3
0
mirror of https://github.com/triqs/dft_tools synced 2024-12-25 13:53:40 +01:00
dft_tools/doc/install.rst
Michel Ferrero f7fad85fca Iteration over the doc
This is an iteration over the doc mainly thank to Priyanka.
I fixed another couple of details on the way.
2013-12-31 14:22:00 +01:00

83 lines
2.4 KiB
ReStructuredText

.. index:: installation
.. highlight:: bash
.. _installation:
Installation
============
TRIQS and its applications are provided *à la carte*:
after you have installed the TRIQS toolkit, you will be able to easily install
various TRIQS-based applications: impurity solvers, realistic DMFT tools, ...
This page describes the installation of the TRIQS toolkit itself. The installation of the applications
is described in their respective documentation.
Prerequisites
-------------
The TRIQS library relies on a certain number of standard libraries and tools described in
the :ref:`list of requirements <requirements>`. Beware in particular to the :ref:`C++ compilers<require_cxx_compilers>`
and to :ref:`Scientific Python tools<python_install>`.
Here are instructions to install
these necessary libraries on two standard systems:
.. toctree::
:maxdepth: 1
installation/ubuntu
installation/osx_lion
Installation steps
------------------
#. Download the sources of TRIQS from github into a ``src`` directory::
$ git clone https://github.com/TRIQS/triqs.git src
#. Create an empty build directory where you will compile the code::
$ mkdir build && cd build
#. In the build directory call cmake specifying where the TRIQS library should be installed::
$ cmake -DCMAKE_INSTALL_PREFIX=path_to_install_directory ../src
If you omit ``CMAKE_INSTALL_PREFIX``, the default
``path_to_install_directory`` is the subdirectory *INSTALL_DIR* in your build
directory. More configuration options are described in the
:ref:`cmake options <install_options>`.
#. Compile the code, run the tests and install the library::
$ make
$ make test
$ make install
.. note::
We remind you that on multicore machine, make's option -jN (where N is the number of cores)
greatly speeds up the compilation, e.g. make -j8 on an 8 core machine...
#. After the installation TRIQS has been installed using the standard UNIX
scheme:
* executables in ``path_to_install_directory/bin``.
* c++ and python libraries in ``path_to_install_directory/lib``.
* c++ headers in ``path_to_install_directory/include/triqs``.
* shared items (e.g. the documentation) in ``path_to_install_directory/share/triqs``.
Further reading
------------------
.. toctree::
:maxdepth: 1
installation/install_options
installation/python
installation/clang