Python

Debian Packages

If you're running Debian, it's often easiest to use the official Debian packages (when present) to install python modules. Search for them via apt. For example:

apt search mvpa2

Once you've found the package, install it:

apt install python-mvpa2

If you can't find a Debian package for the module you want installed, then you can to install it using pip — preferably in a virtual environment (see below).

Virtual Environments

Virtual environments (AKA "venvs") allow you to create isolated environments for Python. These environments still have access to the full filesystem, but their "Python World" is their own sandbox to play in, and is completely independent of everything else.

The advantages of this isolation are numerous (tidiness, one project's dependencies won't affect others, ease of troubleshooting, etc). Using venvs should be your default mode of operation.

Virtual environments can be created anywhere on the filesystem, but I like to keep them all together in a hidden folder called .venvs in my home directory.

To create a venv for your new hyperalignment project, run the following:

python3 -m venv ~/.venvs/hyperHyper

NOTE: Python 2.7 users will need to use the virtualenv command (e.g. virtualenv ~/.venvs/hyperHyper)

To activate your new venv run:

source ~/.venvs/hyperHyper/bin/activate

Your shell's prompt will change to denote the venv you are in. Now you can install whatever packages you need (using pip3 install); they will all be stored in ~/.venvs/hyperHyper and will only be available when this venv is activated.

When you're done, deactivating is as simple as running:

deactivate

IPython

IPython is an interactive shell to compute Python code, similar to the Bash shell or the Matlab prompt. IPython features tab-completion, command history retrieval across sessions, dynamic object introspection, and magic commands that provide some nice quality-of-life syntactical sugar. To begin an IPython session, simply run:

ipython

Resources

Interactive Book: Foundations of Python Programming: A "projects first" approach that focuses on building things using Python rather than focusing on the language itself.
Book: Learn Python the Hard Way: Popular with good content. People either enjoy or strongly dislike the book's approach of typing out every exercise, embracing failure, and working through problems.
Book: Python Crash Course: A good go-to book for learning Python.
Website: The Python Tutorial: The official tutorial from the Python Project.