.. _Miscellaneous:

#############
Miscellaneous
#############

We are collecting various classes and functions that are mainly used by the :ref:`Operator class<Operator>`.
For example, two possible containers to represent the :ref:`Operator<Operator>` are:

.. toctree::
    :maxdepth: 1
    :titlesonly:

    Miscellaneous/SparseMatrix.rst
    Miscellaneous/ShellMatrix.rst

The :ref:`SparseMatrix class<SparseMatrix>` is a simple class that handles sparse matrices. 
It is used by the :ref:`Operator class<Operator>` to store and apply local tensor products.
:ref:`SparseMatrix<SparseMatrix>` does not support :ref:`MPI<MPI_docs>` computation as we strongly suggest :ref:`Petsc and Slepc<Petsc_setup>` in this case, see :ref:`here<Examples_SparseMatrix_petsc_Hamiltonian_imaginary_time_evolution>` for an example.

The :ref:`ShellMatrix class<ShellMatrix>` is used for **matrix-free** matrix-vector multiplications and supports :ref:`MPI<MPI_docs>` and :ref:`openMP<openMP_docs>`.
It is utilized when a `MATSHELL <https://petsc.org/release/manualpages/Mat/MATSHELL/>`_ is created by :ref:`Petsc<Petsc_setup>`, see :ref:`here<ShellMatrix_Petsc>`.
It can be applied the :ref:`Vector class<Vector>`:


.. toctree::
    :maxdepth: 1
    :titlesonly:

    Miscellaneous/Vector.rst


This class simply represent a vector using distributed (if :ref:`MPI<MPI_docs>` is available) or shared memory.
It is mainly used by the :ref:`ShellMatrix<ShellMatrix>` to perform linear algebra routines that can be used to compute ground states, see :ref:`here<Examples_ShellMatrix>`.

Finally, various basic functions and more advanced algorithms like the `Lanczos algorithm <https://en.wikipedia.org/wiki/Lanczos_algorithm>`_ are summarized here:

.. toctree::
    :maxdepth: 1
    :titlesonly:


    Miscellaneous/Functions.rst