560 lines
33 KiB
Plaintext
560 lines
33 KiB
Plaintext
#---------------------------------------------------------------------
|
|
# ChangeLog
|
|
#---------------------------------------------------------------------
|
|
|
|
The ChangeLog will describe feature changes rather than coding detail changes.
|
|
|
|
For detailed information about changes have a look here:
|
|
|
|
https://gitlab.psi.ch/nemu/musrfit/activity
|
|
|
|
or
|
|
|
|
https://bitbucket.org/muonspin/musrfit/commits/all
|
|
|
|
Release of V1.6.5, 2021/01/22
|
|
=============================
|
|
|
|
Add Qt6 version of musredit, mupp, etc.
|
|
|
|
Release of V1.6.4, 2020/08/29
|
|
=============================
|
|
|
|
Allow to view time spectra and Fourier theory only calculated at the data points
|
|
via the option '-1' or '--one_to_one'.
|
|
|
|
Release of V1.6.3, 2020/07/27
|
|
=============================
|
|
|
|
A single asymmetry plot allows to toggle the color of the theory function line
|
|
by the key 't'.
|
|
|
|
Release of V1.6.2, 2020/06/17
|
|
=============================
|
|
|
|
Functions not only can use parameters and maps but now also meta data from the
|
|
data files like, field (B or b), energy (En or en), or temperature (T<x> pr t<x>
|
|
where <x> is a number starting from 0, e.g. t0).
|
|
|
|
Release of V1.6.1, 2020/05/25
|
|
=============================
|
|
|
|
For asymmetry fits: allow alpha, beta in the RUN block to be functions.
|
|
|
|
Release of V1.6.0, 2020/05/16
|
|
=============================
|
|
|
|
Allow to transform parameter vectors of run collections for mupp.
|
|
This allows for instance to calculate the superfluid density directly from the
|
|
depolarization rate and plot it.
|
|
|
|
Release of V1.5.2, 2020/02/03
|
|
=============================
|
|
|
|
Implemented a SECTOR command. It allows to get chisq/maxLH information from
|
|
different time sector slots.
|
|
|
|
Release of V1.5.0, 2019/05/15
|
|
=============================
|
|
|
|
Added beta-NMR support (thanks to Zaher Salman).
|
|
|
|
Release of V1.4.2, 2019/03/08
|
|
=============================
|
|
|
|
The tool chain support for automake is removed. Only the cmake tool chain is
|
|
supported from now one.
|
|
|
|
Release of V1.2.0, 2017/10/26
|
|
=============================
|
|
|
|
Adopted to be ready for rpm builds in a first step. This will be followed by debian style
|
|
packages.
|
|
|
|
Release of V1.1.0, 2017/10/26
|
|
=============================
|
|
|
|
Added full DKS support, i.e. additionally to CUDA also OpenCL for GPU's and CPU's in place.
|
|
|
|
Release of V1.0.0, 2017/05/18
|
|
=============================
|
|
|
|
I feel that musrfit has majored enough to lift it to its first full version, i.e. V1.0.0.
|
|
|
|
A lot of bug fixing went into it since 0.18.0. It is now also fully ROOT 6 compatible which
|
|
is also the main supported version now.
|
|
|
|
New features:
|
|
|
|
* musrWiz: a wizard which will help the user to get started to generate msr-files.
|
|
* musrStep: a little helper program which allows to reset the step size of a msr-file.
|
|
This is useful if you deal with a lot of detectors (e.g. HAL-9500 at PSI).
|
|
* auto-phase correction of Fourier transforms: this is still in its early stage but will
|
|
eventually provide a real phase corrected Fourier transform, which will allow
|
|
a real Fourier detector average.
|
|
* the startup file handling has improved substantially.
|
|
|
|
changes since 0.18.0
|
|
===================================
|
|
|
|
changes since 0.17.0
|
|
===================================
|
|
NEW 2016-12-18 for maxLH it is now possible to get the per-run-block maxLH,
|
|
and the expected maxLH.
|
|
NEW 2016-12-18 musrview allows no to show directly the Fourier transform,
|
|
and the averaged data/Fourier transform.
|
|
NEW 2016-12-18 added a first full version of an automatic phase correction
|
|
for the real part of the Fourier transform. It is based on
|
|
the algorithm published in JMR 158, 164 (2002).
|
|
NEW 2016-12-18 a couple of improvements in dump_header: (i) it is not
|
|
only searching in the current directory, but also in all
|
|
default paths defined in musrfit_startup.xml. (ii) it is
|
|
now possible also to give only run numbers instead of the
|
|
full file name.
|
|
NEW 2016-08-24 in musrview it is now possible to enable/disable a
|
|
crosshair cursor by pressing 'c'.
|
|
NEW 2016-04-28 updated licence info in musredit and added paramList
|
|
feature to the msr2data GUI.
|
|
NEW 2016-04-28 msr2data gets a new option: paramList which allows to
|
|
extract a subset of all the parameters of a collection
|
|
of msr-files.
|
|
NEW 2016-04-22 Added the theory function muMinusExpTF for mu minus fits
|
|
NEW 2016-02-23 It is now possible to export the averaged data/Fourier
|
|
CHANGED 2016-12-18 updated the docu.
|
|
CHANGED 2016-12-18 a couple of little improvements in musredit.
|
|
CHANGED 2016-08-10 drop NeXus support for Version < 4.4
|
|
CHANGED 2016-04-26 start-/endTimeBin are now class members. This reduces
|
|
the number of recalculations.
|
|
FIXED 2016-08-02 run lists are now properly loaded if containing nS-nE elements.
|
|
|
|
changes since 0.16.0
|
|
===================================
|
|
NEW 2016-01-22 update of the docu which now describes the RRF option.
|
|
CHANGED 2016-02-17 move the startup XML-files to $HOME/.musrfit
|
|
rather than having it within $prefix/bin. Increase
|
|
therefore the version number to 0.17.0
|
|
CHANGED 2016-02-16 if the LEM file is created 2012 or later and the
|
|
data-file-format tag is still ROOT-NPP or ROOT-PPC, it
|
|
will automatically switched to MUSR-ROOT. This way red/green
|
|
issues should be prevented. (see MUSR-314).
|
|
CHANGED 2016-02-16 start migrating the icons of musredit_qt5 to the breeze
|
|
icons (https://github.com/NitruxSA/breeze-icon-theme).
|
|
At the same time I start to redesign the musrfit specific
|
|
icons accordingly (not yet happy with it, but ...).
|
|
|
|
changes since 0.15.0
|
|
===================================
|
|
NEW 2016-01-22 add RRF fit option. At the same time, the version has been
|
|
increased to 0.16.0. For details see the documentation.
|
|
FIXED 2015-10-29 TMusrRunHeader didn't map a double vector properly into a
|
|
string. The last element was a copy of the 2nd last instead
|
|
if the proper last. This is now fixed. Fixed also some
|
|
comment typo.
|
|
FIXED 2015-11-11 Fixes the crash of Fourier if the maximal range given in the
|
|
Fourier block is larger than the available.
|
|
CHANGED 2015-11-15 removed unnecessary check form musredit_qt5 msr2data interface.
|
|
|
|
changes since 0.14.0
|
|
===================================
|
|
NEW 2015-10-26 added a first ported musredit version (Qt4 -> Qt5). At the same
|
|
the version has been incremented to 0.15.0.
|
|
NEW 2015-09-24 adding a phase optimized real Fourier to musrFT. This is still VERY
|
|
experimental, and hence only available from within musrFT. Eventually
|
|
it should make its way into musrview as well. Furthermore the
|
|
documentation needs to be written yet. There is only a short not in
|
|
the command line present.
|
|
NEW 2015-02-23 implemented an average-per-data-set option for musrFT.
|
|
NEW 2015-02-21 add proper Mac icon to musredit
|
|
FIXED 2015-09-22 generating global fit msr-files including a GLOBAL block is working
|
|
now as expected.
|
|
FIXED 2015-09-17 in PMsr2Data::PrepareGlobalInputFile() there seem to be 'unmotivated'
|
|
break; commands in some loops. They prevent a proper map handling.
|
|
Since this is a real puzzle I contacted BMW for clarification.
|
|
2015-09-18: there is only one unmotivated break; cleaned up the code
|
|
accordingly.
|
|
FIXED 2015-09-14 any2many export of MusrRoot crashed when first histo group was != 0.
|
|
This happend when exporting to PSI-BIN or WKM. This is fixed now.
|
|
FIXED 2015-09-08 fixed error in view_packing for single histo (wrong norm of the theory).
|
|
FIXED 2015-09-04 minor change in Fourier-block output (PMsrHandler::WriteMsrLogFile
|
|
and PMsrHandler::WriteMsrLogFile) to avoid truncated range labels.
|
|
FIXED 2015-06-30 changed setf/unsetf bug (PMsr2Data.cpp)
|
|
FIXED 2015-05-15 another path check for HDF5 with ubuntu 15.04
|
|
FIXED 2015-03-16 some minor correction for ASCII export in any2many
|
|
FIXED 2015-02-26 some more bug fixing for average per data set.
|
|
FIXED 2015-02-20 for graphical export in the batch mode, menus mustn't be set.
|
|
CHANGED 2015-08-17 updated ChangeLog which was orphaned since V0.13.0
|
|
CHANGED 2015-08-16 updated the docu
|
|
CHANGED 2015-08-16 added the class PStringNumberList which allow to parse a
|
|
generic encoded list of numbers. Used this class to improve
|
|
PMsrHandler, and improved the runList feature of msr2data.
|
|
For details see the documentation of msr2data.
|
|
CHANGED 2015-06-25 adopted the temperature dependence of the gap function
|
|
(see Eq.(8) of the memo), which breaks the self-consistency.
|
|
Makes it more flexible but requires that the user is using
|
|
his brain.
|
|
CHANGED 2015-06-25 small correction of the skewed Gaussian for extrem parameter
|
|
values.
|
|
CHANGED 2015-05-29 updated docu and removed an unsed variable
|
|
CHANGED 2015-05-29 added a group histo selector in any2many for MusrRoot when
|
|
exporting to a too limited file format as PSI-BIN.
|
|
CHANGED 2015-04-01 update of the technical docu
|
|
CHANGED 2015-02-28 check that FOURIER block phase parameter as parX is not leading
|
|
to an error message.
|
|
CHANGED 2015-02-24 added missing background range from msr-file (musrFT)
|
|
CHANGED 2015-02-24 separated PSI-BIN and PSI-MDU file format
|
|
CHANGED 2015-02-21 make sure that musrFT is found
|
|
CHANGED 2015-02-21 needed to adopted includes for dictionary generation due to
|
|
adding fftw.h in PMusr.h
|
|
CHANGED 2015-02-21 have now README and README.md for automake and bitbucket,
|
|
respectively.
|
|
|
|
changes since 0.13.0
|
|
===================================
|
|
NEW 2015-02-23 implemented an average-per-data-set option for musrFT.
|
|
NEW 2015-02-20 add a switch to musrview that it plots initially the Fouier data rather
|
|
than the time domain data.
|
|
NEW 2015-02-19 added a first preliminary user interface for musrFT within musredit.
|
|
NEW 2015-02-16 changed the data export handling from musrview. It is now more
|
|
main line and follows the implementation of musrFT.
|
|
NEW 2015-02-13 first implementation of a standalone Fourier transform/plotter:
|
|
musrFT. Initially it is meant to be used for HAL-9500,
|
|
i.e. Fourier transform WITHOUT lifetime correction.
|
|
A first simple minded lifetime correction is implemented as well.
|
|
NEW 2015-02-04 Integration of libBNMR for fitting beta-NMR relaxation data
|
|
into automake of musrfit.
|
|
NEW 2015-01-17 adding a branch for ROOT 6.x. This needs some minor adaptations due
|
|
to the new rootcint/rootclang and the stricter c++11.
|
|
NEW 2014-12-18 first implementation of a GLOBAL block which allows to shorten
|
|
a typical msr-file. Duplicate entries from the RUN blocks can be
|
|
added here. Furthermore, the 'lifetimecorrection' flag is
|
|
transferred from the RUN blocks to the PLOT block which allows
|
|
a quicker switching between asymmetry and histogram representation.
|
|
In order to make the code more readable a lot of setup collection
|
|
parameters (t0, data, fit) has been encapsulated into its own
|
|
functions.
|
|
|
|
changes since 0.12.0
|
|
===================================
|
|
NEW 2014-12-04 Fourier: added the unit 'Tesla' needed e.g. for HAL-9500
|
|
NEW 2014-12-04 Added the possibility to show the average of currently viewed data.
|
|
This is useful e.g. when looking at the Fourier of lots
|
|
of histogram channels (detectors).
|
|
FIXED 2014-12-04 added proper alignment for averaging. Still missing: error propagation
|
|
for averaging in time based data.
|
|
|
|
changes since 0.11.0
|
|
===================================
|
|
NEW 2014-02-07 added new internal field functions internFldGK and internFldLL to
|
|
PTheory. Online documentation still missing.
|
|
NEW 2013-09-10 add a stretched Kubo-Toyabe function to the PTheory class, including
|
|
an update of the docu.
|
|
NEW 2013-02-14 (i) adding t0 estimate for single histogram fits. The estimate
|
|
procedure is documented under doc/memos/estimateN0.
|
|
(ii) adding a more flexible preference handling in musredit.
|
|
Likely it will not ported anymore to musrgui.
|
|
(iii) adding 'recent files' to musredit.
|
|
In order that this changes are working properly, the configure
|
|
script needs to be executed once more, sorry.
|
|
NEW 2013-01-15 adding an external spin valve related library, currently
|
|
only containing a skewed Lorentzian.
|
|
NEW 2013-01-07 print out the estimated time needed for the Minimize,Minos, etc.
|
|
both, to the stdout and to the MINUIT2.OUTPUT file.
|
|
NEW 2012-11-19 added a flag in the Fourier block (dc-corrected) which can
|
|
be used to subtract a potential DC-offset before the Fourier
|
|
transform is carried out.
|
|
NEW 2012-10-25 (i) add PRINT_LEVEL to the command block (0='nothing' to
|
|
3='everything'). This allows to tune the Minuit2 output. (ii) added the
|
|
possibilty to give the fit range in bins. For details see the docu.
|
|
NEW 2012-09-24 add header information for printing.
|
|
NEW 2012-05-31 added Noakes-Kalvius function (see A. Yaouanc and P. Dalmas de Reotiers,
|
|
"Muon Spin Rotation, Relaxation, and Resonance" Oxford, Section 6.4.1.3).
|
|
NEW 2012-05-25 musredit/musrgui: added a dump muSR data header file information.
|
|
NEW 2012-05-22 added spin rotation angle to the LEM MusrRoot file.
|
|
NEW 2012-05-12 added dump_header. This is a little program which dumps the
|
|
header information of a given muSR data file onto the standard
|
|
output. Since dump_header also includes the option of
|
|
nexus_dump and read_musrRoot_runHeader these latter programs
|
|
were removed.
|
|
NEW 2012-04-24 added a first version for negative muon fitting. At the same
|
|
time substaintial bug fixing has been carried out (mainly the
|
|
logx/logy handling).
|
|
FIXED 2013-12-16 fixed a couple of bugs on level of config. This is needed especially
|
|
for Mac OS X 10.9 (Mavericks).
|
|
FIXED 2013-11-29 in the global mode of msr2data for run lists, instread of the
|
|
value of the global parameter, the positive error was written into
|
|
the DB file. This is fixed now.
|
|
FIXED 2013-11-20 suppress automatic N0 esitmate for chisq-only calculation
|
|
MUSR-270.
|
|
FIXED 2013-01-21 fixed any2many bug concerning input file lists as described
|
|
in MUSR-247. Rewrote the template browser.
|
|
FIXED 2012-12-18 fixed a shortcoming of the outdated TLemRunHeader class
|
|
(see MUSR-246).
|
|
FIXED 2012-11-19 fixes issue with too many t0's in the msr-file (see
|
|
MUSR-238).
|
|
FIXED 2012-11-19 the Fourier transform had 1 Bin too much, this is fixed now
|
|
FIXED 2012-11-06 changed to out formating of the fitting parameters such that
|
|
enough relevant digits are written, which was not always the
|
|
case previously.
|
|
FIXED 2012-10-29 fixed missing read call in PMsr2Data.
|
|
FIXED 2012-09-23 fixed wrong chisq output in musrview if expected chisq is
|
|
present.
|
|
FIXED 2012-05-30 fixed RRF bug in single histo plotting.
|
|
FIXED 2012-05-18 fixed wrong forward/backward tag for ROOT-PPC (MUSR-215)
|
|
CHANGED 2014-11-10 (i) added cross-check for Asymmetry view which makes sure that
|
|
fbg-t0 is equal for forward and backward histo.
|
|
(ii) extended the command block for the test-nonlocal.msr in
|
|
order to have an example which shows the more 'advanced' features
|
|
of minuit.
|
|
CHANGED 2014-10-28 BMWlibs, calculation of superconducting gap functions:
|
|
allow to choose between to different parameterization of the
|
|
temperature dependence of the gap. For details see the memo
|
|
in the source code.
|
|
CHANGED 2014-10-25 updated docu, since git is now available for ALL users
|
|
CHANGED 2014-02-12 since we moved to git, I cleaned up the svn prop's from
|
|
the sources.
|
|
CHANGED 2013-12-20 upgrade of cuba to version 3.2. Merge in from BMW
|
|
CHANGED 2013-12-16 prettyfied the Noakes-Kalvius formulae. Furthermore added
|
|
a sub-folder with cross checks for these formulae.
|
|
CHANGED 2013-11-12 changed normalization of log max likelihood according to S.
|
|
Backer and R.D. Cousins NIM 221, 437 (1984), in order to have a
|
|
"goodness-of-fit" criteria.
|
|
CHANGED 2012-12-11 if multiple SAVE are present in the COMMAND block, append
|
|
MINUIT2.OUTPUT file. Added docu for PRINT_LEVEL (MUSR-244).
|
|
CHANGED 2012-11-19 replaced hard coded gyromagnetic ratio of the muon in the
|
|
whole code.
|
|
CHANGED 2012-11-09 when converting to WKM, take the beamline tag to decide
|
|
which exact WKM flavour should be written.
|
|
CHANGED 2012-08-28 prevent LF from allocating too much memory
|
|
CHANGED 2012-06-29 changed handling of the timeout for musrfit.
|
|
CHANGED 2012-05-10 prevent any2many from overwriting an input file. At the
|
|
some additional bug fixing of any2many has be carried out.
|
|
CHANGED 2012-05-08 updating docu
|
|
|
|
changes since 0.10.0
|
|
===================================
|
|
NEW 2012-04-15 added timeout facility for musrfit, musrview, musrt0. Adopted musredit/musrgui accordingly.
|
|
This is at least a partial workaround until it can be pinned down why sometimes musrfit, etc.
|
|
do not terminate properly.
|
|
NEW 2012-03-28 added a validator program (musrRootValidation) for MusrRoot files.
|
|
NEW 2012-03-22 added a first version of MusrRoot, the next muSR file format at PSI. At the same time a rewrite of
|
|
the low run data handling has been done which enables a proper Red/Green handling with histogram sets addressable
|
|
either via the histogram number or the index in the data vector.
|
|
CHANGED 2012-03-22 removed outdated musrparam.cpp. This program is not needed anymore since msr2data is much
|
|
more powerful and deals with all (and much more) options which where present in musrparam.
|
|
|
|
changes since 0.9.0
|
|
===================================
|
|
NEW 2011-12-22 added nexus_dump. It is used to dump nexus files in an ascii
|
|
form. For instance it can be used to read header info.
|
|
NEW if the background is estimated form an interval rather than fitted, it will be added as a comment
|
|
to the background interval tag. (MUSR-192).
|
|
NEW forward/backward accept now not only c0 c1 c2 ... cn, but also c0-cn cm-cp, or c0 c1-cn cm cx-cy, etc. (MUSR-201, improvement whish).
|
|
NEW added minimal NeXus IDF 2 support.
|
|
NEW Added the online documentation to the repository. It can be found under "doc/html".
|
|
NEW added chisq per run block information (all fit types). The same output rules as for the expected chisq are in place.
|
|
NEW calculate expected chisq (Pearson's chisq) for single histogram fits. It always will send this information to the stdout
|
|
but only writes it to the msr-file if a corresponding flag in the musrfit_startup.xml is enabling it. The following
|
|
expected chisq's are calculated: total expected chisq, and the per histo expected chisq. (MUSR-194)
|
|
NEW the chi^2 calculation in single-histogram and asymmetry fits is parallelized
|
|
if musrfit is built using a compiler supporting OpenMP (e.g. GCC >= 4.2)
|
|
Using --disable-omp this feature can be disabled on the configure level.
|
|
NEW any2many: force the user to define the exact NeXus ouput format (HDF4,HDF5,XML)
|
|
FIXED 2012-01-05 fixed a bug in PRunAsymmetry::PrepareData relevant when combining addrun's and grouping (MUSR-209).
|
|
Remove some commented out stuff from PRunSingleHisto.
|
|
FIXED 2011-12-23 changed Qt parts of configure.ac in order to be compatible
|
|
with BSD MacOS X
|
|
FIXED 2011-12-22 some more work towards a proper NeXus muon data format support.
|
|
FIXED proper phase correction for Re/Im Fourier transform (MUSR-206)
|
|
FIXED when release or restore is called (command block commands), the
|
|
corresponding parameter error is set to 2% of its parameter value (MUSR-188).
|
|
FIXED the wrong normalization when using "SCALE_N0_BKG FALSE" together with view_packing (MUSR-184)
|
|
FIXED crash in non-interactive mode of musrt0 when data file doesn't exist (MUSR-133)
|
|
FIXED wrong asymmetry fit plotting if data range is not provided (MUSR-203)
|
|
FIXED broken run-list interface to msr2data in musredit/musrgui (MUSR-202)
|
|
FIXED some severe issues in the run-number-digit determination of msr2data introduced in late 2010
|
|
FIXED a musrview exception error when starting with t0's very much off to the positive side in the msr-file (MUSR-199)
|
|
FIXED a linking problem when only shared libraries are built on Cygwin
|
|
FIXED the problem that in certain environments XML files could not be parsed (MUSR-122)
|
|
FIXED crash of musrview in case the XML startup file is present but cannot be parsed correctly
|
|
FIXED crash in case a non-existing userFcn is called from an existing library (MUSR-159)
|
|
FIXED slightly wrong chisq since the 0-bin entries were treated wrongly for scaled N0/Bkg data (MUSR-193)
|
|
FIXED the evaluation of the LF depolarization functions (before numerous unnecessary calculations were performed)
|
|
FIXED casting problem between uint32 and time_t for some compilers (MUSR-185)
|
|
FIXED bug reported in MUSR-183: missing background for 2nd histo in asymmetry fits when using musrt0.
|
|
FIXED Makefiles so that the NeXus support will not be built if it has not been enabled during the configure stage
|
|
FIXED ASCII export from musrview in case of a Fourier-Power- or Fourier-Phase-difference plot
|
|
FIXED bug in asymmetry fit with fixed background
|
|
CHANGED Nonlocal: optimized the number of calculations of B(z)
|
|
CHANGED Integrated building and installing the Qt editor into the musrfit installation.
|
|
CHANGED musredit/musrgui msr2data interface such that it is hopefully closer to the command line (MUSR-162, MUSR-173)
|
|
CHANGED the behavior of msr2data so that
|
|
* it proceeds to the next run if a fit did not converge (and does not stop as before)
|
|
* it always tries to read the data files if the nosummary option is not present
|
|
(before this was switched off for the next runs when the data file of the present run could not be read)
|
|
CHANGED in case first good bin (fgb) is not given, it is now fgb=t0+(10ns/time resolution) (MUSR-190)
|
|
CHANGED improved handling of 'old' ROOT LEM data files (2006 and earlier) for any2many (MUSR-178)
|
|
CHANGED improved handling of WKM/ASCII in PRunDataHandler for any2many (MUSR-172)
|
|
CHANGED the data-file-name handling (MUSR-195).
|
|
It is now possible to specify the full file name in the RUN block (including the extension).
|
|
CHANGED cosmetics in the musrview legend concerning the units
|
|
CHANGED the user-function dependency from libPMusr to libPUserFcnBase
|
|
|
|
musrfit 0.9.0 - changes since 0.8.0
|
|
===================================
|
|
NEW added NeXus support. Still a lot of testing will be needed, documentation updated, etc.
|
|
NEW added a magnetic proximity user function to the ASlibs
|
|
NEW added the command SCALE_N0_BKG TRUE | FALSE to the command-block. This can be used to force a single histogram fit
|
|
to use either 1/ns scaling for N0 and background or 1/bins one.
|
|
NEW any2many: some more work, including the PSI-BIN write routines which are officially not released yet.
|
|
NEW extended global mode in msr2data
|
|
NEW any2many: an attempt to write the universial musr-data-file converter. Just started, needs still some work.
|
|
NEW musrt0: added the possibility to show the t0 saved in the data file 's'. Furthermore added the option
|
|
--getT0FromPromptPeak, -g with <firstGoodBinOffset>: will, in non-interactive mode estimate the t0's from
|
|
the prompt peak and write it into the msr-file (MUSR-133).
|
|
FIXED formatting issues in the FUNCTIONS block (MUSR-171)
|
|
FIXED plotting issues described in MUSR-174.
|
|
FIXED accidental casting of a NULL pointer in msr2data
|
|
FIXED crash of musrfit if a run-data-file is not found (MUSR-176)
|
|
FIXED corrected time resolution for WKM from (ns) -> (us)
|
|
FIXED the formating of fixed parameters (step=0) where wrongly formatted (connected to MUSR-167). This
|
|
should be done correctly now.
|
|
FIXED for single histogram fits, the chisq given was wrong (not the fit-result though). I missed some needed scaling
|
|
when normalizing to 1/ns rather than bins. The same is true for log max. likelihood. I am rather depressed that
|
|
nobody so far found this but only R. Scheuermann pointed out there might be a problem.
|
|
FIXED background range correction, such that it is a multiple of the proton cycle time
|
|
FIXED linking of BMWlibs on Cygwin
|
|
FIXED various bugs in msr2data
|
|
FIXED 2 little annoying problems: (i) now it is possible to zoom down to the single bin in musrview.
|
|
(ii) when switching between data- and difference-view, the x-range doesn't change anymore.
|
|
FIXED musrt0 crash for histogram number out of range (MUSR-157)
|
|
FIXED fixes the inadequate attempt to use log max likelihood fit for asymmetry/non-muSR fit (MUSR-148)
|
|
CHANGED the file name of the XML startup file for some of the libraries included in the BMWlibs to "BMW_startup.xml"
|
|
CHANGED added an additional layer for process killing to musrgui/musredit which hopefully should help to reduce the number of
|
|
musrfit/musrview corpses with high CPU load.
|
|
CHANGED the formating of the parameters such that they show the precision corresponding to the error. At the
|
|
same time some other parameter formating is improved (MUSR-167)
|
|
CHANGED the default behavior of msr2data for writing output-file headers (see svn-log 4758)
|
|
CHANGED the behavior of msr2data when non-existing files are encountered---as far as possible they should be ignored now
|
|
CHANGED less strict handling of empty FUNCTION block
|
|
CHANGED cosmetics in the y-labelling (MUSR-154)
|
|
CHANGED maximum possible run number for the use in msr2data to numeric_limits<unsigned int>::max() (MUSR-155)
|
|
UPDATED built-in cuba-library to version 2.1 in BMWlibs
|
|
|
|
musrfit 0.8.0 - changes since 0.7.0
|
|
===================================
|
|
NEW added y-range option to usr_fit_ranges in the PLOT block (MUSR-144).
|
|
NEW added the option to split a user function into a global and run-block related part (see MUSR-134). If the
|
|
global option is used, this can potentially increase the efficiency up to the number of run-blocks since
|
|
the time consuming calculations are done only once per changed parameter set. Examples where this approach
|
|
is profitable are: Abrikosov vortex lattice fits, Nonlocal Meissner screening.
|
|
FIXED compilation of BMWlibs with gcc 3.x
|
|
FIXED hanging musrt0's should be eliminated now (MUSR-153)
|
|
FIXED wrong WARNING message for use_fit_ranges
|
|
FIXED bug in event handling leading to crash for non-muSR when switching between view and difference view
|
|
CHANGED improved y-range scaling for Fourier
|
|
|
|
musrfit 0.7.0 - changes since 0.6.0
|
|
===================================
|
|
NEW added FIT_RANGE RESET | start end | s1 e1 s2 e2 .. sN eN command to the COMMAND block
|
|
NEW added FIX/RELEASE/RESTORE minuit2 command to the COMMAND block
|
|
NEW implemented more checks on the integrity of the msr-file
|
|
NEW [MUSRGUI/MUSREDIT] msr2data default option flags in musredit_startup.xml
|
|
NEW 'global fit' option for msr2data
|
|
NEW implementation of SCAN/CONTOURS/MNPLOT (MUSR-100)
|
|
NEW possibility to use sqrt() and pow(base, exponent) in the FUNCTIONS block
|
|
NEW grouping of histograms
|
|
NEW [MUSREDIT] port of musrgui (Qt3.x) to musredit (Qt4.6)
|
|
NEW [MUSRGUI] better handling of font changes, and adding the possibility to define default font and font size in musrgui_startup.xml (MUSR-105)
|
|
NEW RRF option for single histograms and asymmetry plot
|
|
NEW integration of a few external musrfit plug-ins into the standard build process (disabled by default, see log of svn rev 4313)
|
|
NEW default estimates for 'data' and 'background' entries in RUN blocks
|
|
NEW directory with some msr- and data-files for quick testing of new musrfit installations
|
|
NEW option to musrt0: the key 'T' will set the t0 cursor to the maximum of the histogram
|
|
FIXED chisq/maxLH labels in the statistics block are properly handled now (MUSR-148)
|
|
FIXED y-unzoom failure for musrt0 (MUSR-77)
|
|
FIXED fitting issue for boundaries none/none (MUSR-136)
|
|
FIXED formatting of the ASCII fit output file for long parameter names (MUSR-142)
|
|
FIXED time window frame bug in musrview (MUSR-138)
|
|
FIXED calculation of LF relaxation functions for big static widths (MUSR-139)
|
|
FIXED minor msr2msr issue concerning the statistic block
|
|
FIXED Abragam function
|
|
FIXED removed phase parameter in Fourier power plots (MUSR-116)
|
|
FIXED segmentation fault in musrview when pressing d-f-d (MUSR-115)
|
|
FIXED wrong 'fixed background feeding'
|
|
FIXED [MUSRGUI] system variable access bugs
|
|
FIXED bug in lifetimecorrection-plot (MUSR-123)
|
|
FIXED axis labeling in non-muSR-plot (MUSR-124)
|
|
FIXED non-muSR ascii export first point missing (MUSR-119)
|
|
FIXED handling of wrong run numbers in PLOT block (MUSR-121)
|
|
FIXED reading of MUD data files in 64-bit environments (at least for EM64T)
|
|
FIXED compilation with GCC 4.4/4.5
|
|
FIXED [MUSRGUI] problems if started with an msr-file from the command line which is not in the current directory (MUSR-117)
|
|
FIXED [MUSRGUI] calls to musrview, musrt0 and msr<->mlog when paths contain spaces, etc.
|
|
FIXED usage of signs in the FUNCTIONS block
|
|
FIXED difference of Fourier != Fourier of difference (MUSR-107)
|
|
FIXED handling such that the proper initial phase is fitted
|
|
FIXED segmentation fault if number of histograms is smaller as the specified histogram number (MUSR-110)
|
|
FIXED warning messages
|
|
FIXED db data tag reading error
|
|
FIXED usage of BOOST >= 1.38.0
|
|
FIXED the extraction of the time resolution for the mud data format, since MUD_getHistFsPerBin seems sometimes to return just crap.
|
|
CHANGED musredit: change into the msr-file directory before calling musrview. This is needed if userFcn's are using startup XMLs.
|
|
CHANGED musredit: improved file system watcher notification mechanism to prevent multiple user notifications.
|
|
CHANGED fit range decimal precision handling of PMsrHandler improved (MUSR-150 request). It is now
|
|
handling up to 4 decimal places.
|
|
CHANGED musredit: now it is keeping the directory from which the last msr-file has been loaded. This directory is
|
|
used when a new open dialog is called.
|
|
CHANGED another attempt to get proper zooming/unzooming. This time I tried to disentangle the rather messy PMusrCanvas class.
|
|
Most of the checks seems to work, though still for non-muSR style, musrview is crashing once in a while which seems to
|
|
come from root rather than musrview. I will try to investigate that on root macro level.
|
|
CHANGED in root-5.27.04 the zooming/unzooming handling was modified which forced me to add some more code for
|
|
proper zooming/unzooming via 'u' key
|
|
CHANGED write the ROOT histogram of the correlation matrix to the output file as well as not only the graph
|
|
CHANGED allow case insensitive 'run' as first entry in a runlist-file as input to msr2data
|
|
CHANGED [MUSRGUI/MUSREDIT] '(Un)comment' is splitted into 'Comment' and 'Uncomment'
|
|
CHANGED handling for invalid MUSRFITPATH
|
|
CHANGED build process: checks for MINUIT2, disable static libraries, disable F77 checks
|
|
CHANGED stricter handling of ADDRUN (MUSR-106)
|
|
CHANGED ASCII export from musrview
|
|
CHANGED check for subsequent numbered parameters (MUSR-103)
|
|
CHANGED WKMFULLDATAPATH to MUSRFULLDATAPATH
|
|
UPDATED triumf mud file support. Now the mud source from triumf is officially published under LGPL
|
|
UPDATED source code documentation
|
|
UPDATED PSIBIN/MDU class from AA/AR
|
|
|
|
musrfit 0.6.0 - changes since 0.5.2
|
|
===================================
|
|
NEW support for reading PSI MDU-ASCII data files
|
|
NEW estimate of t0 for the case it is not present, neither in the msr-file nor in the data-file (MUSR-73).
|
|
FIXED 0/1 problem in data and background ranges
|
|
FIXED crash of musrfit in case no 'background' line is present in the RUN-block (MUSR-99)
|
|
FIXED bug with t0 handling in case of ADDRUN and missing t0's in the msr-file, but t0's present in the data file (MUSR-94)
|
|
FIXED 'unwanted empty line problem of mlog-files' (MUSR-95)
|
|
FIXED building of PSI-BIN library on MacOSX
|
|
FIXED severe memory leaks
|
|
FIXED bug in the spin glass function for negative times
|
|
CHANGED improved handling for musrt0
|
|
CHANGED musrfit_startup.xml, added ALC and LEM AFS paths to the default paths
|
|
CHANGED ascii export: Produce now two files, one for the data and one for the theory (MUSR-87, MUSR-92)
|
|
CHANGED encapsulation of data structures in classes
|
|
REMOVED 'save db' support from musrview
|
|
|
|
musrfit 0.5.2 - changes since 0.5.1
|
|
===================================
|
|
NEW initial support for reading TRIUMF MUD data files
|
|
|
|
musrfit 0.5.1 - changes since 0.5.0
|
|
===================================
|
|
NEW Read _measured_ temperature(s) from data files (ROOT, PSIBIN, WKM)
|
|
NEW msr2data rewritten in C++
|
|
NEW in msr2data added support for taking parameters like temperatures and the applied field from the data files (PSIBIN, WKM)
|
|
NEW in msr2data LEM-data RA-HVs are taken from the RunSummary in the ROOT file (if present) - other data is taken from the TLemRunHeader
|
|
|
|
musrfit 0.5.0
|
|
=============
|
|
NEW Initial release supporting building by autotools on Linux, MS Windows (Cygwin), Mac OS X
|
|
|
|
#---------------------------------------------------------------------
|
|
# this is the end ...
|
|
#---------------------------------------------------------------------
|