;----------------------------------------------------------------------- ; Copyright (C) 2002-2004 ; 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 31DEC03 is TST *********************************************************************** *********************************************************************** 11191. December 16, 2002 several Eric Changed 31DEC02 to 31DEC03 in install.pl, update,pl, INSTEP1, SETPAR, GETRLS, AIPSPATH.*SH, UPDCONTROL. Moved nowhere. 11192. December 16, 2002 install.pl Eric Corrected the handling of the MNJ so that the date will be set by default to the tar ball date - 1 not December of last year. Moved change only to 31DEC02 this date. 11193. December 16, 2002 plots Eric CCNTR and CNTR had an excess call to GLTYPE ahead of GINITL which confuses LWPLA. Moved to 31DEC02 this date. 11194. December 17, 2002 LWPLA Eric Added code to prevent writing commands ahead of the first correct line in the PostScript output file (as CNTR caused until yesterday). Issues warnings, does not write bad lines, and continues. Moved to 31DEC02 this date. 11195. December 17, 2002 LOCIT, BASFIT Eric Changed BASFIT.001 to offer DOCOLOR option. Added info to BASFIT help file. Fixed bugs in LOCIT centering phases when some are blanked and printing phases when some are blanked. Moved to 31DEC02 this date. 11196. December 17, 2002 SUL procs Eric Changed FDEFAULT, CCOPTS, LDOPTS to allow g77 as well as SUNWspro compilers in $SYSSUL. WARNING: the g77 are untested. Moved to 31DEC02 this date. 11197. December 23, 2002 Plotting Eric LWPLA The defense against premature plot commands did not take into account the possibility of appending to a pre-existing file. The defense then stopped any appending. Added a second place for FIRST=.FALSE. KNTR Needed to manipulate the choice of LOCNUM for STARPL. Otherwise LTYPEs other than 3 could not plot stars. Moved to 31DEC02 this date. 11198. December 23, 2002 VLBAPIPE.001 Lorant Had the logic for DOUVCOMP switched, corrected in 31DEC02+03 Moved to 31DEC02 this date. 11198. December 23, 2002 TBIN Eric The program did not handle a missing HI file properly. Added code to create one if there was none and to avoid catastrophic errors in there is an HI problem. Moved to 31DEC02 this date. 11199. December 23, 2002 POSSM Eric Had the ability to call GFINIS without having called GINIT. This wrote to the unopened TV and hung both POSSM and AIPS. Moved to 31DEC02 this date. 11200. December 24, 2002 LWPLA Eric Fixed so it will not repeat the setrgbcolor commands when not needed. Also fixed wrong format for setgrey. Moved to 31DEC02 this date. 11201. December 24, 2002 install.pl Eric The question about skipping the tar ball unpack had the wrong code for the default - requiring a Y or y to be typed to skip. Moved fix (not full file) to 31DEC02. 11202. December 28, 2002 KNTR Eric LTYPE=4 got the polarization vectors. Added a LOCNUM game for them too. Moved nowhere - should go to patch. 11203. January 2, 2003 FITLD Eric EIF and ECHAN were set to match the user's data at the first VLBA-correlator file. Thereafter they were used but not changed even if the user's data in a later file had a different number of IFs or spectral channels. Changed it to use the adverbs at each file, so that they get reset as needed. Moved nowhere - should go to patch. 11204. January 9, 2003 CONFI Leonia Two problems found by Mark Holdaway have been fixed. 1. BLC, TRC used for selection the portion of the mask can not be chosen >16384 because of AIPS limit. Adverbs BLC, TRC were substituted by array XPARM. The four first elements of the XPARM are used for BLC, TRC of the selected portion of mask. 2. The algorithm of finding the biggest side lobe of the beam failed to consider the nearest sidelobes. The algorithm that excluded the main lobe excluded some nearest sidelobes as well. I have rewritten the whole algorithm of finding the biggest side lobe. Moved nowhere. 11205. January 10, 2003 CONFI Leonia Some small changes added to the last modified version of CONFI. 1. Input parameter for elevation (BPARM(1)) has been removed and ELEV=90 fixed instead because only 90 has been used for elevation. 2. When optimization is provided together with fitting to topography (DPARM(5)=2) user could forget to put number of steps at X,Y and rotation to 1. Putting the steps (BPARM(7,8,9) zero has given the wrong default values. Now the step values are forced to 1 if DPARM(5)=2. 3. The task is terminated now with the relevant warning if number of the prohibited point at the mask is bigger than the given limit. Moved nowhere. 11206. January 10, 2003 IRING Eric/Patrick Leahy I found an older e-mail pointing out a bug in the coordinates used by IRING. The center was offset by one pixel in both X and Y from that claimed. Moved nowhere, should go to patch. 11207. January 10, 2003 SMODEL Eric/Michael Bietenholz Changed VISDFT to accept PARMOD(1) = 2 to be the same as 3, an optically thin sphere as stated by numerous help file such as CALIB and FRING. Also changed SMODEL.HLP. Moved nowhere. 11208. January 10, 2003 CALREF Eric Corrected several matters for CLCAL, CALIB mostly: 1. Added code to accept subarray 0 as any subarray. 2. Added code to suppress warning about failure when there are no non-zero phases. 3. Fixed it to stop unflagging flagged SN-table records. Moved nowhere. 11209. January 10, 2003 IMDIST Eric New verb to determine the distance between two pixels in one or two images. Changed AU9, POPSDAT, NEWPARMS and made a help file. Also added COORDINA as an output adverb to IMVAL, QIMVAL, and MAXFIT and proc help files TVFLUX and TVMAXFIT. Moved nowhere. 11210. January 10, 2003 output adverbs Eric It is about time that output adverbs be identified as such so that they do not appear on INPUTS only on OUTPUTS. In/out ones appear on both. Changed the help files for: BOX2CC CHKNAME CLR2NAME CLR3NAME CLR4NAME CLRNAME CLRONAME COPIXEL COWINDOW EGETNAME EPOCONV GAMMASET GET2NAME GET3NAME GET4NAME GETHEAD GETNAME GETONAME GETTHEAD GREAD IMPOS IMSTAT IMVAL IMXY MAXFIT QIMVAL REBOX SET1DG SETSLICE SETXWIN TABGET TK1SET TKBOX TKNBOXS TKSET TKVAL TKWIN TKXY TV1SET TVBOX TVFLUX TVMAXFIT TVNAME TVPOS TVSET TVSTAT TVWINDOW Moved nowhere. 11211. January 12-13, 2003 TVDIST Eric Added procedure TVDIST to POPSDAT.HLP; also help file. It reads two coordinates off the TV and invokes IMDIST. Moved nowhere. 11212. January 13, 2003 AU4, UVLOD, FITLD, TAPIO Eric Added test and message to the first three to show the first file name that could not be opened as well as the second (with the 1 appended). Fixed TAPIO to issue a BAKF error only when appropriate. Moved nowhere. 11213. January 13, 2003 ZLWIO, LWPLA Eric It is not correct to assert EPSF-3.0 on text files with more than one page. Changed ZLWIO to convert the EPSF line if present when concatenating a new plot and changed LWPLA to assert the EPSF only when INVER and PLVER are the same (after defaults are filled in). There are now smart "printers" that overprint all pages of a file if it claims to be EPSF. Moved nowhere. 11214. January 14, 2003 LOCIT Eric The Over-the-Top table could be read passed its last record when the times runded off slightly wrong - apparently an unlikley occurrence. Added code to set the OTT time to 10**10 when we have read the full table. Moved nowhere. 11215. January 17, 2003 $YPGNOT/CONFI Leonia The format of the topography file is free now. The help file has been updated. Moved nowhere. 11216. January 20-21,2003 LOCIT Eric When there is no data for an antenna, some parameters did not get set even though they were used in the summary displays at the end. The task also took the absence of an antenna in the last SN table time as reason to ignore the antenna entirely. Corrected the error handling to avoid this. Moved nowhere. 11217. January 21, 2003 LAYER, SCLIM Eric Two new related tasks. SCLIM creates images scaled from 0-1 using corners, function types, pixranges, etc. Clipped values - even at the top - may be set to 0 on putput. LAYER reads as many as 10 images letting each "absorb" the current image and add to that image by emssion with optical depths scaled from the input image values, all optionally colored. This will be tricky to use but good results are possible. Moved nowhere. 11218. January 21-22, 2003 FRING Eric/Amy FRING can apply the solutions and write out data when running on single-source input data sets. It did not prepare the header properly and so did not properly sum up the spectral channels when averaging them and used the wrong number of spectral channels when BCHAN or ECHAN are set. Moved nowhere. 11219. January 22, 2003 ABOUT files Eric ABOUT topics updated: ADVERB ANALYSIS CALIBRAT CATALOG EDITING FITS GENERAL IMAGE-UT IMAGING INTERACT PLOT POPS PROCEDUR RUN SPECTRAL TASK TV-APPL TV UTILITY UV VERB VLBI Moved nowhere. 11220. January 23, 2003 install.pl Eric Dropped page 10 which did nothing useful except confuse the users and renumbered pages 11 and 12. Added remark about host in disk names. Moved from Miller's machine this date. 11221. January 23, 2003 DBCON Amy There was an unnecessary radian to degree conversion in the calculation of DELZ (deltaZ). This effected shifting the phase center of the second dataset, although would only impact large shifts. Moved nowhere. 11222. January 24, 2003 UVCOP Eric UVCOP copies all sources in several different tables but does not copy all sources in the SU table. Certain table processing tasks like CLCOR may then encounter problems with sources in the CL table for which no coordinates may be found. Odd that this has not come up - the CL table has had all sources since OCT98. Dropped SUTIDY. Moved nowhere. 11223. January 24, 2003 GREYS.HLP Eric The colors assigned to images 3 and 4 were confused in the inputs and help - correctly stating grean and blue at least once but incorrectly saying green and red a couple of times as well. Sigh... Moved nowhere. 11224. January 27, 2003 GETSOU Eric Whoever wrote this piece of junk was a lazy bastard. It changed all the variables in common even when it did not work - but all routines outside GETSOU assume that it only changes them when it works. Fixed it to match the assumptions. Moved nowhere. 11225. January 27, 2003 DBCON Eric The code for phase shifting was only correct for very small shifts. It made no change to u,v,w and used Delta RA and Delta Dec linearly, not even sin's and cos's! Changed to use correct shift code with SHISIN, PRJMAT, and PRJMUL. Moved nowhere. 11226. January 27, 2003 CONFI Leonia The new option finding the worst "side lobe" has been added. The worst "side lobe" is found as the beam biggest value outside of the circle of given radius. Therefore side lobe is not necessary determined as an island at the two dimensional beam. It can be some elongated structure of the main lobe. The new option can be used for improving the main lobe shape. Moved nowhere. 11227. January 28, 2003 DBCON Eric Fixed a ZCLOSE LUN=0 error message - due to no FQ tables being present but TABIO ('CLOS' being called anyway. Moved nowhere. 11228. January 29, 2003 CLCOR Leonia The new OPCODE='ATMO' has been implemented. This option is useful to carry out atmospheric correction. The required correction should be given in the input file as a table of zenith delay for the set of antennas and times. This OPCODE was ordered by Mark Reid (CfA). The new version of GETSOU (#11224) is used. CLCOR may encounter problems if UVDATA were UVCOPed by the old version of UVCOP (see #11222) and if SOURCES = ''. Moved nowhere. 11229. January 31, 2003 CONFI Leonia The gain at the optimization loop depends on the distance of the found beam maximum from the beam center. Having the same gain CONFI carried out bigger reshuffling of antennas, which results in making bigger side lobes at other direction. I have introduce the option of the dynamical change of the gain. The gain is changed for each iteration depending on the found beam maximum from the beam center. This option may give smother the optimization process. The function of the gain dependence is determined by the user. Moved nowhere. 11230. February 3, 2003 CLCAL, SNSMO Eric Raised the time smoothing buffers from 2500/5000 to 25000. Moved nowhere. 11231. February 4, 2003 SAD Eric The VLB formatted display neglected to put out the source pixel coords when it could. Moved nowhere. 11232. February 4, 2003 MACPPC Eric For a MacIntosh OS/X port: We have problems with case sensitivity and so cannot name procs the same name as a system-level proc ignoring case. Rename AS AIAS, GREP AIGREP, PWD.C AIPWD.C, WHICH AIWHICH. Remove LGREP, PWD, CPWD.C from $SYSUNIX. Changed procs to use $SYSLOCAL/AIPWD if present else PWD: $SYSUNIX/AIPSCC, AIPSEXEC, BATER, COMRPL, COMLNK, FC, LINK MAKEAT, PP, RUN, SEARCH, SPHOP, STARTPMON, AIWHICH, $UPDUNIX/UPDCOMLNK, UPDCOMRPL, $INSUNIX/FTPGET, INSTEP1 $SYSPERL/update.pl, install.pl AS: $SYSUNIX/COMRPL, COMLNK. Add a case for MACs to STARTPMON, UPDERROR, UPDCONTROL, MAKE.MNJ, ZLPCL2, ZLASCL, TVALT, START_AIPS, LIBR, FCLEVEL.SH Also $SYSUNIX/LIBR does an "ar cr " operation with no modules to replace which Mac regards as an error - so added a case. $SYSPERL/install.pl - added code to recognize MACs and to accept the MACPPC as an architecture. Added /sw/lib in the list of possible locations for libreadline.a. Moved nowhere 11233. February 4, 2003 WRTPROCS Amy Changed FITDISK so that it checks whether INNAME and INCLASS are set. Also changed reference to tapes in FITDISK.HLP, to disk. Moved nowhere 11234. February 5, 2003 UVFIX Amy Added GMRT to the possible arrays with hardwired BZ, BY, BZ. Also fixed a divide by zero error when an antenna had reference coordinates of 0,0,0. Per Eric, made the calculation of DeltaZ more accurate. Moved nowhere 11235. February 5, 2003 LOCIT Eric Zero divide when first source missing in first record but refant not missing. Need to do inits on either antenna. Moved nowhere. 11236. February 5-6, 2003 MACPPC Eric Changed $SYSAIPS/AREAS.DAT and (autmagically) $SYSUNIX/AREAS.SH and AREAS.CSH to define $APLMAC. $SYSMAC, and $INCMAC for $ARCH=MACPPC. MACPPC and LINUXPPC use the Linux Z routines and then their own. Changed: XAS.SHR add MAC to Makefile Added: $SYSMAC/CCOPTS.SH, CDVER.SH, CDVER.CSH, COMLNK.ADD COMRPL.ADD FDEFAULT.SH INCS.SH LDOPTS.SH LIBR.DAT $APLMAC/ZDCHI2.FOR big endian, system type string ZERRO2.C sys_errlist already external - not allowed to do it twice ZFRE2.C statfs needed directly, changed member name ZTACT2,C dirent problems iff POSIX ZTKILL.C dirent problems iff POSIX ZTXMA2.C dirent problems iff POSIX ZMOUN2.FOR stubed for now ZTAP2.FOR stubed for now ZTPWA2.FOR stubed for now $APLLINUX/ZGRLTC.C dropped not needed and not present on Macs Moved nowhere. 11237. February 7, 2003 VLBAFIX Amy VLBAFIX now indexes (with INDXR) after USUBA, MSORT or if nothing needs fixing except indexing. Moved nowhere. 11238. February 7, 2003 MOMNT Eric MOMNT used the file at input catalog number = input disk number to make the first part of the output HI files to which the MOMNT HI is appended. Fixed the choice of pointer into the WaWa structures. Moved nowhere. 11239. February 7, 2003 UPDAREAS Eric Changed location of end if statement - it was only trying to create directories if AREAS.DAT was changed and AREAS.SH was not. This condition is prohibited in putbck. Moved nowhere. 11240. February 11, 2003 VLBALOAD Amy The final thing VLBALOAD does is deletes the NX table and creates a new one. VLBALOAD was croaking because the final INDXR did not have the correct name of the data file. Moved nowhere. 11241. February 11, 2003 POSSM Eric Yet another complex logic with error: plotting multiple plots with some data missing at start led to funny characters written to screen and TV I/o going to terminal.... Moved nowhere. 11242. February 12, 2003 FILLM Eric The channel 0 frequency increment was wrong. Moved nowhere. 11243. February 13, 2003 STARPL Eric Added more checks on the star position before calling the plot routines. Moved nowhere. 11244. February 18, 2003 INSTEP1 Eric Traces of ZTRLOG remained in this old - probably not too functional massive script. Moved nowhere. 11245. February 18, 2003 VLARUN Lorant New procedure for quick and dirty (ie blind) VLA calibration and imaging. Great for a quick look at archive files or your own VLA data, continuum or line, including high-frequency. Currently does not do polarization calibration (sorry!). 11246. February 21, 2003 VLBALOAD Amy Fixed bug that I introduced on Feb. 7 (it was connected to the VLBAFIX bug). Moved nowhere. 11247. February 25, 2003 UVDIF Eric Changed it to use different formats when OPCODE 'UVBX' and 'CLIP'. The old set are used for other OPCODEs. Also help file. Moved nowhere. 11248. February 25, 2003 UVPRT Eric The re-scaling of u,v,w by the frequency used BCHAN after a UVGET INIT call. At that point the correct channel to use is 1.0. Corrected. Moved nowhere. 11249. February 25, 2003 UVFIX Eric/Leonia Changed to do multi-source files and to interpret any "epoch" other than 1950.0 and 2000.0 as an apparent position which is converted to a J2000 coordinate. The use of an aberrated coordinate and aberated antenna position has been eliminated - that probably was not a correct usage. The GMRT appears to use a reversed sign for phase and u,v,w. This is detected and supported. The antenna center, time corrections, etc are taken first from the AN table unless UVFIXPRM overrides. This was corrected for array center and made more clear in the help file for the others. Note: GMRT does not provide the time corrections so the user must do that. Moved nowhere. 11250. February 25, 2003 POSSM Eric Again! A logic error in plotting amplitude of visibility caused it to include uninitialized or bad amplitudes rather than blanking them. Moved nowhere. 11251. February 25, 2003 IMAGR Eric QCLEAN was changed to defend against a BOX such as 343 129 0 0 resulting from a truncated input line into AIPS. Moved nowhere. 11252. February 25, 2003 TECOR Amy Starting 3 November 2002, the IONEX (input) files cover a time range from 0:00 to 0:00 the next day. This caused TECOR to break if more than one file was read in because TECOR ended up reading in two TEC maps with the same time at the day boundary. This has been fixed, so if there is a redundant time at day boundaries then it is ignored. Moved nowhere. 11253. February 25, 2003 ZFREE Eric Cleaned it up to let it be flexible in its choice of format depending on the maximum size of the disks. Moved nowhere. 11254. February 26, 2003 UVFIX Eric Added an option to set the Besselian epoch used to presecc coordinates from B1950 to J2000. Default is old value 1979.3. Moved nowhere. 11255. February 26, 2003 POPS stacks Eric Some very ambitious procedure builders have exceeded the stack size. Changed them from 100 to 1000. Required changes to DPOP.INC, INIT, POPSGN, BATER, GRIPR and lots of compiling. Moved nowhere. 11256. February 26, 2003 TASAV Eric Changed to work on images as well as uv files and to copt PL and SL files as well as tables. Moved nowhere. 11257. February 27, 2003 CONVL, DECONV Eric DECONV was changed to return IERR=0 under all circumstances and indicate failure with deconvolved axes = 0. Chaged it to count the failed axes in IERR. Changed CONVL to quit on a deconvolution error. Changed CONVL.HLP to say this and to say it correctly witout referring to an internal Fortran variable undefined in the help file. Moved nowhere. 11258. February 27, 2003 EDITR Eric QEDIUTIL: In an attempt to avoid excess FG table entries, it was possible to refuse to make flags when the first IF (if ALLIF) or first polarization (ALLPOL) were already flagged. Fixed four places for time flag, time-range flag, flag undo, and flag redo. Changed to copy any FG table applied to the input UV or table data to the (new) FG table. Changed EDITR.FOR to set the right keyword for IN_FGVER. Moved nowhere. 11259. March 3, 2003 SNSMO Eric It was able to return IRET = -1 to higher level routines if the last record in the table was flagged. It ignored and overrode flags in the table as well. Changed to test for flagged rows everywhere and set IRET 0 before branching to the end of the loop many places. Moved nowhere. 11260. March 4, 2003 CLIPM, UVMLN Eric Changed code to accept missing sources in a subarray or even all subarrays as worthy only of a warning message. FLGIT was fixed not to die on UVGET for a subarray having no data. Moved nowhere. 11261. March 4, 2003 VLBAUTIL Amy Fixed VLBAFIX, did the wrong thing if it needed to split into different frequencies and run VLBAFPOL. Moved nowhere. 11262. March 5, 2003 DQUAL Eric New task - removes all qualifiers from a source table, renumbering the sources in the uv data and source table. Moved nowhere. 11263. March 7, 2003 WENSS Survey Eric Fetched and translated the WISH (southern Westerbork -25 to -15 or -9 deg dec) survey. Inserted it in the WE50.0100 and WE00.0100 files in $AIPSTARS and in the full files available from the ftp site. Moved nowhere. 11264. March 11, 2003 UVCON Leonia There has not been any warning that the both image model dimensions must be 2**N. I added reading the image dimensions from the image model header. The task is aborted now with the relevant warning if at least one of the dimensions is not 2**N. Moved nowhere. 11265. March 11, 2003 UVSIM Amy 1) Removed all the unused code. 2) Added true IAT time calculation if the user inputs the Bx, By and Bz of the array site, previously UVSIM simply added 2.5 days to the hour angle and called that the time of the observation. I retained this if the user does not input the site location and warn user of incorrect times. 3) Calculates site latitude from Bx, By and Bz if they are input. Add message to tell user this. 4) Changed observation date from 1 Jan 1950 to 1 Jan 2000 and the observing epoch form 1950 to 2000. Moved nowhere. 11266. March 15, 2003 LWPLA, TVCPS Eric Added the CMYK option to TVCPS. In so doing, I discovered that the similarity between the RGB and CMYK displays had suddenly become greatly reduced. The system people insist that this is due to the repair of an error that previously existed in our printing system. It turns out that good displays can be obtained so long as different gamma corrections are used in TVCPS for RGB and CMYK. Changed TVCPS to apply the gamma corrections also to the graphics colors - this matters for some and for combined graphics planes although not for the pure colors. Changed LWPLA by adding RGBGAMMA adverb to the inputs to allow a similar control over the gamma corrections. I messed with the scaling options (DPARM of 1-4) some to make them somewhat more intuitive. Moved nowhere. 11267. March 16, 2003 IMAGR Eric Changed QCLEAN (s.r. CLRMAX) to add 'FIELDRES' keyword to each Clean field. Changed IMAGR, SCIMG, SCMAP to write this in the history file. Moved nowhere. 11268. March 17, 2003 DOBTWEEN Eric Changed POPSDAT.HLP and NEWPARMS.001 to create adverb DOBTWEEN which will control whether one interpolates and smooths between sources or only between observations of the same source. Moved nowhere. 11269. March 18, 2003 UVCON Leonia The same (as in UVSIM) type of error in time column is found in UVCON. This error and an error at the output AN table have prevented to give correct UVW data at the output of the UVFIX and other AIPS tasks which use the time of the visibility and AN table. The name of the array ANAME=VLBA at AN table has been added. The VLBA name is chosen because the Y coordinate at the AN table corresponds to VLBI (LH) coordinate system. The UVW data at the output of UVCON and UVFIX are coincide by absolute value but differ in sign. The difference in UVW sign is explained by different convention forming the baseline data of the antenna data. If we have antennas 4, 15 (for example) then VLA/VLBA carries out subtraction 4-25 and UVCON 25-4 to form the baseline data. Obviously the sign of UVW in UVCON has to be inverted. But the visibility sign has to be inverted also. I'll do the sign inversion later. Moved nowhere. 11270. March 18, 2003 UVFIX Leonia Added the line IF (ISVLB) XLONG = 0.0D0 to work succesfully at the case when LH equatorial coordinate system (VLBI) is used but the array center is not at the Earth center- the UVCON case. Add couple comments to clarify LH/RH conversion. Moved nowhere. 11271. March 19, 2004 UVSIM Amy Changed U,V,W calculation for baseline A(I):A(J), J>I from A(J)-A(I) to A(I)-A(J), which is VLA, VLBA convention, in order to clarify the calculation. Moved nowhere. 11272. March 19, 2003 UVCON Leonia Changed U,V,W calculation for baseline A(I):A(J), J>I from A(J)-A(I) to A(I)-A(J), which is VLA, VLBA convention, in order to clarify the calculation. Moved nowhere. 11273. March 19, 2003 UVFIX Eric The VLA and VLBA actually use BX = Bx(ant1) - Bx(ant2) etc. So changed the code to compute that as the baseline components and then dropped the arbitrary - signs from the conversion of u,v,w to wavelengths. There is not change in the output. Moved nowhere. 11274. March 20, 2003 FILLM, PRTTP Eric The disk reading handled the case in which there were short physical records in a logical so long as the first physical record was not short. But that occurs (continuum with < 26 antennas in the subarray), so the reading could get off and find no more data. The signal was a read error at the end with fewer than 13*2048 bytes but error code 0. Fixed by testing for short first record and correcting pointer into disk. Moved to patch area this date. 11275. March 21, 2003 install.pl Eric Added the command to rename the TEXT directory to TEXT.OLD just ahead of the unpacking of the tar ball. That unpacking is failing on TEXT, forcing when we are lucky a full cvs update of that area and potentially leaving that area with incorrect contents when the full update does not occur. Moved nowhere. 11276. March 21, 2003 WIPER ERic Added more code to keep the user from making the window too small both before and after the TVMENU call. Also added the DOCENTER option to control where the menu is placed so that one can avoid have data behind the menu. Moved nowhere. 11277. March 25, 2003 IMFIT, JMFIT Eric When the major axis of a deconvolution fails, the GAUSPS routine that changes pixels to arc sec will fail and so is not called. Added the code, already present in SAD, to patch up the problem by at least scaling the result to arc sec. Moved nowhere. 11278. March 27, 2003 UVCON, UVFIX Leonia Modify the AN table created by UVCON to make it be as VLBI observation table: LH equatorial coordinate system; axis X directs to the Green which meridian; ARRAYC=0; Removed the card: IF(ISVLB) XLONG =0 from UVFIX which is not necessary more having the new AN table created by UVCON. UVFIX is now does not have anything special to work with UVCON. Moved nowhere. 11279. April 4, 2003 data weights Eric The correct weight for an Ipol measurement is 4 * Wr * Wl / (Wr + Wl) where Wr and Wl are the > 0 weihts of RR and LL, resp. The routine DGGET used (Wr + Wl) / 2 as did GET1VS and GETVIS. Also improved the coding of VISCHK although it appears never to be used. This forces many link edits. Moved nowhere. 11280. April 6, 2003 FILLM Eric The last logical record could consist of one short physical record. Fixed the task to handle the case of short records, checking that the number of bytes = n*2048 and that the Modcomp structure is correct for the number of bytes found. In that case no significant error is generated. Moved nowhere. 11281. April 7, 2003 AIPS start up Eric To allow the $HOST name to be different than uname -n, I have created scripts SETNAME and SETUNAME which set a name in lower and upper case, resp. The main use is for laptops which, when they are plugged in, get names that change with time and plug-in location. The main trouble with this change is that sites will need to update the versions of these files that are in their $AIPS_ROOT. The name LOCALHOST and localhost are used so that Inet things will work. Changed: SETNAME Returns host-name in lower case - either the uname -n or localhost SETUNAME As SETNAME except in upper case. HOSTS.SH To call SETUNAME HOSTS.CSH To call SETUNAME START_AIPS To call SETNAME numerous places STARTPMON To call SETUNAME to set $HOST if empty. TVDEVS.SH To call SETNAME for TVDISP and TVHOST. LOGIN.SH Added environment variable LAPTOP LOGIN.CSH Added environment variable LAPTOP install.pl Move SETNAME and SETUNAME to AIPS_ROOT, add pause to allow edit of LOGIN.{C}SH. Moved nowere. 11282. April 9, 2003 On-line tapes Eric Added code to ZTPOP2.C and ZTPMI2.C to re-try and re-start the on-line modes for FILLM. Moved to AOC and VLA ONLINE areas this date. 11283. April 11, 2003 $APGNOT/SHADO.FOR Leonia The new task SHADO estimates the sensitivity loosing of an array because of shadowing of the array antennas by each other. The array geometry is given by INFILE and can be presented by four coordinate systems similar to those used by UVCON. The fractional sensitivity is determined as the ratio of the number of unshadowed antenna to the total number of antennas. Only antennas with elevation exceeded the given elevation limit are taking into account. The antenna is called shadowed if the blockage exceeds the given limit of blockage. The output file (table) includes the fractional sensitivities at the given set of declinations and hour angles. The table format can be chosen by user as LATEX or TEXT. The OUTFILE is not created if number of hour angles is .GT. 11. But the print out of the table still appears at the display. If LATEX then the output file is ready to create the output as a table in .ps format using LATEX commands (latex, dvips). You may want to edit the table capture before command 'latex' Moved nowere. 11284. April 13, 2003 errno.h Eric The standard unix error code should be obtained by #include not with the obsolete extern int errno RedHat 9 can detect that the latter was used! Changed numerous versions of ZPRMPT.C and a variety of other routines. Moved nowhere. 11285. April 14, 2003 ZERRO2, ZACTV9 Eric Another change is to complain about direct references to sys_errlist and sys_nerr. Changed ZERRO2 and ZACTV9 to call strerror - the thread safe version strerror_r is not available on Solaris. Moved nowhere. 11286. April 14, 2003 LISTR Eric The matrix average listed at the bottom of a MATX group was a simple average of the values in the matrix although, when the values are phases, they were at handled with sines and cosines. Added a line which averages the data entering in to the matrix during the averaging that produces the matix values. This is a vector average when the MATX values are vector - while the summary is ampscalar of the individual vector averages. If that is not complicated enough, it turns out the LISTR did a literal scalar average of phases when asked, thus giving 0 for -179 averaged with + 179 rather than 180. The rms's were not computed with a correct formula either. Changed both LIST and MATX for these. The MATX summary printed the average and the uncertainty of the average (i.e. avg/sqrt(N-1)) calling it sigma. Changed to give avg(uncer avg) sigma, where sigma is the rms of the numbers entering the average. Moved nowhere. 11287. April 14, 2003 BPASS Eric There were typos in the history writing that caused BPASSPRM(10) values of 1 and 2 only to appear in the HI file. Also added the intrpolation option. Moved nowhere. 11288. May 5, 2003 POSSM Eric Changed the units of RAEPO and DECEPO to degrees when shifting but then, in display code, changed the units again. Fixed the when shifting section to leave the units alone. Moved nowhere. 11289. May 5, 2003 IMSTAT, IMEAN Eric Changed AU6D and IMEAN to allow for circular apertures, with BLC = -1,radius and TRC = Xcen,Ycen. Moved nowhere. 11290. May 5, 2003 PLCUB Eric Added option (APARM(3) <= -11) to omit the frame axis labeling. Moved nowhere. 11291. May 6, 2003 GREYS Eric LOCNUM confusion caused LTYPE 4 to be done as LTYPE 3 when contouring. Moved nowhere. 11292. May 8, 2003 SNSMO, CLSMO, CLCAL Eric Worked on the smoothing operations. The subroutines are BOXBSM New call seq include list of source number; rewrote algorithm to honor support size, return blanks if needed. Previously it extrapolated a "good" value in all directions if needed to avoid blanks. SOUR(1) < 0 causes source array to be ignored. MWFBSM as BOXBSM FUNBSM New call seq - list of source numbers and cutoff for sum of conv. function. Add EXP function. TPTBSM New call seq - list of source numbers. CALREF Call BOXBSM rather than BOXSMO with source = -1. Call only if smooth time > 1.1e-6. REFFAZ Call BOXBSM rather than BOXSMO with source = -1. MBDREF Call BOXBSM rather than BOXSMO with source = -1. Call only if smooth time > 1.1e-6. QEDIUTIL Fix call seq to functions, adding source=-1. Allow EXP smoothing. SNEDT Help file had EXP function added plus remark about smoothing across all sources. HLPEDISN Ditto. CALSEL Changed to re-write the CL/SN table rather than write the selected records as flagged. The latter run the risk of becoming unflagged. Replaced BOXSMO everywhere, moved it to $APLCONTR. Also moved GNFSMO, REFDLY, REFRAT from $APLNOT to $APLCONTR since they are not used. The tasks are: SNSMO New adverbs DOBTWEEN (smooth all sources together?), new values for SAMPTYPE (was INTERPOL), new function parameters (BPARM of 6 - 10), CUTOFF to limit function smoothing, DOBLANK with new values (>= 0 replace blanks, <= 0 replace good values). The biggest changes are in honoring the function support size and in the control over what os replaced by smoothed values. CLSMO This little known task was changed just like SNSMO. CLCAL Changed default GAINVER and GAINUSE to highest and highest+1, resp. If GAINUSE=0, GAINVER is copied to highest+1 before the operation begins. Added adverbs SAMPTYPE, BPARM, ICUT, DOBTWEEN to control smoothing of the merged SN tbale independent of the interpolation methods to be used in converting the SN table to the times of the CL table. The smoothing for phase is now done in real and imaginary rather than in phase (with some attempt to avoid 2 pi ambiguities). The full range of smoothing functions from SNSMO is now available, although the latter will handle full VLBI SN tables better. Moved nowhere. 11293. May 8, 2003 $YPGNOT/SNPLT.FOR Leonia GRCHN in SNPLT has not been change to 1 if GRCHAN=0, although it is said in the .HLP file. I have fixed it. Moved nowhere. 11294. May 9, 2003 VLBACALA (VLBAUTIL) Amy Changed DOBLANK=1 to DOBLANK=-1 for SNSMO in VLBACALA, because of changes in meaning of DOBLANK mentioned above. Moved nowhere. 11295. May 12,2003 $YPGNOT/ELINT.FOR Leonia RMSs of fit parameters have been calculated erroniously as their variances. I have fixed it adding SQRT. Moved nowhere. 11296. May 12, 2003 repair Eric Somehow, the old version of BOXBSM was restored, not the new one I had written and tested. Replaced it with a new one in $APLSUB, deleting the $APLNOT version. Also fixed FUNBSM since it had an error and I used it to make the newly recovered BOXBSM. Fixed a VLBI woarning in CLCAL also. Moved nowhere. 11297. May 12, 2003 SNPLT Eric The above "fix" by Leonia is incorrect. GRCHAN=0 has special meaning - to use multiple graphics channels. Fixed the help file to stop saying 0 => 1 and to mention the special meaning attached to GRCH=0 in ALL plot tasks. Fixed the help file description of 'SUM' plots and the labeling of them as extrema. Changed scaling to fit the TV when DOTV true. Moved nowhere. 11298. May 14, 2003 INDXR Eric It used WXTABL = ' ' as an indication not to do any opacity or gain correction even though the WXTABL is not needed for gain or for user-specified opacity. Fixed that and some other minor oddities. Moved nowhere. 11299. May 14, 2003 VLA.GAINS Brian Butler Updated for runs in Feb 2003. Moved nowhere. 11300. May 15, 2003 data weights again Amy/Eric The Ipol weights fixed 4/4/03 are now okay - but it was getting wierd (double) weights when doing only RR or only LL etc. Fixed DGGET, GETVIS, GET1VS to halve the weights when the "two" parts of the vis are the same. Moved nowhere. 11301. May 15, 2003 FILLM Eric Added DOWEIGHT=10 to set the weights larger by the factors suggested by Bryan Butler in AIPS Memo 108. Moved nowhere. 11302. May 16, 2003 TVRGB Eric Corrected several errors related to XINC and YINC. Changed adverbs BLC/TRC/XINC/YINC to the TBLC/TTRC/TXINC/TYINC set. Then added BLC/TRC to contro the part of the image that is written out at the end. That file always ignored the Xinc and Yinc and reread the image from disk. However, the wondow was restricted by the TV size. That restriction is now lifted. Moved nowhere. 11303. May 20, 2003 TVHUI Eric Changed the help file to correct spelling, to use TBLC et al rather than BLC et al. Made more clear the fact that TBLC, TXINC etc are ignored on output of the RGB image. Moved nowhere. 11304. May 20, 2003 OPTIMIZE.LIS Eric Added GINITL for no optimize on SUL machines. Chris Phillips had trouble with the Solaris 5.0 f77. It will not hurt to have this one no-opt so let's do it. Moved nowhere. 11305. May 21, 2003 INDXR Eric The variable VLBMOD confused me in April 2002 - it is for some special VLB subarray mode, not VLB generally. So, when I entered the VLA gain correction code I did an IF-THEN-ELSE on VLBMOD which kept the IM and MC tables from setting values in the CL table for many VLB experiments. Moved to the 31DEC02 patch area this date along with 5/14 fix. 11306. May 22, 2003 MCUBE Eric Check INSEQ, IN2SEQ, IN3SEQ to see if anything to write and quit if not with a message. Moved nowhere. 11307. May 29, 2003 MOMNT Eric Overhauled the code and fixed it to handle blanked pixels correctly. Smoothing 0's as if the image were zero is not correct. It must really know that the image was blanked. Raised the buffer sizes to 4096 channels and 4 million for the in-core smoothing buffer. Moved nowhere. 11308. May 29, 2003 LISTR Eric 1. The rms computations I added recently were in error. They should be okay know. What was I smoking? 2. The gain display did not check for flagged data underflowing with the sign character. Replace overflows with ******** and underflows with --------. Moved nowhere. 11309. May 29, 2003 UVCOP Eric UVCOP did not flag data that had FQ set in the flag table unless one also specified a single FQ to be copied. This error was made in August 2000 to fix the fact that in the select one FQ case, the FQ number was already altered to 1. Now it does a proper IF-THEN-ELSE on FRQSEL to get an FQ to compare to the flag. Moved nowhere. 11310. May 30, 2003 UVCON Leonia Simulation of a random phase and amplitude for each antenna has been added. It can be done for only one(the first) clean component of the model (as Rick P. asked) or for all CC components of the model. The new random try is done each given time. Moved nowhere. 11311. May 30, 2003 ZLPCL2 Eric The script attempted an lpq on ghostview - fixed the chcmd to be null in that case. Also gv does not read from standard-in under any circumstance. Changed proc to create $file.ps, then view it, then delete it. Now seems to work. Moved nowhere. 11312. June 2, 2003 FILLM Bryan Butler Added option to choose to not load visibilities for which one of the antennas failed its last reference pointing attempt (bit 10 of CPARM(2)). The default is to do as it does now - load all the data. but if you do that, then it makes an OF table which holds that bit (along with the antenna shadowed bit). the OF table routines (DOFV.INC, OFINI.FOR, and TABOF.FOR) are also attached. the OF table is also created if you choose to pass shadowed data at any level. Later, could make a task that does the flagging after the fact (convert the OF table into entries in an FG table). Moved nowhere. 11313. June 3, 2003 MOMNT, XMOM Eric Added the primary beam correction option to MOMNT as implemented in XMOM. Corrected the routine in XMOM to avoid excess subroutine calls - it was okay, but did way too much work. Moved nowhere. 11314. June 4, 2003 GRIPE Eric Changed ZGMAI2.C to send gripes only to aipgripe and daip. Pat and I do not need private copies and Ernie does not need to discard the aipsmail one. Moved nowhere. 11315. June 5, 2003 IMAGR Eric Changed the beam-size setting: if IMAGRPRM(10) <= 1.5, then the beam is 2048 or less on a side. If it is set to 2, then the limit is 16384. Chaged help file. Moved nowhere. 11316. June 5, 2003 TVILINE, IM2TV Eric New verbs: draw a line on the TV between image pixels PIXXY and PIX2XY and find TVXY given PIXXY. Changed POPSDAT, NEWPARMS, Au5B, and made a new help files. Moved nowhere. 11317. June 5, 2003 install.pl Eric Registration does not work so comment out the registration portion. Moved nowhere. 11318. June 6, 2003 GFINIS Eric Changed the TV delay function. If you hit button A, the 30-second delay is made indefinite while B & C resume sooner and D quits. After A is hit, all buttons are resume except D which is quit. Moved nowhere. 11319. June 11, 2003 install.pl Eric Changed default $AIPS_ROOT to `pwd` rather than $HOME/AIPS. The latter is almost never wanted. Moved nowhere. 11320. June 14, 2003 FILLM Eric Bryan missed one changed call sequence which caused aborts on Linux, at least with some compilers, and a failure to separate bands on Solaris. Moved nowhere. 11321. June 17, 2003 FDEFAULT Eric Changed $SYSLINUX version to 3.2 level options, leaving the 2.95.3 as comments. Added comment to install.pl about 2.95.3. Updated web pages some about this. Moved nowhere. 11322. June 18, 2003 TVCPS Eric Drop the AIPS USER number string at the bottom and the APARM(8) option to suppress it. Moved nowhere. 11323. June 18, 2003 TVHUI Eric The offset was not correct in the scaling of the hue and saturation images when writing the output image. Also fixed the final effective pixranges which used the wrong scaling array in their computation. Moved nowhere. 11324. June 24, 2003 FUNBSM Eric Drop error message on 0 width - there is no problem there. Moved nowhere. 11325. June 24, 2003 INDXR, FITLD Eric The dispersive delay (unlike clock and other things in the CL table is actually applied to correct data. Therefore, when the CL table 1 is created, it must be forced to 0. Any dispersive delay applied in the IM table has been applied and hence must be ignored here. Moved nowhere. 11326. June 24, 2003 MBDREF Eric The change to make MBDREF call BOXBSM only when the smoothing time was meaningful caused it to fail to fill the WORK1 array from the input data array. The WORK1 array (in dynamic memory) the was trash at least on DEC alphas and so led to math problems. It did not work right of course on other machines. Moved nowhere. 11327. June 25, 2003 VLBAPIPE Lorant Fixed a bug in VLBAPIPE that caused the FRINGE rates to be zeroed. Moved nowhere. 11328. June 25, 2003 SOUFND Eric When there is no SU table it still called the TABIO CLOSE and this makes a ZCLOSE: INVALID LUN = 0 message. Fixed. Moved nowhere. 11329. June 25, 2003 CLCAL Eric The author of this task was, when he wrote the single-source aspects, smoking some illegal substance. Fixed it to merge and smooth SN tables suitably for single-source files. It would do only one SN file, probably to smooth it, but used some of the GAINUSE/GAINVER adverbs to control it partly. Also fixed the code to be more readable - the author had not used IF-THEN-ELSE much. Dropped the SMOO operation - MERG suffices with smoothing depending on the now separate adverb SAMPTYPE. It will quit on CALI for single-source files. Added adverb INVERS to allow a range of SN tables to be used, more than one but less than all. Changed POLY to use the CUTOFF to set the order. It conflicted with the smoothing which was a problem I forgot to solve. Moved nowhere. 11330. June 27, 2003 SETFC, FIELDF Eric Both had formatting problems with field numbers > 999. Moved SETFC to patch area. 11331. June 29, 2003 VLARUN Lorant Now that we're really announcing it, I made it more robust in in the inputs (i.e. no defaults allowed). Also some minor fixes and additions (e.g. allow phase cal sources='*' for continuum). Moved nowhere. 11332. July 1, 2003 3 color plots Eric Changed POPSDAT New adverbs: AVGIF (average IFs?), POL3COL (color polarization vectors), CON3COL (color contours), DO3COL (use 3-color vectors elsewhere). NEWPARMS Ditto. DGPH.INC Add current 3-color line color GINIT Init 3-color line color to black G3VCOL New: add 3-color vector values to plot file G3VEC New: write a 3-color vector IMVECT Changed to call TVVECT. TVVECT New: write a TV vector in any channel of a specified intensity. COLOR3 New: returns an RGB color for vectors. CONDRW Change call sequence to add DOCOLR and take that as a vector choice. LWPLA Fixed "feature" - if OFMFILE was set but there was no grey-scale the colors got botched on DOCOLOR false. Added support for the new plot command types. TVPL Added support for the new plot commands. Fixed the init channel to turn off all colors when turning on only one. TKPL Ignore the 3-color color command and draw vector on color vector command TXPL Ignore the 3-color color command and draw vector on color vector command PCNTR Added POL3COL option including color spray to show the color vs angle. Add CON3COL option to overplot multiple contour drawings in colors. KNTR Added POL3COL, CON3COL options. AU8A Update PCNTR and KNTR. Moved nowhere. 11333. July 1, 2003 TVPLCOL Eric New standard procedure in POPSDAT.HLP - sets adverb PLCOLOR to the colors of the graphics channels that DOTV would use given the GRCHAN setting. Also help file. Moved nowhere. 11334. July 3, 2003 VLBAUTIL Amy Edited VLBALOAD, VLBAMCAL, VLBACALA, VLBAPANG and VLBAFPOL. Added DEFAULT to all tasks that are run by these precedures in order to prevent the creation of "hidden parameters". This could arise with Chris Flatter's system of setting all parameters to a task ifinputs were added to a task. Moved nowhere. 11335. July 3, 2003 Usage Eric Found by Fortran 95 compiler (run time abort): AXEFND Declared string as *8 when calling routine might use (did use) something less. You must use *(*). Moved nowhere. 11336. July 5, 2003 TVCOLORS Eric The proc name TVPLCOL cause GO TVPL to fail. Renamed TVCOLORS. You will need to say SCRATCH TVPLCOL; COMPRESS to clean this mess up. Moved nowhere. 11337. July ??, 2003 3-color plotting Eric VPLOT Add adverbs APARM(7) (buffer size), AVGIF, and DO3COL to help file and Fortran. Rewrote code to use dynamic memory and allow for multiple IFs in a single plot w or w/o 3-colors used in symbols. BPLOT Corrected scaling on large FACTORs, added DO3COL to color the multiple spectra drawn on a single plot. UVPLT Added DO3COL to color the channels and IFs. SNPLT Added DO3COL to color the channels and IFs or antennas in the SUM mode. Moved nowhere. 11338. July 7, 2003 tapes Eric Changed $APLUNIX/ZONLIN.C and ZTPOP2.C so that an error code from the stubbed ZONLIN (olopen) will be recognized as a non-online attempt to open online "tape". Moved nowhere. 11339. July 7, 2003 AIPSPATH Eric These procs have set MANPATH. The Unix utility man now has an elaborate scheme that we should not be messing with. Changed them to add AIPS to a pre-existing MANPATH, but to leave an empty MANPATH alone. New $SYSUNIX/AIMAN will do man aips for those with no $MANPATH. Moved nowhere. 11340. July 10, 2003 PCNTR, KNTR Eric Failed to handle CON3COL=0 properly. ZINC was zero -> abort, but setting it to 1 also caused problems. Both KNTR and PCNTR overwrote FACTOR thereby destroying its value for later planes. Changed PCNTR to plot multiple planes of polarization vectors colored by axis 3 position when DOCONT<= 0, CON3COL > 0, and DOVECT > 0 and the I, Q, and U images are matching cubes. Upgraded wording in both help files. Moved nowhere. 11341. July 10, 2003 Tic drawing across Eric CTICS, ITICS, TKTICS had things to detect excess skips in DOCIRCLE that failed on small images. Computers are faster now so change things to do finer increments especially in small images. Moved nowhere. 11342. July 11, 2003 FILLM Eric The VLA shifts its center for Q band data with Pie Town - too bad no one told us. I have added a trap for this with a shift back in the antenna coordinates that end up in the AN table. This change had to take place two places in the task and was not as trivial as it sounds. Moved nowhere. 11343. July 11 & 13, 2003 Plot programs Eric VPLOT Let flagged IFs leak through as zeros when averaging. UVPLT Put wrong NPARMS in plot file, did not insert defaults for many things. SNPLT Set Stokes to 5 characters at one point. UVHGM Addressing error in extra parameters added for EXTLIST. BPLOT Fill in adverbs, use type 30. Moved nowhere. 11344. July 13, 2003 EXTLIST Eric Overhauled AU8A so that it no longer combines things together to try to save a few lines of code at the cost of greater complexity. Then for VPLOT Updated a few pointers, corrected quite a few CLPLT Corrected a bunch of pointers. BPLOT Added a section UVPLT Upgrade Changed to allow more than one plot file inputs type (will allow up to 3). Ones with 2 versions now are UVPLT (no diff), SNPLT (no diff), VPLOT differ, PCNTR differ, KNTR differ, BPLOT (no diff). Time-range displays now can say beginning to end. Moved nowhere. 11345. July 13, 2003 IMFIT, JMFIT Eric Put the deconvolved Gaussians into the CC file with the wrong units for width (asec not degrees) and wrong type (convolved not deconvolved Gaussian). SAD appears to get these right. Moved nowhere. 11346. July 14, 2003 START_TVSERVERS Eric It was passing $DISPLAY through to XAS et al. even when the value was misleading (e.g. localhost:11.0 from ssh). When the tvhost and the tvdisplay are the same anyway, it should set $DISPLAY to ${tvdisp}:0. This may cause problems if people have things set to e.g. :1.0. START_AIPS: type "screen" added to list of terminals THIS HAS TO BE MOVED BY HAND TO $AIPS_ROOT TO TAKE EFFECT. Moved nowhere. 11347. July 14, 2003 LISTR, UVFLG Eric Added digits to the display of scaling in LISTR. Changed UVFLG (and help) to allow 40000 entries in the external file. Moved nowhere. 11348. July 14, 2003 SPFLG, TVFLG Eric These could get two times at the same apparent time due to averaging and a change of source. Fixed the row selection routine to understand this circumstance. Otherwise, error messages came from the clip be set numbers. Moved nowhere. 11349. July 17, 2003 FILLM, PRTTP helps Eric Changed the direct address of the archive data into info about ordering it from the analysts. Naive (stupid?) people were abusing this information. Moved nowhere. 11350. July 17, 2003 tick labeling Eric Cell sizes in micro-arcsec with good coordinates! Changed TICINC Increased the range of numbers possible. TICSTR New: make the units string label (so only 1 place has to know the meaning of ITRY) ITICS Use TICSTR instead of own labeling CTICS Use TICSTR instead of own labeling TKTICS Use TICSTR instead of own labeling CHNTIC Use TICSTR and just take max from that. CLAB1 Used AXINX(i,LOCNUM) = 0 as clue not to call CTICS. That could not work unless both were 0 because CTICS on one axis would fail if the other had 0. Changed to trap CTYP='NO TICKS' to avoid label, tick marks, and tick labels. CLAB2 As CLAB1. PROFL Use TICSTR. Also added code to avoid doing overlapped strings. FRPLT Use TICSTR instead of own labeling. POSSM Use TICSTR instead of own labeling. FGPLT Drop local CLAB1 in favor of new version. IBLED Use TICSTR instead of own. LWPLA Check for and comment out overlapping strings. Works only for tick labels on the X axis KNTR Get rid of local version of GTICS (identical to $YSUB/GTIC) Also changed comment that used the name TICINC in $YPGNOT tasks: CLPLT, CONFI, ELINT, LOCIT, PBEAM, SNPLT, VPLOT, WETHR Moved nowhere. 11351. July 18, 2003 AXSTRN, JMFIT Eric Changed AXSTRN to use greter accuracy at times. Created COORDI as a version of COORDD with SEC double precision rather than single. Revised displays in IMFIT and JMFIT for very small cell sizes. Lots of link edits are forced. Moved nowhere. 11352. July 21, 2003 FILLM Eric On-line FILLM quits when a new observer is found with data. Added a counter to do this quit only if data were actually written to the disk for the current user. Moved nowhere. 11353. July 21, 2003 POSSM Eric While debugging the gnu 3.2.2 compiler I noticed that the GETFQ call used IERR not JERR and the SOUFIL call did not test its error return. Moved nowhere. 11354. July 21, 2003 SN2CL Eric The CUTOFF adverb was applied badly - it cut off a solution if either of the calibrator scans was too far away - even if one of them was quite close. Relinked CLCAL, BLAPP. Moved nowhere. 11355. July 21, 2003 DATBND Eric The shifting for VLBI used an IF- but not source-dependent frequency for to lower numbered antenna and the header frequency for the other antenna. This does not match BPASS which uses source-dependent frequencies and the same for all antennas. It was easy to fix since everything is known at this point. Moved nowhere. 11356. July 22, 2003 DBCON Eric Add warnings about mismatched SU tables on NUMIF. Moved nowhere. 11357. July 22, 2003 Mac FDEFAULT.SH Eric Apparently the new Mac compilers do not like "frame pointers" whatever they are. Added -fomit-frame-pointer to the standard options list. Unfortunately, system things like readline are compiled with frame pointers and the newest systems do not have any library to resolve references to restFP and saveFP! Moved nowhere. 11358. July 27, 2003 POSSM Eric The UV averaging just advance the counter by NCHAN when the full polarization/IF was empty. This left the buffer marked good although all zero. Fixed to ignore the SUMWT here and process each channel properly. Also defended against pure zero amplitudes. Will we ever remove all these egregious bugs from this task? Moved nowhere. 11359. July 28, 2003 VPLOT Eric The averaging routine did not handle the end of data properly when no data were found. It looped for more under a variety of conditions with no check on end of data. Moved nowhere. 11360. July 28, 2003 $YPGNOT/DELZN Leonia The new task restores the residual atmosphere delay at the given elevation observing calibrators at wide range of elevations. The task reads the SN table, taking the information about the delays at given antenna observing given calibrator. This delay is contributed by delay at the atmosphere and delay due to the clock error. The task separates the atmosphere and clock delays using the least square method to fit polynomial of the chosen degree to the clock error and polynomial of the chosen degree for the zenith atmosphere delay. Having known these polynomials the task calculates the zenith atmosphere delay and records it to the output file ready for use as input file for the task CLCOR (option 'atmov'). The fitted zenith atmosphere delay and clock error is used to correct the delay for target sources putting the correction to the output CL table by the same way CLCOR does at option 'atmov'. Moved nowhere. 11361. July 29, 2003 TVMOVI Eric Corrected both AU5D and TVMOVI for bugs related to actually using MAG=1 for an image (e.g. 1024x1024x4). Moved nowhere. 11362. July 30, 2003 SNFLG Eric New adverb DOIFS added to POPSDAT, NEWPARMS, and a new help. SNFLG changed from a single IF and polarization to allow two polarizations and any number of IFs. Adverbs to control whether all IFs and.or all polarizations are flagged when one is were also added. The code now uses dynamic memory, correctly picks up missing antennas as "bad", flags bad antennas as antenna rather than baseline flags, and works hard to keep the number of flags down. Upgrade DOSTOKES.HLP also. Moved nowhere. 11363. July 30, 2003 SNSMO Eric Suggested correction by Olaf Wucknitz, University of Potsdam, Germany: The attempt to make phases deal with phase wrap in SMOAPH would not function across failed solutions. Fixed it to keep track of the last good phase and to use that for wrap rather than using only the previous if it is good. Moved nowhere. 11364. July 31, 2003 FITLD Leonia FITLD has not filled out the CQ table column NO_BITS, forcing it to be equaled zero always. Fixed considering all antennas of the given subarray have identical digitizers. Moved nowhere. 11365. August 1, 2003 IMAGR Eric Added IMAGRPRM to history and help files. In QCLEAN, made IMAGRPRM(19) limit the weakest component relative to the strongest in any major cycle. It was going too deep in high dynamic range cases and getting lots of bad + and - bumps to fix up later. Also added an initial value to the ATLIM magic parameter that cuts of a major cycle. It is 1% of the difference between the max and min values loaded except near the start. Moved nowhere. 11366. August 4, 2003 CLCOR Leonia CLCOR has been corrected the given CL table. This required the user to take a precaution to escape possible undesirable corruption of the CL table. The new version of CLCOR copies the given CL table to the (high+1) number and the last one is corrected. Moved nowhere. 11367. August 4, 2003 ELINT Leonia ELINT has been applied the correction to the output SN table. The user should convert the SN table to the CL table using the task CLCAL taking care about possible undesirable smoothing of the SN table data over the different sources. The new version of ELINT applies the correction directly to the output CL table. The user should select the CL table number. This CL table is copied to to the (high+1) number and the last one is corrected. Moved nowhere. 11368. August 4, 2003 LTESS, UTESS, VTESS Eric The primary beam corrections were not up to snuff here. Changed DTESS.INC Add PBPARM(7) instead of PBCO APLPBI Changed to implement the PBCALC call. APPLPB Changed to implement the pBCALC call. LTESS Change meaning Of PBSIZE: 0 => use new adverb PBPARM, > 0 => Gaussian, < 0 => no beam corr UTESS As LTESS VTESS As LTESS STESS As LTESS Moved nowhere. 11369. August 4, 2003 OPTIMIZE.LIS Eric Added IOBSRC and PRJMAT to the no-opt list for LInux 3.2.2. This will be easier than installing 3.3 everywhere. Moved nowhere. 11370. August 4, 2003 CAL Adverbs Eric The calibration adverbs are not all present in all tasks. Changed AU8A For FRMAP, FRPLT (was rather wrong) FRMAP Add BLVER FRPLT Add BLVER IBLED Add DOPOL POSSM Add BLVER SPFLG Add DOPOL TVFLG Add DOPOL CLIPM Add DOPOL, BLVER FLGIT Add DOPOL, BLVER SPCAL Add DOBAN, BPVER, BLVER UVFIT Add BLVER UVFND Add DOPOL, BLVER, DOBAND, BPVER UVMLN Add DOPOL, BLVER Moved nowhere. 11371. August 5, 2003 VPLOT Eric Changed EIF to the usual default, fixed STOKES s.t. ' ' actually means 'I' and 'HALF' becomes 'RR' (or 'LL' is there is only 'LL'). Moved nowhere. 11372. August 5, 2003 FRING Eric Changed APARM(6) meaning - at level 0 the time comes out every 10th time as a progess message. At level -1, no messages. Moved nowhere. 11373. August 5, 2003 GPHAS Eric Corrected bug I put in in 2000 with 'Inputs' rather than Ketan's non-standard POPS. Added Kogan's remarks about FRING with DPARM(1)=3 to steer people away from this task. Moved nowhere. 11374. August 5, 2003 CCEDT Eric There was one array declaration in error and circular windows were not properly supported. Also, the HI file was not written with windows from the BOXFILE. Moved nowhere. 11375. August 5, 2003 calibration Eric Cleaned up a few of the help files from yesterday and added calibration to RFI and MAPBM. Moved nowhere. 11376. August 6, 2003 GPHAS Eric This task is not known to work and FRING solves its problem according to Leonia, so move it to $APLCONTR. Moved nowhere. 11377. August 6, 2003 CLCOR Eric Added GAINUSE to adverbs. If not GAINVER then a new CL is written, else the one CL is used in place. Moved nowhere. 11378. August 6, 2003 cal adverbs Eric When calibrating some things, need to be able to do all: BLCAL Added DOPOL, DOBAND, BPVER BPASS Added BLVER, DOBAND, BPVER, changed BPVER to OUTVERS CALIB Added BLVER, DOPOL. CPASS Added BLVER, DOBAND, BPVER, changed BPVER to OUTVERS DEFLG Added BLVER, DOPOL. FRING Added BLVER, DOPOL. KRING Added BLVER, DOPOL. LPCAL Added BLVER, DOBAND, BPVER but not DOPOL PCAL Added BLVER, DOBAND, BPVER but not DOPOL BLING Added DOPOL, BLVER. OMFIT Added DOPOL, BLVER, DOBAND, BPVER SCIMG Added DOPOL, BLVER, DOBAND, BPVER SCMAP Added DOPOL, BLVER, DOBAND, BPVER Moved nowhere. 11379. August 7, 2003 $YPGNOT/DELZN.FOR Leonia Fitting Polynomial at the axis of time DELZN has used actual time axis of the data. Fitting could be better if the time starts at the center of the data. It is done. Calculating of the atm. zenith delay rate and clock drift has been correct only for three term at the polynomial fitting of zenith delay. It is fixed. Moved nowhere. 11380. August 12, 2003 SMOTH Eric Improved the code to handle blanked pixels, replacing them where possible with smoothed values. Did some cleaning up of the typing as well. Added comments to the help about units and CONVL. Removed the erroneous claim that a history is not written. Moved nowhere. 11381. August 12, 2003 Model scaling Eric FACSET New subroutine: Sets FACGRD to correct for (1) beam area in Clean image models and (2) the sum of the model CC flux to the limits set by standard adverbs such as FLUX, NCOMP, BCOMP in order to match the total flux given in the source table (if any). BLING I cannot find a way to insert FACSET here - added comment to that effect in the help file and cleaned up the code typing some. CALIB Depended on a task called CCFND to set a header keyword SUMFLUX based on some dubious assumptions. Then, for the first wanted source, if a flux was set, the model scaling parameter was adjusted to produce this total flux. UVSUB Added FACSET - since UVSUB only works on single-source files, the CC scaling part is not in effect. Added cals to FACSET in BLCAL, BPASS, CPASS, FRING, KRING, LPCAL, PCAL, CLPLT, IBLED, VPLOT as well. Moved nowhere. 11382. August 18, 2003 Source azimuth Craig/Eric Craig has supplied us with a new version of SOUELV that computes azimuth as well as hour angle and elevation. Changed TABLEUTIL routine OSUELV to return all 3 of these. Changed QUVUTIL, ACFIT, ANCAL, CLCOR, LISTR, CALIB, APCAL, WFCLN, CL2HF, and INDXR to use the new call sequences with az ignored. Changed MAPBM, APGPS, and TECOR to use the new call sequence and to skip computing azimuth on their own. Changed UVPLT and WIPER to use the new call sequence and to offer a new option to plot against azimuth. Moved nowhere. 11383. August 13, 2003 installation Eric Changed MAKE.MNJ to use $COMPILER internally and SETNAME rather than uname -n. Changed install.pl to 1) copy MAKE.MNJ to $SYSLOCAL 2) edit MAKE.MNJ to use the declared C compiler 3) to make skipping the download of the tarball the default. Moved nowhere. 11384. August 14, 2003 UVHOL Eric Fixed up STOKES to do true Stokes as well as RR/LL/RL/LR. Moved nowhere. 11385. August 14, 2003 COODEFIN Eric Created new verb COODEFIN in AU7B with new adverbs COOREF and COOINC. It redefines the RA/Dec or xLon/xLAT values including increments, reference pixels, and rotation all at once. Also POPSDAT.HLP and NEWPARMS.001. Moved nowhere. 11386. August 19, 2003 VLBAUTIL Amy Fixed VLBALOAD so that it will not crash during INDXR if the file loaded is not in TB order. Changed a message in VLBAFXPL, about the possibility of switched polarizations for foreign stations. Changed VLBAPANG per new CLCOR, it no longer runs a TACOP first. Moved nowhere. 11387. August 26, 2003 more 3-color Eric Enable 3-color plotting: PL3VEC New: Gary-package to call G3VEC PLOTWO Add COLOR to call sequence PNTPLT Add COLOR to call sequence AU5B Change call seq to PNTPLT CTICS Change call seq to PLOTWO STARPL Change call seq to PNTPLT CCNTR Change call seq to PLOTWO IRING Change call seq to PNTPLT LOCIT Change call seq to PNTPLT PLOTR Change call seq to PNTPLT Moved nowhere. 11388. August 28, 2003 PLOTR Eric Changed to have an extra color value column optionally and to use PL3VEC, PNTPLT coloring. Moved nowhere. 11389. August 21-29, 2003 FILLM Eric Created ZOLAVF in $APLGEN (stub) and $APLUNIX (real) to call oladvf to position the on-line tape device at start of yesterday's file, start of today's file, or current end of today's file (i.e. now). Created a dummy oladvf_ in ZONLIN.C. Changed FILLM to call ZOLAVF for ONLINE using the NFILES argument. Moved nowhere. 11390. September 2, 2003 $YPGNOT/DELZN Leonia DELZN has been modified following request of Mark Reid: 1. Editing of the input data based on high level of deviation of the data points relatively found solution. The edited points are blanked and are not used at the following edit iterations. 2. The OUTFILE has been created only if the output CL table created also. Now creating of OUTFILE and CL table is splitted. 3. The default value for OPTYPE (DELA) has been added. Moved nowhere. 11391. September 2, 2003 FIXWT Eric Overhauled the code to get different antenna-based weights for each IF and polarization, taking advantage of multiple spectral channels when present and using both real and imaginary parts of the vis. It no longer does medians for rms and does not require the data to be in exact TB order. Moved nowhere. 11392. September 3, 2003 CALIB Eric Repaired minor logic problem that caused negative ANTUSE to be handled wrongly. Moved nowhere. 11393. September 4, 2003 install.pl Eric Added to message about edit LOGIN.SH for LAPTOP YES. The HOSTS.LIST file must also be edited for LOCALHOST. Moved nowhere. 11394. September 4, 2003 XAS.SHR Eric The order of colors seems not to be fixed even for Solaris systems. This means that we need a 4-way branch on ByteSwapDisp and RedBigger to determine which colors go to which bytes in the image word. RedBigger was previously assumed to be exactly determined by the byte order of the displaying machine. Also fixed numerous unmatched quotes (in comments) that caused emasc to color the code improperly. Moved nowhere. 11395. September 5, 2003 install.pl Eric Changed it in regard to LAPTOP=YES. Added a page 5a to ask about this option and to change ThisHostNmae to LOCALHOST. If this is set, LOGIN.SH and LOGIN.CSH are not autoedited to change the LAPTOP logical. LOCALHOST will then appear in the HOSTS.LIST, DADEVS etc etc Updated the web pages for this. Moved nowhere. 11396. September 8, 2003 CLCOR Eric I forgot several calls to SOUELV when adding the AZimuth argument. Cleaned up usage of PSTD.INC. Moved nowhere. 11397. September 15, 2003 FILLM Eric Added progress messages for when data are rejected at hour intervals. The on-line fillm read is too slow and people worry too much... Moved to on-line area today. 11398. September 18, 2003 IMAGR Eric IMAGR spelled NORESTORE without the last E so it was not picked up. Moved nowhere. 11399. September 19, 2003 Y2K and others Eric Changed Y2KLOAD.001 and Y2KEXEC.001 to support a new HUGE Y2K test. It uses Cygnus A with 4 A-array frequencies and one each in B, C, and D. It uses 4-resolutions in IMAGR and does not use APCLN. Changed PRTAC to recognize the Y2KH list of tasks. Changed VTESS to load the TV every n pixels - the others like it should do the same. Moved nowhere. 11400. September 23, 2003 EXTLIST Eric Added a second version for PLOTR although the display is unchanged, the number of adverbs increased. Moved nowhere. 11401. September 25, 2003 MSORT Eric The sort problem in MSORT degrades as non-linearly as the record size, number of records, and average separation between the current record position and desired record position increase. The buffers were not fully utilized - so I changed it to use them. I also added a very brute force record read from disk for long records which works only when the input and output files are not the same. The meaning of PRTINLEV was adjusted so that 0 produces some progress info. Moved nowhere. 11402. September 25, 2003 VTESS, APCLN, et al Eric Changed all TV loads to use pixel increments to fit the picture on the TV. Added a message ahead of the load giving the pix range and increment and cleaned up the instructions and replies. Affected APCLN, SDCLN, UTESS, VTESS, APGS, APVC, STEER. FNDVAR would not do 4096 images, setting the beam area to 10 without computation. Moved nowhere. 11403. September 29, 2003 ALLTAB Eric The standard extension file copy routine did not check that the input and output were identical! Added that check. A large COMLNK ensues... Moved nowhere. 11404. September 30, 2003 $APGNOT/DFCOR Leonia The new task is a version of CLCOR with the only difference that it records the differential correction SOURCE-CALSOUR(1) into the copy of the given CL table. The only OPTYPE='ATMO' is available now. DFCOR reads the CL table twice. First time the data corresponded to the given calibrator (CALSOUR(1) are read and the relevant delays and rate of delays recalculated from the input file of zenith delays. The found delays and rate of delays are stored at the array together with the arrays of time and antennas. Second time the data corresponded to the given source list (SOURCE) are read and the relevant delays and rates of delay are calculated from the input file of zenith delays. Then the stored delays and rates of delay for the calibrator are interpolated to the time of the CL table row and the interpolated values are subtracted of the calculated source delays and rate of delays. The differences are recorded to the CL table Moved nowhere. 11405. September 30, 2003 FRING Amy FREQX and FREQY where declared as arrays of size 1024, changed this to MAXFRQ, the maximum number of frequecies. This caused problem for experiments with more than 1024 channels. Added a warning message if the specified pre-integration time (DPARM(4)) is larger that the smallest pre-int. time found in the experiment. This can cause array problems since the pre-int. time is used to calculate the maximum number of times etc.. Moved nowhere. 11406. October 1, 2003 FITLD Eric Most of the main FITS header is ignored for IDI uv data. Added code to pick up INSTRUME and put that in the output data header. Moved nowhere. 11407. October 1, 2003 ATNF suggestions Eric/Chris Phillips Changes to incorporate ATNF ideas: VKEY.INC Bring date format up to date. ZABOR2.C Solaris - add stuff to get rid of ieee warning messages UVFIT Increase XMXVIS by an order of magnitude to 250000 Adjusted the help file accordingly. GETJY Have AT data avoid a warning about FQ mis-match on sources and flux. Fix errors in which 1 was used where BIF was supposed to be. GETANT Have ATLBA be same as ATCA - they both use earth centered coords but do not reverse Y LGEOM Upgrade the help file with info on the order of shift and scale. Removed old task GEOM. MCUBE Added a simplistic description of things to the help file. BPASS Made one equivalenced buffer larger on general principles although the equivalence made the space much larger than needed. CALIB Changed Min number antennas to take into account small arrays - fixed help file too. UVLOD Raised all UVBFSS to UVBFSL buffer sizes, fixed standard Stokes test to include AT FITLD Raised all UVBFSS to UVBFSL buffer sizes, fixed standard Stokes test to include AT Moved nowhere. 11408. October 2, 2003 FITTP, FITAB Eric Added range of catalog numbers to which DOALL applies. Changed POPSDAT, NEWPARMS to create adverb CATNO. Moved nowhere. 11409. October 2, 2003 more AT Eric/Chris Phillips More changes: CLCOR Extend the VLA override to include VLBA and AT. Skip an FQ test for ATCA, ATLBA. PCCOR Add ATLBA as non-VLBI like coordinates. UVCRS Add ATLBA as non-VLBI like coordinates. ACCOR Add ATLBA as non-VLBI like coordinates. SPLIT Add ATLBA as particpating in some FQ/QUAL mixup, selecting qual SPLAT Add ATLBA as particpating in some FQ/QUAL mixup GETJY Add ATLBA as particpating in some FQ/QUAL mixup - suppressing some warning LOCIT Add ATLBA as non-VLBI like coordinates. CVEL Add ATLBA as non-VLBI like coordinates, help with user array type parameter VPLOT Add ATLBA as non-VLBI like coordinates, help with user array type parameter CLPLT Add ATLBA as non-VLBI like coordinates, help with user array type parameter UVFIX Added option to not change u,v,w. Added code to detect -NCP, warn the user, and change the header to -SIN. Moved nowhere. 11410. October 5, 2003 DBCON Eric Added coordinate tests for when DOPOS <= 0 in case coords do not agree. The test is 1 microarcsec not 3 as the message claimed. Moved nowhere. 11411. October 5, 2003 GREYS Eric Added INFILE option to overplot Holography panel and bolt layout. Moved nowhere. 11412. October 6, 2003 FITTP, FITLD Eric Fixed an egregious error I made Thursday which caused it to write out the wrong image/uv data set(s). Updated the MNJ tar ball immediately. 11413. October 7, 2003 COMLNK Pat/Eric Move test on $PREP being writable to a place where LPWD is known. Moved nowhere. 11414. October 7, 2003 APCAL Amy APCAL can now read WX tables instead of external text weather tables to do the opacity corrections. Added INVERS to the help file as the weather table version number and updated the EXPLAIN file. APCAL also now prints a message stating it is writing an SN table and gives the number. Moved nowhere. 11415. October 9, 2003 CYG Eric Put a new timing test (very large) back - called CYG with help files CYG, CYGSAVE and run files CYGLOAD, CYGEXEC, CYGLBOX. The data files are on the ftp site and amount to about 1 Gbyte. Changed PRTAC to handle this one too. Moved nowhere. 11416. October 10, 2003 BPASS Eric $APLNOT/ITBSRT did not do the test on buffer size correctly. It asked if the number of records was less than 0, not less or equal. So an obscure UVINIT message came out instead of a good message. BPASS makes the table buffer for sorting for other reasons some of which can make it way too small. Added a minimum. Moved nowhere. 11417. October 10, 2003 FITS includes Eric Changed form of includes to avoid unused variable messages. PARAMETER works very well indeed. DFUV, DFIT, VFUV, VFIT. Moved nowhere. 11418. October 10, 2003 FRING Eric Moved some code around to allow a check on REFANT before it is used for things. Otherwise, near the end a REFANT greater than the number of antennas would cause all solutions to be discarded. Moved nowhere. 11419. October 13, 2003 $QPGNOT/UVCON Leonia At March 11, 2003 (#11264) I add the test if the model dimensions are 2^n. Erroneously this test was put at the are where model could be point source. The task FHEAD finds the dimension of the model(?) image even if its NAME and CLASS are given as blank. This is strange (for me). FHEAD uses the first image at the given disk. Anyway I have fixed it testing the dimensions of the model image only if the model is not a point source and NAME/CLASS of the model image are not blank. Moved nowhere. 11420. October 14, 2003 APCAL Amy Added some warning messages for the opacity correction; i.e. that OPCODE, DOFIT, TREC and TAU0 should all be set to get good opacity corrections. Added the sec Z limit (APARM(5)) to opacity plots. Added a line in line in the help file clarifying what the opacity fitting actually calculates. Removed most references to ANTAB from the help file. Moved nowhere. 11421. October 15, 2003 FITAB Eric Empty extension files need to be written. Otherwise, some facets will end up with no CC file and CALIB will not be able to use them. Changed DKEY and VKEY includes to reduce the warning messages. Moved nowhere. 11422. October 15, 2003 CDi_j, PCi_j Eric Created PCCARD and PCHDR to parse and interpret the CDi_j and PCi_j world coordinate header cards into aips nomenclature. We may not move into the full WCS but we should read the headers written in simple ways by HST et al. Added calls to these in IMLOD, UVLOD, FITLD, PRTTP, and FWRITE (used by TPHEAD in AIPS). Dropped the inversion of the input image based on the second axis increment in IMLOD and FITLD. Masked a message in PRTTP that gave a very odd disk name. Moved nowhere. 11423. October 16, 2003 FITAB Eric The extension file writing goes from record IRNO1 to IRNO2 which may be set by a time range. If there are no data records in the ext file then they were 0 and 1 and TABIO was called. This resulted in only 1 message but may have made some garbage out. Moved to AOC load area today, nowhere else. 11424. October 16, 2003 FITLD Eric Changed the code to allow more than one disk file to be read in a single FITLD run. This is especially useful for VLBA data in which there may be more than one mode and so more than one output file. The disk file names must be postfixed with consecutive numbers 1-NCOUNT with no leading 0's. Moved nowhere. 11425. October 17, 2003 install.pl Eric Some people apparently insist on blowing away the tar ball before doing most of the installation. This made it hard to find the date of the tar ball for the MAKE.MNJ. These same people do not read and attempt to interpret the messages that come out and then are upset that the MNJ did not build. Changed install.pl to look for dates on several files and to issue a different message if no valid date was found. Moved nowhere. 11426. October 20, 2003 $APGNOT/FIT2A Leonia The new task reads one plane image at FITS format and records it to the output file at ASCII format in three columns: X,Y in pixel number; value at format E10.3. Moved nowhere. 11427. October 20, 2003 misc Eric Changed MacIntosh LIBR.DAT to ncurses from termcap Changed ending of $SYSUNIX/LIBR so that all of the data files in $PREP are kept when there is an error. Moved nowhere. 11428. October 22, 2003 DSTOK Eric New task - eliminates RL and LR or Q, U, V from a uv data set. UVCOP does not offer this choi2e. Moved nowhere. 11429. October 22, 2003 Misc Eric Changed to fix mispelling in XTRAN.HLP one-liner and to give a better one-liner and category for COODEFIN.HLP - we could not find either with apropos coordinat before. Updated LSAPROPO.HLP - removes do not get removed. Updated HLPIT.LIS - new and removes fail to get added Updated ABOUT files for ADVERB ANALYSIS AP CALIBRAT CATALOG COORDINA FITS GENERAL IMAGE IMAGE-UT IMAGING OOP PLOT POPS PROCEDUR RUN TABLE TAPE TASK TV UTILITY UV VERB VLA VLBI Moved nowhere. 11430. October 22, 2003 POSSM Eric Yet again a whole area of code did not work. In this case, the option to plot auto and cross correlations wrote over themselves and were rather confused about what to plot - but that was okay since it was wrong anyway. Now I think it is okay. Moved nowhere. 11431. October 23, 2003 Compiler bug Eric A couple of compilers have made problems with simple STOP statements in subroutines. We have character arguments to STOP many of the places where it is used so added to the others: $APLGEN FOR routines : ZC8CL ZCLC8 ZDM2DL ZI16IL ZI32IL ZI8IL ZILI16 ZILI32 ZILI8 ZIPACK ZM70M2 ZMCACL ZR32RL ZR64RL ZRLR32 ZRLR64 $APLNOT/FFTV2 Changed variable name STOP to TSTOP in $APLNOT/UVTIME. Moved nowhere. 11432. October 27, 2003 CONFI Leonia Topography file can be given now at the FITS format as well as the ASCII format. The restricted pixels correspond now to the pixel value .NE.0 removing the inversion existed at CONFI long time. Moved nowhere. 11433. October 27, 2003 Bad addressing Eric The offset to particular correlators may be confused by the switch between compressed data input and uncompressed work area data. The VOFF computed by GETCTL needs to be rescaled: (VOFF / INCS) * JNCS Changed UVCON, ALGSTB, ALGSUB, VISDFT, VSCDFT, VSFDFT Changed ALGSTB to do some of its recomputation of pointers before calling GETCTL or even the above fix will not work. Changed GRDSUB to make the choice between ALGSTB and ALGSUB only once - ALGSTB changes INCS to 3 and leaves it there so it would be possible for ALGSTB to be called for the first field and ALGSUB for the rest which is not a good idea. These fixes should affect UVSUB of compressed data only - either division of cross-hand subtraction. All else should have worked. Dropped one test from GETCTL as well. Moved nowhere. 11434. November 5, 2003 APCAL, POPSDAT, NEWPARMS Amy Changes to APCAL: 1. changed TREC parameter in APCAL to TRECVR and increased the array size from 30 to 60. There is a receiver temp. for each polarization so an array size of 30 means that opacity corrections could only be done for 15 antennas. POPSDAT and NEWPARMS changed for this. 2. Added subroutines DFTAU and DFTRC which estimate an initial TAU0 and TRECVR if they are left blank and opacity corrections using method 'OPAC' or 'GRID' are chosen. 3. Added another opacity solution method, 'LESQ'. This method used a least squares fitting routine in addition to manipulating the errors to fit a line to the lower edge of the Tsys vs sec(z) curve. This method does not need an initial guess for the TRECVR and TAU0. 4. Fixed various small bugs: a) the opacity correction would fail if a TIMERANGE was set; b) the TREC and TAU0 for an antenna were not accessed correctly if one selected ANTENNAS; c) TSPILL (spillover temperature) was not added correctly in one formula. Moved nowhere. 11435. November 5, 2003 VLBAUTIL Amy VLBAFPOL - fixed bug which automatically set OUTDISK = INDISK VLBACALA - Added ability to do opacity correction by adding parameter DOFIT VLBALOAD - fixed bug that ran TAMGR and INDXR on one file, even if FITLD produced more than one (this happens if you observe with different frequency setups). Moved nowhere. 11436. November 5, 2003 PBCALC Eric Modified PBCALC to look at telescope array name. It has values for the VLA and ATCA and also returns a logical OUTSID indicating that the position is outside the reliable region of the beam fit. PBCALC New call seq, cleaner organization. Added ATCA parameters provided by Chris Phillips. Changed call sequence using CATH(KHTEL) in character*8 PBCOR, SAD, XMOM, IMFIT, JMFIT, MOMNT, and FLATN Added calls to ANTINI to get array name in BOXES, FACES, and SETFC Added OPCODE to PATGN to do ATCA beams and set array name. Added ANAME to DTESS.INC, use it in calls to PBCALC in APLPBI and APPLPB, initialize it from image headers in LTESS, STESS, VTESS, UTESS, and IMERG. Moved nowhere. 11437. November 6, 2003 MSGSUP Eric MSGSUP used to be used excessively where ZEXIST should be used with ZOPEN, ZCREAT, ZDESTR et al. used based on the result. Changed AU2. AU2A, AU3A, SGLAST, STORES in $AIPPGM plus ISTAB in $APLSUB and OTBSRT in $APLNOT. MSGSUP prevents all conditions from being seen, when the normal condition (file exists or does not exist) is all that one wants to mask. Moved nowhere. 11438. November 10, 2003 UVCON Leonia/Eric The standard input files for four VLA configurations and VLBA configuration have been added. The five files are under $AIPSTARS and should be called as: INFILE 'AIPSTARS:VLA-A_UVCON' INFILE 'AIPSTARS:VLA-B_UVCON' INFILE 'AIPSTARS:VLA-C_UVCON' INFILE 'AIPSTARS:VLA-D_UVCON' INFILE 'AIPSTARS:VLBA_UVCON' If a user want to change the data of VLA, VLBA configuration, he/she can copy the relevant file(s) to his area, edit it and create his own version. The comment cards can be added now to the input file. Moved nowhere. 11439. November 11, 2003 GETCTL, DROUND Eric Created a DROUND to round double precision to integer. GETCTL got into trouble saying K = CATD(KDCRV+3) which was probably 1 bit less than 1.0 on a Mac G5. This has worked okay for a long long time but should really be K = DROUND (CATD(KDCRV+3)). Moved from Kevin's G5 today. 11440. November 11, 2003 READLINE.SHR Eric Fixed READLINE.SHR so that an sh command will unpack it - also RLSUPP.SHR, RLEXAMP.SHR, RLDOCS.SHR. Like XAS.SHR this is simple: the cat line has to have quotes around the magic string and the end line does not have quotes e.g. cat > colors.c << "--XYZZY--" ..... --XYZZY-- Moved nowhere. 11441. November 12, 2003 Mac procs Eric Added the IBM compiler suite as an option to $SYSMAC files FDEFAULT.SH, LDOPTS.SH, and CCOPTS.SH. Moved nowhere. 11442. November 12, 2003 FRING Amy FRING calculated the number of frequencies by taking the band- width and dividing by the smallest channel width. If channel widths were less than 0, then the number of frequencies was negative and this caused FRING to die. Added an absolute value number of freqs calculation. Moved nowhere. 11433. November 12, 2003 install.pl Eric Changed: 1. Replace building UNSHR and then using it with a simple sh command for XAS, READLINE, etc. 2. In compiling PP, changed if IBM test to if compiler is xlf to get the -qextname option in. Moved nowhere. 11444. November 14, 2003 FITLD, PRTTP Eric FITLD, PRTTP, IMLOD, UVLOD all had rather small limits for the number of FITS header records that they would read before deciding they were in trouble. Raised the limits by a factor of 100. Moved nowhere. 11445. November 16, 2003 TABHDR, FITLD Eric FITLD has been neglecting the fact that the weather table has source and antenna number in it. Since the antenna number can change for a particular antenna (and source number for source) this means that the data in the WX table produced for VLBA data is likely to be wrong. Changed TABHDR to turn a WEATHER table to a WR aips table and added WR2WX in FITLD. Moved nowhere. 11446. November 17, 2003 install.pl Eric When it has to unpack the READLINE.SHR there must be a link file to unpack. Added the code back again to make the link file. Moved nowhere. 11447. November 17, 2003 Mac teachings Eric Changed $APLNOT/CALREF to avoid calling SLAMCH. This routine goes into infinite loops with any reorganization which the Mac linker apparently is doing. Changed LIBR.DAT in $SYSMAC, SYSLINUX, SYSALPLHA, SYSAXLIN, SYSDEC, SYSHP, SYSHP2, SYSIBM, SYSPPLIN, SYSSGI since $AIPNOT does not need AIPSUB subroutines and AIPSUB subroutines call things in Y areas. Global optimization can cause trouble with this. Moved nowhere. 11448. November 18, 2003 WRTPROCS, READISK Amy Fixed READISK so that it worked with new FITLD parameter, NCOUNT. Moved nowhere. 11449. November 18, 2003 WX table Eric The VLBA table lacks the array number it should have. Changed: DWXV.INC Added additional column WXINI Added additional column TABWX Handle new column incl its absence. TABLEUTIL Change call seq to OTABWX and TABWX. FITLD Append old TABWX and DWXV as WR things. The subarray will be 1 in the FITLD table. FILLM Always writes subarray 1 to all of the files. APCAL Include only those rows matching the subarray. Increase number of adverb values for Amy's fix. WETHR Added SUBARRAY as adverb to limit if desired to one subarray. Check SUBARRAY column value (if present) to match. INDXR Pass through subarray and limit to the correct subarray. Moved nowhere. 11450. November 20, 2003 EXTLIST and APCAL Eric Worked over APCAL so that info for EXTLIST would be present. Changed AU8A to say something about APCAL. It can even do the parameters of the first 1 or 2 plots on a page. Changed the plot file: PLRANG Corrected bad error in logic, changed call seq to allow a LEAST difference between max and min PLAXES Allow for axis labeling in DD/HH MM form if axis label is 'TIME (HOURS)' PLCLOS Return an error code so that button D pushes will actually stop the plotting APCAL Made it support page full on DOTV, change right hand plot to be apparent zenith opacity versus time, add GRCHAN as adverb supporting 0. OBPLT Changed call sequences. Moved nowhere. 11451. November 20, 2003 IMLOD FITLD Eric In IMLOD there was a limit of 250 extensions of one type on a file - the same limit in FITLD was 1200+ (2 digit base 36). Changed both to 46000 (3 digit base 36). Moved nowhere. 11452. November 20, 2003 IMPOS Eric Changed help file and AU5 to return COORDINA from IMPOS. Moved nowhere. 11453. November 24, 2003 APCAL Amy, Eric Made another option for DOFIT (-1), so antennas can be left out of the opacity correction. Also, if ANTENNAS is set then DOFIT applies to only those antennas, i.e., DOFIT(1) applies to antenna in ANTENNAS(1). This is also true for TRECVR and TAU0. Fixed problem with plotting if gain goes negative. Moved nowhere. 11454. November 30, 2003 SETPAR Amy Increased AIPSmark maximum to 200. Moved nowhere. 11455. December 1, 2003 CookBook Eric Changed $DOCTXT/LaCOOK?.tex and $AIPSPUBL/COOK?.PS: 1 - add Mac OS/X, revise bits and pieces 2 - change start up messages, screen server description 3 - OUTPUTS, DEFAULT, disk files in FITLD, FITAB, FILLM 4 - VLARUN mentioned, FILLM from disk added to, CLCAL seriously changed. 5 - FITLD multiple disk files 6 - FITS disk reading, gamma corrections in LWPLA, line coloring, task pause at page full, UVPLT option, POSSM plots BP and correlation funcs, polarization vectors in color, contours in color, TVCOLORS, output adverbs, TVILINE, 2 more color plot pages 7 - PBCOR was way out of date, IMDIST, IMEAN circular aperture 9 - FITLD disks, CLCOR/DELZN, SNSMO, CLCAL, APCAL changes, GPHAS dropped 13 - update Also Index and Table of Contents. Created PDF files for all chapters in $AIPSPUBL. Rearranged Recipe.tex for changed usage. Moved nowhere. ====================================================================== ====================================================================== Changes while 31DEC03 is NEW ====================================================================== ====================================================================== 11457. December 2, 2003 FILLM Eric Fixed time advances while skipping tests to handle day roll over. There was a problem only when no data had yet been selected. Moved to 31DEC03 this date. 11459. December 3, 2003 MAKE.MNJ Eric The part in do_daily that defines AIPS_ROOT should come ahead of the line that uses it. Moved to 31DEC03 this date. 11460. December 4, 2003 install.pl Eric Grammar error caused special hack for HPs not to work. Corrected $arch to $ARCH. Moved to 31DEC03 this date. 11461. December 8, 2003 SNSMO Eric There was no default for SMOTYPE despite what the help said. The clipping routines had a variety of troubles. Put in counters for what was deleted in CLPAPH, CLPRAT, CLPDEL. The amp/phase clip assumed that both polarizations needed to be present and so showed both bad if one was. This caused odd behaviors and failures to clip. Dropped this both pol nonsense. Fixed the setting of the BAD flag so that fewer records would be updated - only those actually changed. Moved to 31DEC03 this date. 11464. December 9, 2003 UVFIX Eric Two major errors in processing multi-subarray data: 1. The counter used to show start of data only worked for the subarray of the first sample. All others used parameters based on that one. 2. When bug 1 was fixed, several parameters were found to be set to the output values by the first subarray. This included EPOCH and the coordinates and so made handling the second one problematical. Made a bunch of parameters be picked up and saved at the beginning and then reused when needed. Moved to 31DEC03. Cannot patch 31DEC02 since UVFIX differs in other ways too. 11466. December 11, 2003 BLOAT Eric The alternate reference pixel was not reset when the frequency reference pixel was. Moved to 31DEC03 this date. 11467. December 11, 2003 VPFLG Eric Add option to flag all IFs if 1 is bad. Moved to 31DEC03 this date. 11468. December 11, 2003 UVFLG Eric Expand error message when trying to flag ampl or weight on a multi-source (FG table) data set. Moved to 31DEC03 this date. 11470. December 16, 2003 PLOTR Eric Corrected symbol size by multiply or divide by SQRT(XYRATIO) to keep the symbols square when XYRATIO is not 1. Moved to 31DEC03 this date. 11471. December 18, 2003 CookBook Eric Corrected DELZN remarks in LaCook9.tex and COOK9.PS and COOK9.PDF. Moved to 31DEC03 this date. 11472. December 22, 2003 UVPRT Eric Changed displayed remark: u, v, w are scaled to selected frequency, not the reference frequency. Moved to 31DEC03 this date. 11473. December 23, 2003 ALGSUB, ALGSTB Eric Added field number to messages to show progress when more than one field is involved. Moved to 31DEC03 this date. 11474. December 31, 2003 ELINT Leonia ELINT has forced the imag part of the output CL table amplitude to zero. As a result ELINT has given the right output CL table only if the input CL table has zero for imag part of amplitudes. It fixed. Moved from 31DEC04 this date by Eric. 11475. January 1, 2004 install.pl Eric Changed 31DEC03 version of install.pl to myver NEW (stops the MNJ). Found that it had called itself 31DEC04 since 4 Dec! Moved nowhere. 11476. January ??, 2004 CLCAL et al. Eric Changed CLSMO Fix spelling of comment SNSMO I swear I did this once already: dropped an initial insistence on having both polarizations at each time. This got in bad trouble in some cases. CLCAL The smoothing of real and imaginary is not ideal and was changed back to smoothing phases keeping track of the last good phase to try to avoid wrap troubles. Moved from 31DEC04 - thawed it just for this!