||Starting version developed from RMC_combined (combining RMC_new
error occurring only for EXAFS data in case of consecutive or multi-threaded
execution with 1 thread was fixed.
concerning the sigma parameter scaling of the potential related interactions,
introducing NB weight mode=2, and zero for sigma. The npotaccepted was
introduced, it is displayed and read only if potential<0, but it is
written in the configuration file, if _POTENTIAL is on. A small error
concerning the original chi2 calculation for the coordination number
constraint, if scalable sigma was given was corrected. A small error
preventing the program to run in case of the total number of atoms were smaller,
than the desired number of atoms in a gridcell was fixed.
||Some small error were corrected. In DataMat:CalcdV for local
invariance it was prevented, that in case the last bin should exceed the
sqrt(3), array elements of sfactorcube would go over the existing elements of
sfactorcube. In RunParams lead series index=0 was prevented.
||The cosine distribution of bond angles constraint was updated
with the possibility to use an experimental distribution for fitting read
from a file (method 3).
||The local invariance calculation was updated, the distance based
method, and separate dr_loc and xmax_loc was added, the input format changed,
see the manual.
||RunParams.cpp, ChiSquared.cpp, History.cpp
||An error preventing old style (ntotal is the first) *.bcf file
reading, if only *.bcf is present was
corrected. To count the number of FNC pairs out of range is detailed for each
FNC constraint type.
||SimpleCfg, RunParams.cpp, FNC_POT.cpp, RMC_POT.cpp, classes1.h
||A small error caused that in case of bin-based local invariance
calculation if loc_ratio1 is greater than 0 the chi2 loc was a little off.
||It was fixed, that if the atoms in the configuration are more
than one gridcell out of the box, there could be segmentation fault or the
program could stop.
||RMC_POT.cpp, classes1.h, SimpleCfg.cpp, Neighbourlist.cpp
case of potential runs for encad type force field there was some problem for
the 1-4 interactions, as this is calculated by scaling.
||Topology.h, Topology.cpp, RunParams.cpp, FNC_POT.cpp
||Changed the place of the PrintRunStatus to after the CalcSigma
was fixed, that the program should
run without problem, even if only the *.bcf file is present.
||Topology.cpp, SimpleCfg.cpp, RMC_POT.cpp
||The type of the cctype variable was changed to the user defined
longint type, so the compiler option _USE_INT64 will decide, maximum how many
atom types (consequently partials) can be used in coordination constraint runs. If _USE_INT64 is turned on
maximum 10 atom types can be used, if not max 7 types. The old code should
not be used in case of more than 7 atom types, unfortunately the old code
although give a warning, but did not stop. Now this was fixed.
||The program did not stop in case of GROMACS type mismatch, also
it indicated that it will, this was corrected.
||Only if compiled with the potential option, and if Coulomb
potential was calculated. If there were an atom with the same GROMACS type,
but different charge, than previously, and only if a new GROMACS type follows
that, then there was a mistake with the resizing of the charge array, and so
forth infinite value appeared for the charge. Easy to recognize, as the
Coulomb potential become infinite too. The error was fixed.
case of cosine distribution constraint an error (which did not effect the
simulation) is given at the end of the simulation when an array which was not
created was tried to be deleted. This was fixed.
||In case of potential runs the ResizeArray of the top_def_option
gave a compiler warning, which was fixed, and the correct size
(FILE_NAME_SIZE) was given at the end of the line. This did not result in
cubic term was added to the S(Q) and F(Q) data renormalization, and the
constant+linear+quadratic+cubic term was added to g(r) renormalization. This
can only be used, if a 1 is specified in the *.dat file after the maximum
index of used data points to make it possible to use the old format files.
||RunParams, RMC_POT.cpp, ChiSquared.cpp, ExptsData.cpp,
||The possibility to use RMC++ for the simulation of spherical
sample without periodic boundary condition was included, which can be useful
for the calculation of nanoparticles. Use the _NO_PERIODIC compiler option
||RMC_POT.cpp, RunParams.cpp, DataMat.cpp, History.cpp,
CalcData.cpp, chiSquared.cpp, Move.cpp, makemove.cpp, SimpleCfg.cpp,
classes1.h, classes2.h, alter.h, Move.h and the linux Makefile
||The possibility to calculate the displacement of the atoms due
to atomic vibrations by convoluting the histogram with a Gaussian function
has been included. Use the _VIBR_AMP compiler option for compilation.
||RMC_POT.cpp, altern.h, classes1.h, classes2.h,
Threads.h,.RunParams.cpp, PPCFSet.cpp, History.cpp, HistoSet.cpp,
CalaPart.cpp, DataMat.cpp, Threads.cpp, and the linux Makefile
error was corrected, which occurred only if the program was compiled with the
_POTENTIAL option on, and potential based chi2 was separately evaluated from
the normal chi2, and average coordination number constraint was used during
the run . If the move was rejected because of the increase of the potential,
and in this move the average coordination number changed, then the previous
average coordination number was not copied back, so the update of the average
coordination number was incorrect.
||An error was corrected which occured during the histogram
update, if local invariance was used with more than one moved atom and/or
||Some modification was made to avoid innecessary error message,
if during the run the chi2 of the leading series went down to zero.
_AV_MOVE option was extended to save the moved distance of each atoms to the
||RMC_POT.cpp, classes1.h, global.h, SimpleCfg.cpp
average coordination number constraint was changed a bit for _NO_PERIODIC. If
EXAFS is used for one component system,
the data reading form the *.dat file would stop, this was fixed.
RunParams.cpp, ExptsData.cpp, AvCoordConst.cpp, classes1.h, Threads.h,
Threads.cpp, CoordNumbConst.cpp, ChiSquared.cpp, PPCFSet.cpp
in case of _NO_PERIODIC _VIBR_AMP and/or EXAFS is used, than the ppcf
calculation was wrong, this was amended.
||If in case of _NO_PERIODIC _VIBR_AMP the load histogram option
was used, then hist_T was not loaded and recalculated, so the ppcf was wrong.
errors were corrected for the old format output, only important in case of
OLDFORMAT_OUT compiler option
||Formatting errors were corrected for the old format output, only
important in case of OLDFORMAT_OUT compiler option
||The PrintStatus and Print Params was changed for local
||In some cases with the linux version using potential, the empty
line seemed to contain the carriege return character for some reason, which
was not handles and crashed the program during reading of the topolgy file.
This was rectified.
||The possibility to fit
r*[g(r)-1] was inclused, which can be swithed on with the R_SWITCH_GR_MIN_1
CalcPart.cpp, CalcData.cpp, History.cpp, altern.h
||Go to RMC++ home