A user manual can be generated by: make docs;

The following packages are required to build and use the client library, and can probably be had from your favorite package manager:

• gsl (>= 1.15): http://www.gnu.org/software/gsl/
• hdf5 with c bindings: http://www.hdfgroup.org/HDF5/
• C++ compiler with C++11 support
• SQUIDS (>= 1.2): https://github.com/jsalvado/SQuIDS/

Run online jupyter notebook example

Additionally for compiling and using the python bindings:

Required:

• boost (>= 1.63): http://www.boost.org/
• numpy: http://www.numpy.org/
• matplotlib: http://matplotlib.org/

Recommended:

• ipython: http://ipython.org/

The LaTeX and pdf files are included in nuSQuIDS/doc/ folder.

The path for the GSL libraries can be specified by running:

./configure --with-gsl-incdir=GSL_include_path --with-gsl-libdir-=GSL_library_path

or, assuming a standard installation into 'include' and 'lib' subdirectories within a common prefix:

./configure --with-gsl=GSL_prefix

The path for the HDF5 libraries can be specified by running:

./configure --with-hdf5-incdir=HDF5_include_path --with-hdf5-libdir=HDF5_library_path

or more simply

./configure --with-hdf5=HDF5_prefix

Finally, the path for SQuIDS can be specified by:

./configure --with-squids-incdir=SQuIDS_include_path --with-squids-libdir=SQuIDS_library_path

or commonly just

./configure --with-squids=SQuIDS_prefix

To compile the python interface it is additionally necessary to pass the --with-python-bindings option to configure. Also, the boost python library, a working python installation, and numpy must be available. The location of the boost library can be specified using:

./configure --with-boost-incdir=boost_include_path --with-boost-libdir-=boost_library_path

or in simple cases

./configure --with-boost=boost_prefix

The python executable (which will also be used to locate numpy) can be specified with:

./configure --python-bin=PYTHON_EXECUTABLE

Once configuration is complete the library can be compiled by running:

make

Unit tests to verify correct behavior can be run using:

make test

A set of example programs demonstrating usage and functionality can be compiled with the command:

make examples

The resulting example executables can then be found in the various subdirectories of examples

Finally the library can be installed using:

make install

By default this will attempt to install the library within /usr/local; this can be changed by using the --prefix option when running configure:

./configure --prefix=$HOME

If you have activated the python interface by doing --with-python-bindings then proceed according to this instructions Even when configured the python interface is not built with the main library. To compile it do the following:

cd resources/python/src/
make

After successful compilation the bindings will be stored in resources/python/bindings/. To make them available from within python, modify your PYTHONPATH:

export PYTHONPATH=$(PATH_TO_nuSQUIDS)/resources/python/bindings/:$PYTHONPATH