.. include:: ../images.rst


CVODE installation
~~~~~~~~~~~~~~~~~~~~~~~

whatis : CVODE solves initial value problems for ordinary differential equation (ODE) systems.

source :
`https://computation.llnl.gov/casc/sundials/download/download.html <https://computation.llnl.gov/casc/sundials/download/download.html>`__

.. warning:: The version used (cvode 2.7.0) is outdated (2016-09-26), but the more recent versions do not work
          with the current eBBH cWB library. The cvode library is contained in the sundials distribution.

.. note::
   CVODE/eBBH is **not** available as a conda-forge package, unlike the
   other auxiliary libraries: it must be built from source as described
   below regardless of which cWB build system is used. For the
   **CMake** build (see :doc:`cwb`), eBBH support is opt-in: pointing
   ``-DEBBH_PATH=/path/to/eBBH/install`` (passed through ``CMAKE_ARGS``) at
   an eBBH install enables it; if not given, the build simply proceeds
   without eBBH.

Installation :

.. code-block:: bash

    cd $HOME_LIBS/CVODE
    1. Download the SUNDIALS/CVODE package version 2.7.0 from

       wget https://computation.llnl.gov/projects/sundials/download/sundials-2.7.0.tar.gz
       or use the package :
       tar xvf $PCKS_REPO/sundials-2.7.0.tar.gz

    2. Uncompress it

       tar -xzvf sundials-2.7.0.tar.gz

    3. Configure it to produce shared libraries and provide the location
    for the installed code ("dist" in this example)

       cd sundials-2.7.0

       mkdir dist
       cd dist
       cmake -DCMAKE_INSTALL_PREFIX:PATH=. ..

    4. Make and install it

       make all install

    5. You should find in dist/lib the two shared libraries sundials_cvode and sundials_nvecserial.


.. note::
   set _USE_EBBH and HOME_CVODE environment variables in watenv script.