Brian tutorial at CNS 2014

Modelling of spiking neural networks with Brian

Tutorial T7 at CNS 2014, Québec City (July 26th 2014)

The tutorial will take place in room 2101 at the Québec City Conference Center. Please consider installing Brian 2 beforehand, see the instructions below.

Material

All the material from the tutorial is available in the “2014-CNS-tutorial” folder in the “brian-material” github repository: https://github.com/brian-team/brian-material

You can clone the repository using git, but you can also download all the material in a single ZIP file.

Schedule

In the morning sessions, we plan to give an introduction to Brian, no prior knowledge of Brian is necessary. The afternoon sessions will be about more advanced topics and delve a bit into Brian’s internals.

Brian 2 installation

We recommend that you install Brian 2 before coming to the tutorial. If you already have a working Python installation with the main Brian dependencies (numpy and scipy), you can try to skip directly to the step “Installing Brian 2” .

For setting up a Python environment with the libraries that Brian depends on, we recommend using the Anaconda distribution by Continuum Analytics. On Windows, an alternative to Anaconda is the Python(x,y) distribution, which is in particular recommended if you want to use the C++ code generation/standalone mode on Windows. Download its installer from code.google.com/p/pythonxy/wiki/Downloads?wl=en.

Installing the Anaconda distribution

• Download the appropriate installer from here: continuum.io/downloads
• Follow the installation instructions on the same package
• (Optional) Update the anaconda packages to the latest versions
  conda update conda conda update anaconda

Installing Brian 2

• If you have just installed Anaconda as described above, make sure to open a new Terminal/Command Prompt window and make sure that the Anaconda binary directory is in your path (the Anaconda installer automatically takes care of this by default)
• Install Brian 2 using
  pip install brian2 --pre
• If you already have an older version of Brian 2, use
  pip install brian2 --pre --upgrade --no-deps

Working with Brian 2

• The full Anaconda distribution comes with a Matlab-like development environment for Python called Spyder, you can start it by typing spyder in a Terminal. You can also start it using Anaconda’s “launcher” application, for more details and other IDEs that you can use, see docs.continuum.io/anaconda/ide_integration.html
• You don’t necessarily need a full IDE to work with Python/Brian, starting ipython in a terminal window and using any text editor to edit files is good enough.