTIL and the two tasks to allow editing solely on dispersion, much like multi-band delay editing. Moved nowhere. 15165. April