• Download HEASOFT
  • ******************************************************************************
                            RELEASE NOTES FOR HEASOFT 6.4a
                                 December 12, 2007
    The "6.4a" patch release fixes problems in the 'mgtime' task (all platforms)
    and in the fv GUI (Mac OS X & Solaris only).
    A change to 'mgtime' in HEASOFT 6.4 improved the precision of the output
    MJDREF keyword.  However, a consequence of the change also meant that
    input MJDREF keywords with full double precision notation are interpreted
    incorrectly (e.g. input MJDREF=5.19100007432345E+04 is interpreted as
    MJDREF=5.1910000743).  The 6.4a patch fixes this read error in mgtime.
    fv (fitsTcl):
    On Mac OS X & Solaris, WCS information was not being constructed
    properly when plotting data with WCS projection.
                            RELEASE NOTES FOR HEASOFT 6.4
                                  December 6, 2007
    The HEASoft 6.4 release is primarily driven by the release of the Suzaku
    (version 7) and Swift (version 2.8) software, but also includes updates
    to several libraries and packages which have been revised to mirror the
    current development versions in use at the HEASARC.
    As before, configuration-related changes thoughout the package means that
    users who already have HEASoft v6.x.x installed will need to completely
    replace their software installation; there is no clean way to simply
    overlay this new release or to update an existing installation.
    - Improved support for Mac/Darwin OS 10.5.x (Leopard)
    - Compilation of HEASOFT's copy of the readline library (v5.2) is now
      optional.  If --enable-readline is not given to the top-level
      BUILD_DIR/configure, it will check for readline among installed system
      libraries and use it (if found). 
                          APE (Parameter Interface Library)
    Release Notes for Ape Version 2.3
    Changes since version 2.2
    The following functional and behavioral changes are new in this version:
    o If the environment variable APEDEBUG is set, debugging messages
      will be enabled throughout the Ape library at run-time.
    o The plist and pquery2 utilities now assume queried and learned
      behavior ("ql") for automatic parameters if there is no mode
      parameter in the file. Previous versions silently returned an
      error code. (Note that an "automatic parameter" is one whose
      third field, the mode field, is "a". The "mode parameter" is
      a parameter named mode, whose value specifies the behavior for
      automatic parameters, usually "ql".)
    o The pcheck utility now reports an error if the parameter file
      contains one or more automatic parameters but there is no mode
      parameter. Previous versions did not detect this error. (See
      also change to plist and pquery2 described above.)
    o Ape now handles correctly all file types associated with the
      SAOHost interface, including "fe" (file exists), "fn" (file does
      not exist), "fr" (file is readable), "fw" (file is writable), and
      "f" (no file check performed). These modes may also be combined,
      e.g. "frw" (file is readable and writable), but it is therefore
      possible to create modes that will never be accessible, such
      as "fen" (file exists and file does not exist).
    o When Ape is built without readline (configure --without-readline),
      prompts are now redirected to the user's terminal (/dev/tty) on
      Unix systems. This change made Ape's behavior consistent
      regardless of whether it is built with or without readline.
    o When built with Windows/Visual Studio, Ape's file access checking
      code is now consistent with the Unix version.
    o The ape_test unit test binary now takes an optional argument
      giving the name of the log file into which to redirect all
      output. This makes it easier to run the unit test from the
      command line while debugging.
    The following code changes are new in this version:
    o A new API function, ape_util_get_file_check_func, returns a pointer
      to the current file access checking function. This complements
      the extant function ape_util_set_file_check, and allows client
      code greater flexibility when implementing custom file access
    o A new API function, ape_util_interpret_env, reads all necessary
      environment variables and sets the internal state of Ape
      correctly based on their values.
    o A new API function, ape_par_redirect_prompt_stream, allows
      prompts to be redirected to the given C FILE* stream.
    o The functions ape_par_check and ape_io_check_file_format, which
      check the validity of indivual parameters and parameter files,
      respectively, were rationalized to perform more precise error
      checking, and to provide more explicit error messages.
    o The functions PILInit and ape_trad_init, which set up Ape
      for standard parameter access, no longer call the strict file
      format checking function ape_io_check_file_format. Serious errors
      will still cause these functions to return an error code, but
      without the extensive output given by ape_io_check_file_format,
      which is not always desired simply while opening a parameter file.
      This change also allows client code to document and recover from
      errors by leaving the parameter file as open as possible. However,
      it is now especially important to call ape_trad_close in all cases,
      whether or not an error occurred in ape_trad_init.
    - imagetrans:
      Implemented another transform method=FLAT.
      Initialize IMAGE weight member to null.
    - rbnrmf: Updated to fix bugs in the energy binning method: the result for
      energy-binning should be averages, not sums, and there was an off-by-one
      error for zero-based channels.
    - callib (evexp routine): Convert all cbdvals to uppercase to fix problem
      that occurs when the calibration boundaries are given in lower case in
      the caldb files.
    New version 1.8:
    - Fixes made to bugs in Column write and writeArrays functions which were
      preventing the writing of variable-width columns.  Also now allows writing
      to fixed-width columns with arrays that are shorter than the fixed width.
    - The HDU::readAllKeys() function will no longer throw if it is unable
      to read a particular keyword.  Instead it will skip it and move to the
      next keyword.  This was done primarily to prevent it from tripping on
      undefined keywords.
    Additional changes beyond version 3.07:
    - Fixed bug with PLIO compression routine, and possible memory overflow.
    Version 3.07 - October 2007  (GSFC internal release)
       - fixed a possible, but unlikely, memory overflow issue in iraffits.c.
       - added a clarification to the cfortran.doc file that cfortran.h
         may be used and distributed under the terms of the GNU Library
         General Public License.
       - fixed bug in the fits_modify_vector_len routine when modifying
         the vector length of a 'X' bit column.
    Version 3.06 - 27 August 2007
       - modified the imcopy.c utility program (to tile-compress images)
         so that it writes the default EXTNAME = 'COMPRESSED_IMAGE'
         keyword in the compressed images, to preserve the behavior of
         earlier versions of imcopy.
       - modified the angsep function in the FITS calculator (in eval.y)
         to use haversines, instead of the 'law of cosines', to provide
         more precision at small angles (< 0.1 arcsec).
    Version 3.05 -  July 2007 (GSFC internal release only)
       - extensive changes to imcompress.c to fully support implicit data
         type conversion when reading and writing arrays of data to FITS
         images, where the data type of the array is not the same as the
         data type of the FITS image.  This includes support for null pixels,
         and data scaling via the BSCALE and BZERO keywords.
       - rewrote the fits_read_tbl_coord routine in wcssub.c, that gets the
         standard set of WCS keywords appropriate to a pair of columns in a
         table, to better support the full set of officially approved WCS keywords. 
       - made significant changes to histo.c, which creates an image by binning
         columns of a table, to better translate the WCS keywords in the table
         header into the WCS keywords that are appropriate for an image HDU.
       - modified imcompress.c so that when pixels are written to a
         tile-compressed image, the appropriate BSCALE and BZERO values of
         that image are applied.  This fixes a bug in which writing to
         an unsigned integer datatype image (with BZERO = 32768) was not
         done correctly.
    - fcreate: Increased the maximum supported line length of the input data
      file to 30000 characters.
    - fmerge: For integer data type, use ftgcfd/ftpcld to read/write columns
      instead of ftgcfe/ftpcle (and use 'dvalues' array instead of 'evalues')
      since 32-bit floating point variables cannot store the full precision of
      a 32-bit integer.
                                  FV / POW / FITSTCL
    FV 5.1 (December 2007):
     * Passing Fits Header string and cnt to client to display wcs info
       correctly for Hera and POWplot operation.
     * Fix zoom in and out bugs on POW image.
     * Support displaying images that straddle the 24h = 00h boundary on the
       sky with WCSLIB.
     * Horizontal scrolling bug of 0 rows table.
     * Release memory when deleting/closing an image in POW (on-going).
     * Fix editing graph parameter values bug.
     * Fix "Duplicate Graph" casuing seg-fault bug.
     * Fix two identical file names from two directories overwrite each other
       in POW bug.
     * Fix image flipping bug with FITS keyword CDn_n.
     * Fix displaying linear WCS projection and coordinates image.
     * Add 2 new icons: Hera and Student Hera during installation for
       Microsoft Windows and Apple Mac OS X.
    FV 5.0 (September 2007):
     * Overhauling POW world coordination calculation logics. Now fV/POW use
       standard WCSLIB library to generate world coordinates/pixel positions.
     * Allow generating of plots that retains its world coordinate projection
     * Allow limited editing capability on the property of graph objects
       (projection, rotation, etc) of a graph/image on the fly.
     * Allow switching between multiply world coordination system (i.e. WCSa,
       WCSx, etc.) on the fly.
     * Allow the flipping of image/plot of direction on X or Y or Both axes.
     * Allow display of 64 bits integers in fv data table and in POW imagery.
     * Fix white-wash grey-out POW imagery on Scientific Linux platform bug.
     * Fix saving 64 bits integer data to a text file bug.
     * "Run Ftools..." option is no long available. Users are encouraged to
       use the new "fgui" task that is distributed with ftools.
     * "Skyview..." catalog selection is now only available via "Catalog..." option.
    - extractor: Improved checking of valid filter expressions attached to
      filenames. Previously, was silently ignoring an invalid expression -
      now generates an error message.
    - grppha / rbnpha: Use new updpha routine in order to copy any additional
      columns (other than those which updpha writes) to the output file.
    - rebingausslc - rebin an irregularly sampled light curve (such as BAT or ASM)
    - swco.dat v030, valid through: 2007-11-28T23:50:32
    - ftcreate: increased the maximum supported line length of the input data
      file to 30000 characters.
    - Added a small amount of leeway to bounding box creation since in a few
      cases pixels of text are being chopped off.
    - mgtime: Increased precision of output MJDREF keyword
    - Update suggested by M. Calabretta to address memory violation that
      appeared during Swift testing on Darwin 9 (Leopard).  (Will be propagated
      into WCSLIB 4.3 release)
    - Increased arbitrary limit on number of WCSs that can be handled
      simultaneously from 20 to 60.
    - xselect.mdb:
      Corrected bad channels for Suzaku XIS - should have started at 0, not 1.
    - Switched way of setting working directory pathname to a method that does
      not resolve soft links. This is useful on Macs where directory names may
      contain spaces and soft links can be used to get round the problems these
    - Increased sizes of arrays.
                                       XSPEC 11
    XSPEC v11.3.2ag:
    - rgsrmfsmooth: new ftool to modify the XMM RGS RMF for extended sources.
    - Corrected bug in xsgabs function: previous version depended on the bin size.
    - Fixed bug introduced when removing BLOCK DATA from cpmod routine. This
      caused a bizarre error when adding a model component which was not the
      final component - the resulting final component was renamed CPCMOD.
                                       XSPEC 12
    Version 12.4.0 incorporates all of the fixes and enhancements that were
    originally distributed as patches (a-ap) for v12.3.1 (see the XSPEC
    final issues/patches page for 12.3.1 for details).  It also contains
    the following improvements:
    Monte Carlo Markov Chain
    The MCMC code run by the chain command has been greatly expanded with
    many new options.
       - The default output format is FITS (with an option to use the old
         ASCII text file).
       - Proposal distributions can now be defined in a variety of ways
         and new proposal classes can be loaded dynamically.
       - A temperature option enables annealed chains.
       - The error, flux error, lum error, and eqwidth error commands all
         use information from chains if they are loaded. This should make
         these commands more reliable.
       - New tclout options chain and margin provide access to chain
       - plot mcmc has been renamed plot chain.
       - The tclout covar option previously output the values in order of the
         upper half of the covariance matrix. This has been changed to the
         lower half matrix order so that it may be directly input into the new
         "chain proposal matrix" command.
       - The chain stat <par> option now assumes <par> to be an XSPEC parameter
         index, of the usual form [<model name>:]<par idx>. Previously it was
         treating this as a number indicating the column position in the chain
       - The xset delta <val> command sets parameter fit delta values
         proportional to the parameter values. This should eliminate the
         problem that the fit delta can end up too large relative to the
         parameter value leading to convergence problems.
       - The method command has new options minim, simplex, and monte from
         the Minuit lbrary. A new command improve can be used with the
         Minuit methods to test for local minima.
       - The projct mix model can now run on a subset of the loaded
         spectra. Previously it expected all loaded spectra to have
         XFLT keywords even if they were not intended to take part
         in the mixing.
       - Very large table models are no longer loaded in their entirety into
         RAM, but instead accessed by deferred read from disk.
       - There is a new (experimental) cyclical designation for those
         parameters which are periodic (for instance inclination angles)
       - The XSPEC v11 rgsxsrc model (which was missing in XSPEC 12), is now
         implemented with the new ftool rgsrmfsmooth.  This applies the
         correction to the XMM RGS response for an extended source, removing
         the need to have the model inside XSPEC.
       - XSPEC recognises the home directory path '~/' notation when
         running scripts (using the '@' method),
       - Initpackage can take relative paths for its 3rd argument.  Also,
         initpackage will append the prefix "lpack_" to the C++ library init
         file it produces. This should eliminate name conflicts with local
         model files.
       - Fakeit output has channel columns with type J rather than I.
       - Doing "xset <var>" with no value argument now removes <var> from the
         string database rather than leaving it in with an empty value.
    - Plot sensitivity/insensitivity was crashing when used on spectra with
      ignored channels.
    - Fix made to table model's out-of-range and exactness testing for
      cases of negative value parameters.
    - Editmod was not recognizing blank spaces as separators between
      multiple values entered for a parameter in a script.  It was only
      allowing commas to be separators.
    - Some cases of editmod on multiple-nested models were being prevented
      due to incorrectly identifying the operation as attempting to change
      more than one component.
    - In the data command, entering the invalid 0:0 for
      <data group>:<spectrum number> was causing a segmentation fault.
    - Specific only to Mac Leopard with fink-installed gcc-4.2.1: the
      data group number was improperly read in from the data command's
      <m>:<n> syntax.
    for information about Xspec patches beyond v12.4.0.
    - Fix for 64-bit bug.
                               MISSION-SPECIFIC SOFTWARE:
    - pcarmf: New version submitted by Nikolai Shaposhnikov:
      10-29-07 hardcoded escape_width to be 1.5 keV (removed escape_width
               from par file)
      10-29-07 V 11.5 removed L escape line.  EscNormL1-2(layers) are no more!
      08-15-07 Just Ka and Kb escape line are left
      08-15-07 Found trend in xenon layer 1 consistent with leak to the propane.
               Adding the term to xe_gmcm2_l1
      08-07-07 V 11.2 added L escape and Kb escape line. All escape lines are
               individually normalized with five new parameters EscNormL1-2(layers)
               and EscNormKb
      05-18-07 V11.1 Modified by N. Shaposhnikov for the escape_width parameter
               to be read from parameter file
      05-15-07 v11.0 Modified by K. Jahoda to account for non-vetoed K-escape
               events through escape_width parameter.
    - xpcaarf: New version with updated geometric areas, delivered by
      Nikolai Shaposhnikov.
    Suzaku software Version 7 - Changes since HEASoft v6.3.2:
     HXD specific tools:
     hxdgrade :
      - Fix a bug when an input file has no NEVENTS keyword in the
        GTI extension.
     hxdpi :
      - Fixed a critical bug in PI-value caliculationg part.
        This affects the gain scale of the GSO data at the
     hxdgtigen :
       -  Added a parameter "fifo_full=yes"
     XIS specific tools:
    xisnxbgen [NEW TOOL]:
       - Claculate  XIS non X-ray background generator 
    xis5x5to3x3 [NEW TOOL]:
       - Convert  the XIS 5x5 mode event file (EDITMODE='5x5')
        into the 3x3 mode event file (EDITMODE='3x3').
    xissimarfgen :
       - bug fix  when executed with "source_mode=DETFITS" option
         (reported by Keith)
    aemkpinudhk [NEW TOOL]:
       - Calculate PINUD count used for non X-ray backgroung estimation 
         from HXD HK files
    Swift software Version 2.8(Bld22) - Changes since HEASoft v6.3.2:
    General tools 
     - Add some more informative diagnostic messages for high chatter
    swiftround [NEW TOOL]:
     - Rounds pixel values in image HDUs to reduce disk size of compressed
       UVOT sky images
     - New hidden parameter round=<real> to request rounding be performed
       after coordinate transformation (rounding is disabled by default).
     - Added parameters to constrain the separation between sources used
       for matching and to filter matches.  
     - Store residual rotation quaternion.  
     - Transfer input source table columns to output.
    BAT  specific tools
    batimgstatpos [NEW TOOL]:
     - This task computes BAT-detected source positions, based on BAT
       housekeeping "image status" data.
     - Fix text of an error message
     - Fix text of an error message
     - Modifications to make the task more robust regarding input values
     - Handle remote CALDB properly  (fixed in v6.3.2)
     - If a RESPFILE column exists, then use it instead of the keyword
       (useful for Type II spectra which could have different responses)
     - Fix a bug which results in a crash when users specify aperture=CALDB:xxxx
     - Handle remote CALDB properly  (fixed in v6.3.2)
     - Add the 'shortfix' parameter to work around a problem with short
       burst trigger time intervals.
     - The task estimates the duration from several differently binned
       light curves; it now also checks 16 msec bins.
     - The default aperture is now CALDB:FLUX for best flux measurements;
     - Updates to adjust to library changes.
     - Fix bug when there are no partial coding constraints but
       theta/IMX/IMY constraints are applied.
     - Handle remote CALDB properly  (fixed in v6.3.2)
     - Handle remote CALDB properly  (fixed in v6.3.2)
     - Major changes to improve the how the burst durations and
       uncertainties are measured.  A new parameter 'durerrmeth'
       controls how uncertainties are estimated.
     - Handle deletion of keyword properly when a column is created
     - Other code clean-ups
     UVOT specific tools:
    uvotdetect :
     - Propagate columns added to the SExtractor .param file to the output
     - Implemented support for calculating exposure maps based on aspect
       following packets accessed when method=SHIFTADD.  
     - Since event mode data does not have aspect following packets, 
       artificial ones are generated based on attfile.
     - Implemented support for generating raw images from event mode data
       where the raw event positions are corrected for attitude drift.
     - Output raw image files (...ff_rw.img) now contain both image and
       event mode HDUs.  
     - If 'ra' parameter is negative, take RA, Dec and roll
       from corresponding -_PNT keywords
     - The badpixfile parameter now allows the special value NONE
       which causes an artificial quality map to be generated with no
       bad pixels.
     - Added the weightfile parameter which can be used to specify multipliers
       for each input HDU.
    uvotlss [NEW TOOL]:
     - Calculate UVOT large scale sensitivity correction
     - Renamed the method parameter to apercorr with values NONE|CURVEOFGROWTH.
     - Allow user to control output plot type by given specifying plotfile
       parameter with extensions matching PGPLOT devices.
     - Now generates magnitude history table for all exposures; 
       multi-filter light curves; finding chart with several positions (ds9 v4+
       required); summary report with key processing parameters and results.
     - Only use sources with count rates < 50/s in attempt to find aspect solution.  
     - Store residual rotation in output
     - Renamed the method parameter to apercorr with values NONE|CURVEOFGROWTH.
     - Added output columns RA, DEC, FILTER, DETX, DETY, LSS_FACTOR, LSS_RATE,
     - Set MET column to middle of exposure and TIMEPIXR to 0.5
     XRT specific tools:
     - New input parameter 'enfile' corresponding to the engineering 
       spacecraft housekeeping file. From this file new information
       are placed as columns in the makefilter file (MKF). The new columns
       in the MKF are 
             extracted from the engineering file 
       -  'STLOCKFL' and 'STLOCKST' calculated from the engineering information
     - New input parameter 'enfile'.
     - Added GTI selection based on spacecraft parameters from new extension 
       'SAC' in hkrangefile.
     - Bug fixed in xrtexpocorr task call.
     - Apply the caldb query for position file depending on the observation
       start time.
     - Bug fixed in 'TIMEDEL' keyword updating.
     - Added check on exposure map file that cannot be input 
       if 'psfflag' parameter is set to "no".
    XRT libraries:
     - Bug fixed checking RA and DEC input parameters.

    Help lines: FTOOLS or xanprob@athena.gsfc.nasa.gov

    Web page maintained by Bryan K. Irby

    HEASARC Home | Observatories | Archive | Calibration | Software | Tools | Students/Teachers/Public

    Last modified: Thursday, 20-Nov-2008 15:07:58 EST