Setting up Python for EEG analysis

Python
EEG
Basics
Installing Python is pretty straightforward. In this post, I describe how to set up Python for EEG analysis on Windows, macOS, and Linux.
Published

October 9, 2017

Modified

February 9, 2024

Installing Python

The official installer, which is available for Windows and macOS, is the easiest way for beginners to install Python. If you use Linux, chances are that Python is already installed (if not, use the package manager to install it).

When running the installer on Windows, make sure to check the option “Add python.exe to PATH” (by default this setting is disabled). I strongly recommend to use the default values for all other settings.

On macOS, run both “Install Certificates” and “Update Shell Profile” commands available in the application folder after the installation is complete:

Note

Anaconda, a distribution with tons of pre-installed packages specifically tailored towards scientific computing, provides yet another way to install Python. If you want to use it, consider installing Miniforge instead, which is more lightweight and uses conda-forge as its package repository.

Tip

If you are an experienced Pythonista on macOS or Linux, I recommend pyenv, which makes working with multiple Python versions (and virtual environments) a breeze.

Additional packages

Although Python ships with an extensive standard library, most scientific packages are not part of Python itself. However, installing third-party Python packages is not difficult with the package manager pip, which is bundled with every Python installation. Note that you need to open a terminal to use pip (for example, Windows Terminal on Windows or Terminal on macOS).

Important

On macOS, the tool is available as pip3, just like the Python interpreter is called python3. So whenever you see pip in this document, you should use pip3 instead.

To install a package, you need to know its name. Therefore, the first step is to determine if a particular package is available in the so-called Python Package Index (or short PyPI). Currently, the only way to find out is to search directly on the website.

If the package is available, you can install it with the following command in your terminal (replace <package> with the actual package name):

pip install <package>

The following packages are useful for scientific computing in general and EEG processing in particular:

  • NumPy provides a multi-dimensional array data type, which is the basis for almost all scientific packages.
  • Pandas provides a flexible data frame type similar to the one available in R.
  • SciPy contains numerous algorithms for scientific computing.
  • Matplotlib is the most popular package to create all kinds of plots in Python.
  • Scikit-learn is a powerful machine learning package for Python.
  • MNE is a package for EEG/MEG analysis.
  • MNELAB is a graphical user interface for MNE.
  • IPython provides an enhanced interactive Python experience.
  • Python-Picard is an extremely fast and efficient ICA implementation.

I recommend that you install all of these packages right now so you can use them later in your EEG analyses.

Note

Package names are case-insensitive, so for example pip install NumPy and pip install numpy should both work.

Housekeeping

It is generally a good idea to use the most recent Python version. If a new release becomes available (you might want to check the official Python website once in a while), install it as previously explained. You can even have multiple versions of Python installed on your system if you want (but usually you only need the latest version).

In addition to Python itself, you probably also want to keep all installed packages up to date. Package updates are independent of new Python releases, so you should check for new package versions more frequently (for example, once a week). You can use pip to get a list of all outdated packages:

pip list --outdated

The following command upgrades a package to its latest version (replace <package> with the actual package name):

pip install -U <package>

Unfortunately, you need to upgrade each outdated package individually.

You can also completely remove a specific package (again, replace <package> with the actual package name):

pip uninstall <package>
Tip

If you are already familiar with Python, you should consider working with virtual environments. This will isolate packages for different projects, which is useful for reproducibility and avoids conflicts between different package versions. The official documentation for the venv module is a good starting point.

Visual Studio Code

A code editor or integrated development environment is an essential tool for writing Python programs. Good editors include support for syntax highlighting, indentation, line numbers, linting, code inspection, and more.

I recommend Visual Studio Code if you have never used Python before. However, feel free to try other programs such as PyCharm, which also provides a great Python development experience.

If you decide to install Visual Studio Code, there are two things you should do before you start writing Python code:

  1. Click on the Extensions section in the left sidebar, search for “Python”, and install the official Python extension.
  2. I recommend that you follow the PEP8 coding style. In Visual Studio Code, install the Flake8 extension to enable automatic PEP8-aware linting.

That’s it, you are now ready to start working with Python!