.. _install:
============
Installation
============
The easiest way to install Elephant is by creating a conda environment, followed by ``pip install elephant``.
Below is the explanation of how to proceed with these two steps.
.. _prerequisites:
*************
Prerequisites
*************
Elephant requires `Python `_ 3.8, 3.9, 3.10, 3.11 or 3.12.
.. tabs::
.. tab:: (recommended) Conda (Linux/MacOS/Windows)
1. Create your conda environment (e.g., `elephant`):
.. code-block:: sh
conda create --name elephant python=3.8 numpy scipy tqdm
2. Activate your environment:
.. code-block:: sh
conda activate elephant
.. tab:: Debian/Ubuntu
Open a terminal and run:
.. code-block:: sh
sudo apt-get install python-pip python-numpy python-scipy python-pip python-six python-tqdm
************
Installation
************
.. tabs::
.. tab:: Stable release version
The easiest way to install Elephant is via `pip `_:
.. code-block:: sh
pip install elephant
If you want to use advanced features of Elephant, install the package
with extras:
.. code-block:: sh
pip install elephant[extras]
To upgrade to a newer release use the ``--upgrade`` flag:
.. code-block:: sh
pip install --upgrade elephant
If you do not have permission to install software systemwide, you can
install into your user directory using the ``--user`` flag:
.. code-block:: sh
pip install --user elephant
.. tab:: Development version
If you have `Git `_ installed on your system,
it is also possible to install the development version of Elephant.
1. Before installing the development version, you may need to uninstall
the previously installed version of Elephant:
.. code-block:: sh
pip uninstall elephant
2. Clone the repository and install the local version:
.. code-block:: sh
git clone git://github.com/NeuralEnsemble/elephant.git
cd elephant
.. tabs::
.. tab:: Minimal setup
.. code-block:: sh
pip install -e .
.. tab:: conda (with extras)
.. code-block:: sh
conda remove -n elephant --all # remove the previous environment
conda env create -f requirements/environment.yml
conda activate elephant
pip install -e .
***********
MPI support
***********
Some Elephant modules (ASSET, SPADE, etc.) are parallelized to run with MPI.
In order to make use of MPI parallelization, you need to install ``mpi4py``
package:
.. tabs::
.. tab:: conda (easiest)
.. code-block:: sh
conda install -c conda-forge mpi4py
.. tab:: pip (Linux)
.. code-block:: sh
sudo apt install -y libopenmpi-dev openmpi-bin
pip install mpi4py
To run a python script that supports MPI parallelization, run in a terminal:
.. code-block:: sh
mpiexec -n numprocs python -m mpi4py pyfile [arg] ...
For more information, refer to `mpi4py
`_ documentation.
***********************
CUDA and OpenCL support
***********************
:ref:`asset` module supports CUDA and OpenCL. These are experimental features.
You can have one, both, or none installed in your system.
.. tabs::
.. tab:: CUDA
To leverage CUDA acceleration on an NVIDIA GPU card, `CUDA toolkit
`_ must installed on
your system. Then run the following command in a terminal:
.. code-block:: sh
pip install pycuda
In case you experience issues installing PyCUDA, `this guide
`_ offers a step-by-step
installation manual.
If PyCUDA is detected and installed, CUDA backend is used by default in
Elephant ASSET module. To turn off CUDA support, set ``ELEPHANT_USE_CUDA``
environment flag to ``0``.
.. tab:: OpenCL
If you have a laptop with a built-in Intel Graphics Card, you can still
leverage significant performance optimization with OpenCL backend.
The simplest way to install PyOpenCL is to run a conda command:
.. code-block:: sh
conda install -c conda-forge pyopencl intel-compute-runtime
However, if you have root (sudo) privileges, it's recommended to install
up-to-date `Intel Graphics Compute Runtime
`_ system-wide and then
install PyOpenCL as follows:
.. code-block:: sh
conda install -c conda-forge pyopencl ocl-icd-system
Set ``ELEPHANT_USE_OPENCL`` environment flag to ``0`` to turn off
PyOpenCL support.
.. note::
Make sure you've disabled GPU Hangcheck as described in the
`Intel GPU developers documentation `_. Do it with caution -
using your graphics card to perform computations may make the system
unresponsive until the compute program terminates.
.. _no-compile-spade:
****************************
Resolving compilation issues
****************************
Some modules in Elephant make use of C extensions to speed up computation.
However, those extensions need to be compiled before use. In some cases, this
causes problems. For example, the compiler on the current machine does not
fulfill the requirements for the extension, certain libraries are missing,
or no compiler is available at all.
In order to circumvent this problem, the following commands allow to avoid the
compilation for specific or for all C extensions.
.. tabs::
.. tab:: general
Use the following to install elephant without C extensions:
.. code-block:: sh
pip install elephant --install-option='--no-compile'
.. tab:: spade
To avoid compilation of the c++ extension ``fim.cpp`` used in :ref:`spade`, install the package with:
.. code-block:: sh
pip install elephant --install-option='--no-compile-spade'
In this case the pure python implementation of :ref:`spade` is still available.
************
Dependencies
************
Elephant relies on two special packages, installed by default:
* `quantities `_ - support for physical quantities with units (mV, ms, etc.)
* `neo `_ - electrophysiology data manipulations