mainpage.dox

/*!
@mainpage Walker

@m_div{m-col-t-4 m-col-m-3 m-right-t}
<img src="images/Eo_circle_light-green_letter-w.svg"/>
@m_enddiv

_Walker_ is a mathematical tool to analyze and design the behavior of
stochastic differential equations.

Using the [Charm++](http://charmplusplus.org/) runtime system, it employs
_asynchronous_ (or non-blocking) parallel programming and decomposes
computational problems into a large number of work units (that may be more than
the available number of processors) enabling _arbitrary overlap_ of parallel
computation, communication, input, and output. Then the runtime system
_dynamically_ and _automatically_ homogenizes computational load across the
simulation distributed across many computers.

@section mainpage_tools Tools

Walker consists of the following tools:

@ref walker_main --- __Time-integrator for stochastic differential
equations__

_Walker_ is a mathematical tool to analyze and design the behavior of
stochastic differential equations. Solving a multivariate
[Fokker-Planck equation](https://en.wikipedia.org/wiki/Fokker%E2%80%93Planck_equation),
it allows the estimation of arbitrary coupled @ref statistics_output
"statistics" and probability distributions as they evolve in time and
is used for the design of @ref papers "statistical moment approximations"
for turbulent flows.

@ref rngtest_main --- __Random number generators test suite__

_RNGTest_ is a test harness to subject random number generators to stringent
statistical tests, enabling quantitative ranking with respect to their
quality and computational cost. Multiple generators can be tested from
Intel's [Math Kernel Library](https://software.intel.com/en-us/intel-mkl),
[RNGSSE](https://doi.org/10.1016/j.cpc.2013.04.007), and
[Random123](http://www.thesalmons.org/john/random123/releases/latest/docs/index.html).

@ref unittest_main --- __Test suite for synchronous and asynchronous
functions__

_UnitTest_ is an asynchronous distributed-memory-parallel test harness,
capable of testing serial, synchronous (e.g., MPI) parallel, and
asynchronous parallel (e.g., Charm++) functions.

@section mainpage_build Build

Install prerequisites: _Debian/Ubuntu linux_

@code
apt-get install git autoconf cmake gfortran g++ gmsh libblas-dev liblapack-dev liblapacke-dev libopenmpi-dev libhdf5-dev libhdf5-openmpi-dev libnetcdf-mpi-dev libboost-all-dev tao-pegtl-dev libpstreams-dev librandom123-dev libbackward-cpp-dev libdw-dev binutils-dev
@endcode

Clone, build external libs, build, test

@code
git clone https://codeberg.org/wlkr/walker.git && cd walker
mkdir external/build && cd external/build && cmake .. && make -sj$(grep -c processor /proc/cpuinfo) && cd -
mkdir build && cd build && cmake ../src && make -sj$(grep -c processor /proc/cpuinfo) && cd -
cd build && ./charmrun +p $(grep -c processor /proc/cpuinfo) Main/unittest -v -q && ctest -j $(grep -c processor /proc/cpuinfo) --output-on-failure -LE extreme && cd -
@endcode

All executables will be under `build/Main/`.

@section mainpage_license License

GPL v3.

@section mainpage_history History

Walker was originally developed as part of
[Quinoa](https://quinoacomputing.github.io/archive/) and later was split off
from the mesh-based solvers to contain only the tools in this repository.

*/