|
RMC++ Documentation |
|
|
|
General documentation |
The manual for the latest version, RMC_POT
can be downloaded from here.
The latest speed-test for parallel simulations
can be downloaded from here.
Documentations
for previous versions
RMC++ offers
various options and some improvements from the RMCA code. They are described in
the RMC++ for RMCA users manual (pdf file 92
kB), together with the changes in input/output.
The
changes and additional features and usage of the new versions, RMC++_new
and RMC++multi are described here.
The
RMCA manual (1993) by Robert McGreevy, Jim Wicks and Malcom Howe is a
very useful document, very relevant to RMC++. It can be found here
(pdf file 1 MB).
The
original RMC++ can be seen as a C++, optimized and flexible version of
the existing Fortran codes RMCA (available from the ISIS website) and its
hungarian offspring rmc_fi (which includes fixed-neighbours
constraints).
|
Input and Output formats |
The reference for the development of RMC++
was the existing Fortran RMCA code. Therefore, the input and output formats of
RMC++ have been kept as close as possible to the RMCA ones.
RMC_POT from scratch:
Read the RMC_POT manual, which
describes the file formats as well.
Newcomers to RMC could have a look at the RMCA manual
. Variations in the input format between the original RMCA code and RMC++ have
been kept to a minimum.
Documentation for the auxiliary programs can
be found in the package of the given auxiliary program, and can be downloaded
from the download page.
Previous
versions
RMC++_new and
RMC++_multi for RMC++ users:
Users of RMC++ can use easily the new versions. The
format of the *.dat file and some of the output files has slightly
changed, as described in description
of the changes document.
RMC++ for RMCA users:
Users of the original RMCA programme should be able to use RMC++
straightforwardly.
The programme can be used as such, without any specific RMC++ option.
The input format of files is strictly identical to the RMCA format (apart from
stricter requirements for separators).
The output files format is very close to the RMCA format.
The few differences in input/output files format and screen prompting are
described in the leaflet `RMC++ for RMCA users'
(pdf 92 kB), and concern mostly
The use of options is specified in an additional .add
optional input file.
Using RMC++ without options will just result in a gain of computing time (about
50%). However, some options such as the FNCs are very useful and provide an
essential a priori information to the algorithm.
RMC++ for rmc_fi users:
Users of the rmc_fi version of RMCA (i.e. the version which
includes the fixed-neighbours constraints) should be able to use RMC++ without
much effort.
They must however use a .dat file fully compatible with RMCA (as described in
the RMCA manual), i.e. the additional line corresponding to the FNC
switch must be suppressed. The FNC switch is included in the optional .add
file.
Otherwise, the differences between RMCA and RMC++ formats are described in the
leaflet `RMC++ for RMCA users' (pdf 92 kB).
|
Source code documentation |
The RMC_POT++ source code is available under
the GNU Public License.
You are free to download, use and modify
the code (although at your own risk!).
The brief description of the files of the source code and the classes used can
be found in the RMC_POT manual, and there
are lot of comments in the code as well.
Last modified 29/06/2010 by Orsolya Gereben