Installing FSLeyes

There are several ways to obtain and install FSLeyes. FSLeyes comes as a standard component of an FSL installation, but can also be installed independently of FSL.

Note

The terminal commands in these instructions may require administrative privileges, depending on where you have installed FSL.

Install from PyPi (advanced)

Warning

This is an advanced option, recommended only if you are comfortable working with Python environments, and installing packages using your OS package manager.

FSLeyes is available on PyPi, and should work with Python 3.5 and newer. The best way to install FSLeyes from PyPi is to create an isolated python environment with a virtual environment, and install FSLeyes into it. To get started:

python -m venv fsleyes-virtualenv
. fsleyes-virtualenv/bin/activate

macOS users

Once you have activated your virtual environment, you should be able to install FSLeyes like so:

pip install fsleyes

Linux users

Before installing FSLeyes, you first need to install wxPython. The easiest way to do this on Linux is to use the pre-release wxPython builds available at https://extras.wxpython.org/wxPython4/extras/linux/. For example, if you are using CentOS 7:

pip install -f https://extras.wxpython.org/wxPython4/extras/linux/gtk2/centos-7 wxpython
pip install fsleyes

You will also need to install the wxPython runtime dependencies. Under CentOS 7, you will need to run the following command:

sudo yum install SDL

Similarly, under Ubuntu:

sudo apt-get install libsdl1.2debian

Another option is to install wxPython directly from PyPi - if you do this, you will need to have C/C++ compilers installed, and all of the dependencies required to compile wxPython. Under CentOS 7, run the following commands:

sudo yum groupinstall "Development tools"
sudo yum install gtk2-devel gtk3-devel webkitgtk-devel webkitgtk3-devel
sudo yum install libjpeg-turbo-devel libtiff-devel SDL-devel gstreamer-plugins-base-devel libnotify-devel

Under Ubuntu, run the following:

sudo apt-get install build-essential
sudo apt-get install libgtk2.0-dev libgtk-3-dev libwebkitgtk-dev libwebkitgtk-3.0-dev
sudo apt-get install libjpeg-turbo8-dev libtiff5-dev libsdl1.2-dev libgstreamer1.0-dev libgstreamer-plugins-base1.0-dev libnotify-dev

Then you should be able to run pip install fsleyes.

Install standalone build (no longer updated)

Note

FSLeyes 0.32.3 is the last version for which standalone builds are provided. From FSLeyes 0.33.0 onwards, conda is the recommended installation method.

If you wish to install FSLeyes independently of FSL, you can download and install a standalone version from the FSLeyes home page, using these instructions.

  1. Download the zip or tar.gz file for your platform.

  2. Unzip/untar the downloaded file to a location of your choice.

  3. Call FSLeyes like this (you may wish to put the directory containing the fsleyes command on your $PATH environment variable):

    Platform

    Command to run FSLeyes

    Linux

    /path/to/downloaded/FSLeyes/fsleyes

    macOS

    /path/to/downloaded/FSLeyes.app/Contents/MacOS/fsleyes

Install into a Singularity image

FSLeyes can be executed from Docker or Singularity containers. Here is an example Singularity definition file which installs a standalone version of FSLeyes:

Bootstrap: docker
From: centos:7

%help
  FSLeyes Singularity image

%post
  yum -y update
  yum -y install epel-release
  yum -y install wget mesa-libGL mesa-libOSMesa
  wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
  sh Miniconda3-latest-Linux-x86_64.sh -b -p /miniconda3
  /miniconda3/bin/conda create -p /fsleyes-env -c conda-forge fsleyes

%environment
  source /miniconda3/bin/activate /fsleyes-env

%runscript
  fsleyes "$@"