The NMpi code (nmpi2.f90) is written in GFortran and can be compiled
using the free compiler GFortran which is a part of GCC, the GNU
compiler collection (https://gcc.gnu.org/).

For Microsoft Windows users, the binary file (nmpi2.exe) is provided
in the NMpi package.

For Linux users, the code can be compiled readily using
the GCC tool which is pre-installed in most Linux distributions. If it
is not, please follow the Internet instructions how
to install GCC on Linux (e.g. https://gcc.gnu.org/wiki/GFortranBinaries).

For Mac users, the code can be compiled using GCC tools. However,
in this case, GCC needs to be installed first. Please follow
the Internet instructions how to install GCC on MacOS (e.g.
https://gcc.gnu.org/wiki/GFortranBinaries).

Generally, the code can be compiled using the command:

gfortran -Wall -O3 -ftree-vectorize -fopenmp -o nmpi2 nmpi2.f90

As a result, a binary file (nmpi2) is generated, that enables
parallelized computations (via the OpenMP library) and other
computational optimalizations.

The simplest way to run the analysis is to paste the binary file to
the folder where a data file is located. In advance, please make sure
that all the required input files are prepared and located in
the same directory. Information how to prepare input files is given
in the User manual (pdf document). Since input files provide all
the necessary settings, the analysis starts immediately after launching
the nmpi2 program (i.e. without any additional run parameters).
The analysis progress is updated on the screen.

Users running nmpi2 on a Linux server may find that way
to be worthwhile a try:
nohup ./nmpi2 > nmpi2.prg &

As a result, the nmpi2 program will continue to run even if the user
logs out. All the contemporary information regarding the progress
will be redirected to the file nmpi2.prg. For details, see e.g.
https://en.wikipedia.org/wiki/Nohup.
