Conda ecosystem#
Conda is a language-agnostic, cross-platform package management ecosystem. conda-forge is a community-led infrastructure for building and distributing packages. These projects provide convenient ways to install and use packages across different platforms.
Important
We recommend avoiding the historical Anaconda distribution, as it can lead to technical and legal issues (see Conda & the conda-forge project). If Anaconda is installed on your computer, it would be advisable to uninstall it and use the alternative solutions instead.
Pixi#
Pixi is a modern package management tool for developers based on the Conda ecosystem. To install Pixi, follow one of these methods:
powershell -ExecutionPolicy ByPass -c "irm -useb https://pixi.sh/install.ps1 | iex"
Download and execute the official installer.
curl -fsSL https://pixi.sh/install.sh | sh
Once the installation is complete, open a new terminal and verify that the pixi command
is available:
pixi --version
Note
For more details about the installation, see the Pixi documentation.
The conda and mamba programs#
Note
The installation of conda and mamba is not mandatory for this training, as Pixi is
sufficient for our needs and is overall a better and more modern tool. However, it is
still beneficial to have these applications installed, since many online examples and
tutorials use these programs.
One can use Pixi to globally install the command-line applications conda and mamba:
pixi global install -e env-conda conda mamba
conda init --all
conda config --set auto_activate false
You should then be able to run:
conda --version
mamba --version
An issue with this installation method?
One can also install conda and mamba with Miniforge, which is a modified version of
Miniconda that uses the
community-driven conda-forge channel by default (we will discuss these projects in more
detail during the training; see Conda & the conda-forge project). Install
Miniforge by following the
instructions presented on this page
(open this link in a new tab).
Danger
If you see (base) when you open a new terminal, this indicates a problematic
configuration! To resolve this issue, run the following command:
conda config --set auto_activate false
After running this command, you can verify that (base) is no longer present when
opening a new terminal.