ChangeLog for rmfgen ================= Outstanding SPRs: Version 1.53.5 - 2005-08-05 (RDS) ------------ Documentation update Version 1.53.4 - 2005-08-04 (RDS) ------------ + (config/rmfgen.par) - Changed detxbins, detybins defaults from 5 to 20. - Updated documentation and added in changes from 1.54 Version 1.54 - 2005-08-03 (RDS) ------------ + (config/rmfgen.par) - Changed default for withbadpixcorr to no to avoid a spurious warning Version 1.53.3 - 2005-07-19 (RDS) ------------ + (config/rmfgen.par) - Changed default for detmaptype parameter to 'psf' Version 1.53.2 - 2005-07-05 (RDS) ------------ + (config/rmfgen.par) - Added new parameter ignoreoutoffov Version 1.53.1 - 2005-06-17 (RDS) ------------ + (src/Rmfgen.cc,rmfgen,cc) - Removed symbol definition for arfDataServer Version 1.53 - 2005-05-30 (RDS) ------------ + (src/Detmap.cc/h, - Produces an RMF which is spatially src/Rmfgen.cc/h, averaged. This uses the filtered src/CalMOSResponse.cc/h, detector map concept of arfgen src/CalPNResponse.cc/h, and uses ConcreteArfDataServer directly src/CalResponse.cc/h, (SPR-2515) test/mosspectrum.ds) Version 1.52.7 - 2005-02-25 (RDS) ------------ + (src/CorePosData.cc/h) - Uses the DatafileAttributes object (src/Rmfgen.cc/h) from arfgen to simplify code and solve SPR-2601. Version 1.52.6 - 2005-01-14 (RDS) ------------ + (src/CorePosData.cc) - Added a fix for source areas where the pnline number can appear to be greater than 199 (SPR-2591) Version 1.52.5 - 2004-12-21 (RDS) ------------ + (DEPEND) - Added selectlib,attcalc,oal,emsaplib, datautils,rgslib to DEPEND (SPR-2572) Version 1.52.4 - 2004-10-03 (RDS) ------------ + (src/CorePosData.cc) - Added support for image masks (cf SPR-2090) Version 1.52.3 - 2004-08-03 (RDS) ------------ + (src/ResponseRow.cc) - Added extra code to avoid MAC out of bounds problem. Version 1.52.2 - 2004-01-20 (RDS) ------------ + (test/runrmfgen) - change in fverify output (SPR-2492) Version 1.52.1 - 2004-01-05 (RDS) ------------ + (src/Rmfgen.cc) - Added TMLIN, TLMAX keywords to EBOUNDS (SPR-2414) Version 1.52 - 2003-11-12 (RDS) ------------ + (src/Makefile) - reordered USELIBS for static build Version 1.51 - 2003-07-10 (RDS) ------------ + (src/ResponseRow.h) - Small changes for devtrack Version 1.48.7 - 2003-07-07 (RDS) ------------ + (src/CorePosData.cc) Changes for PositionUtils change Version 1.48.6 - 2003-04-25 (RDS) ------------ + (src/CorePosData.cc/h) Changes for gcc 3.2 compatibility Version 1.48.5 - 2003-01-02 (RDS) ------------ + (src/Rmfgen.cc) Opens and closes the RMF before writing the Ebounds extension to fix a problem which occurs in low memory mode. Version 1.48.4 - 2002-12-20 (RDS) ------------ included a 'var' format run in the test harness Version 1.48.3 - 2002-12-18 (RDS) ------------ documentation update Version 1.48.2 - 2002-12-10 (RDS) ------------ + (src/CorePosData.cc) A fix for MOS timing mode to use single+double pixels to form the RMF Version 1.48.1 - 2002-12-05 (RDS) ------------ + (src/ResponseRow.cc) Zeroes the N_CHAN and F_CHAN values (SPR-2363) Version 1.48 - 2002-11-20 (RDS) ------------ + (src/Rmfgen.cc, CorePos.cc) Sets artifical DETX/Y region for Timing mode (SPR-2348) Version 1.47 - 2002-09-13 (RDS) ------------ + (src/Rmfgen/cc) Addition of OGIP standard HDUVERS keywords (SPR-2336) Version 1.46 - 2002-08-19 (UL) ------------ + a few transparent changes to improve standard C++ compliance Version 1.45 2002-05-15 (RDS) ------------ + (config/rmfgen.par) Changed to new parameter file format. Version 1.44.6 2002-05-15 (RDS) ------------ + (src/CorePosData.cc) Fixed problem with source centre falling in a chip gap (SPR-2298) Version 1.44.5 2002-04-11 (RDS) ------------ Documentation update to explain which modes are supported in this release. Version 1.44.4 2002-03-14 (RDS) ------------ + (config/rmfgen.par/lyt, test/runrmfgen) pnline parameter removed (SPR-2281) Version 1.44.3 2002-02-26 (RDS) ------------ + (src/ConcreteRmf.cc, Implemented variable length rows (SPR-1809) src/ResponseRow.cc, and set as default. config/rmfgen.par) (src/CalPnResponse.cc/h, ConcreteRmfServer.cc) - use rmf->at(chipx, chipy) (src/Rmfgen.cc) Added an error message if too many channels test/mosspectrum.ds) Reduced the number of channels Version 1.44.2 2002-02-08 (RDS) ------------ + (config/rmfgen.par) Changed q -> h (optional) for rmfset. (src/CalResponse.cc) Removed some debug Version 1.44.1 2002-01-28 (RDS) ------------ + (config/rmfgen.par) Changed a -> q. (src/CorePosData.h) - make datasubspace non-constant Version 1.44 2001-12-20 (RDS) ------------ + (DEPEND) Now links with the arfgen library (src/Makefile) (src/Rmfgen.cc, src/CorePosData.cc/h, src/CalPNResponse.cc, src/CalMOSResponse.cc/h, src/ConcreteRmfDataServer.cc) (SPR-2043, get the PN line number from the spectrum) + (src/Rmfgen.cc, src/CorePosData.cc/h) Pass the eventGrades to the Cal (SCR-72) ConcreteRmfDataServer.cc/h, src/CalPNResponse.cc, src/CalMOSResponse.cc/h, CalResponse.cc/h) (src/CalResponse.cc, DummyResponse.cc, ResponseRow.cc, RmfTableIterator.cc) (SPR-2047 - remove asserts) Version 1.43.4 2001-05-21 (RDS) ------------ + (doc/rmfgen_description.doc) - documentation update Version 1.43.3 2001-05-09 (RDS) ------------ + (src/CalResponse.cc) - set elements less then 30 eV to zero (SPR-2024) Version 1.43.2 2001-05-03 (RDS) ------------ + (src/CalResponse.cc) speed-up by calc of matrix once (SPR-2019) Version 1.43.1 2001-04-23 (RDS) ------------ + (src/CalResponse.cc) Get channel energies from the CAL (SPR-1993) + (srcRmfgen.cc/h, ConcreteRmfDataServer.cc/h) set the window mode in CAL (SPR-1991) Version 1.43 2001-04-05 (RDS) ------------ + (src/ResponseRow.cc/h, RmfTableIterator.cc) changed F_CHAN by one channel (SPR-1983) Version 1.42 2001-02-21 (UL) ------------ + brought in line with cal-3.99 Version 1.41.3 2000-11-14 (HS) -------------- + Bug in test harness fixed (SOC-SPR-1842). Version 1.41.2 2000-11-08 (HS) -------------- + INTERFACE CHANGE: Removed redundant parameters --responseset, --photondistset --modelpileup (SOC-SPR-1781). Version 1.41.1 2000-11-01 (HS) -------------- + INTERFACE CHANGE: Added new parameter --pnline to enable user to generate a PN RMF for a specific CCD line (SOC-SPR-1780). Version 1.41 2000-10-06 (HS) ------------ + (src/CalMOSResponse.cc): MOS response generation performance improved by taking grade probability calculation outside the energy loop. + INTERFACE CHANGE: renamed param --responseset to --rmfset. Now consistent with naming convention used for arfgen. Version 1.40 2000-09-21 (HS) ------------ + (src dir) PN support added. + (config src dirs) INTERFACE CHANGE: replaced --withenergyranges with --withenergybins. + (test dir) cleaned up. Version 1.39 2000-08-14 (HS) ------------ + (config/rmfgen.info): corrected entries following failure under new chkdist. Version 1.38 2000-08-04 (HS) ------------ + (src/Makefile ./DEPEND): added datautils package. Version 1.37.2 2000-06-15 (HS) ------------ + (doc/rmfgen_description.tex): Cleaned up latex macro usage (SOC-SPR-1567: use separate latex macros for lists). Version 1.37.1 2000-06-15 (HS) ------------ + (config/Makefile): Added rmfgen.info file to CONFIGFILES (SOC-SPR-1527:task info file missing). + (config/Makefile): Replaced rmfgen.prm by rmfgen.par in CONFIGFILES list (SOC-SPR-1565 : config/Makefile has erroneous 'rmfgen.prm' entry) Version 1.37 29 May 2000 (HS) ------------ + Fixed SOC-SPR-1361 : rmfgen supports PHA spectra, not PI spectra Version 1.36 23 May 2000 (HS) ------------ + DEPEND file now includes param-1.54 instead of param 1.55 (which was removed) Version 1.35 22 May 2000 (HS) ------------ + New CAL MOS response model VERY slow - speeded up test harness temporarily by reducing number of rows in o/p RMF. Version 1.34 19 May 2000 (HS) ------------ + INTERFACE CHANGES + rebinned spectra supported + doc in sync with code Version 1.33 22 Feb 2000 (HS) ------------ + Fixed bug in the Rmfgen class that lead to lowmem dal failure and long executin times Version 1.32 1999-12-17 (HS) ------------ + Fixed dummy response data such that it starts at channel=0 instead of channel=1 + Default value for --spectrumset is now spectrum.ds Version 1.31 1999-10-29 (HS) ------------ - Fixed SPR 924 : Strange values for channel numbers in rmf rmfgen now pads out column entries of rows with no data with zeroes. - Minor updates to documentation. Version 1.30 1999-10-29 (HS) ------------ - Compatible with cal-3.31 - Fixed SPR 908 : Inconsistency between XMMSLECT spectra and RMFGEN - Fixed SPR 918 : pad missing rows with 0s - Fixed SPR SPR 923 : Channel number mismatch - Modified interface: removed --instrument and --nchannels parameters, and added a new parameter, --spectrumset. This version is therefore NOT backwards compatible. Version 1.29 1999-09-23 (HS) ------------ - Fixed SPR-838 : Parameter useall should be replaced by withenergyranges (this means a change to the parameter i/f, so this version is NOT backwards compatible) - Added .lyt file Version 1.28 1999-09-15 (HS) ------------ - Fixed SPR-812: Make rmfgen documentation more 'user-friendly' Version 1.27 1999-09-15 (HS) ------------ - Fixed test harness. Mandatory parameter instrument now specified in script. Version 1.26 1999-09-15 (HS/UL) ------------ - brought in line with cal-3.21 Version 1.25 1999-09-06 (HS) ------------ - Removed local latex commands in documentation Version 1.24 1999-06-24 (HS) ------------ - Renamed parameters according to SAS convention (GV). - Rmfgen no longer requires a dataset corresponding to the value of the photondistset parameter, if pile-up is switched off. Version 1.23 1999-06-08 (HS) ------------ - CLOSED SPR-683: rmfgen.cc doesn't compile: wrong semantics results in multiple copies ** PENDING CONFIRMATION *** Version 1.22 1999-06-07 (HS) ------------ - Now compatible with cal-3.0 (ie modified DEPEND file) Version 1.21 1999-05-28 (HS) ------------ - removed lengthy debug messages - fixed HUGE_VAL error raised by Achim Bohnet - The photon dist file is no longer accessed if pile-up (default) is not selected (it is not used in this mode anyway). This means that a photon dist file does not need to be created in the default mode. Version 1.20 1999-05-20 (HS) ------------ - made compatible with error-1.18 Version 1.19 1999-04-27 (HS) ------------ - made compatible with CAL 2.19 Version 1.18 1999-04-13 (HS) ------------ - user interface now accepts energy ranges in keV, whereas the internal units are in eV. Version 1.17 1999-04-11 (MB) ------------ - changed assertion for zero normalization (HS) - fixed eV - keV conflict (HS) Version 1.16 1999-03-23 (MB) ------------ - upgraded to cal-2.10 (new interface for rmf redistribution vector) - added missing test file. Version 1.15 1999-03-15 (MB) ------------ - contains pre-assembled test files, no longer needs deceit to run Version 1.14 1999-03-15 (MB) ------------ - fixed problem with tracer off. - no longer uses dalif. Version 1.13 - 22-02-1999 ------------ - Code now compatible with cal-2.4 and dal-1.43 Version 1.12 - 05-02-1999 ------------ - Code now compatible with CAL-1.21 - Problems when using DAL SubTableProcessor class - no longer inherit from this class as subtable iteration not utilised for time being. Version 1.11 - 04-02-1999 ------------ - Using more flexible RmfDataServer instead of RmfSettings class - Rmfgen takes as input a photon distribution model file instead of a counts spectrum - Changed test harness to generate a photon distribution model using deceit instead of a counts spectrum generation executable. Version 1.10 - 29-01-1999 ----------- - Fixed unsigned int/int classin rectangularSpec.cc Version 1.9 - 11-11-98 ----------- - Corrected test harness: added a ./ in front of rectangularSpec call. Version 1.8 - 10-11-98 ----------- - Removed spurious files (TODO etc) from release. Version 1.7 - 09-11-98 ----------- - New framework from which to implement features such as pile-up correction, "dummy" matrix generation etc has been added. Test script has been modified to test rmfgen with input spectrum. NOTE: Documentation is not currently in line with code. Version 1.6 - 2-10-1998 (MB) ----------- - resolved conflict with new DAL addAttributes. Version 1.5 - 17-08-98 ----------- - Keywords TLMIN4 and TLMAX4 are now added in terms of COLMUN ATTRIBUTES (of F_CHAN) instead of explicitly adding them to the dataset. Version 1.4 - 30-06-98 ----------- - C++ executable modified to become an SAS task proper (top-level fn now called void rmfgen() instead of int main(...) ) Version 1.3 - 18-06-98 ----------- - Refined test script to exit with an error on the failure of any ftool test, and added an additional ftool test using dmprmf. - Added a mechanism to group data above a specified threshold; this will reduce the size of the output rmf file. - closed SOC-SPR-270: F90 file names should be lower-case: the library filledColumnFns and associated f90 source files are no longer used. - Ported rmfgen to C++ Version 1.2 - 03-06-98 ----------- - Declared variables row, nchannel and nenergy as standard integers. This corrects for a NAGACE compiler problem raised by J Ballet (2 Jun 1998) - Separated call to CAL_getRedistribution to two lines (minor). Version 1.1 - 20-05-98 ----------- - Fixed bug related to the number of detector channels assigned to output file. - closed SOC-SPR-218: do not check SAS_DIR and/or source sas-setup.csh (test/testrmfgen) - The following SPRs have been closed as a consequence of changes made for the package sas-1.1: SOC-SPR-217 : remove config/rmfgen.prm Version 1.0 - 04-05-98 ----------- Initial version for SAS v1 integration