Anaconda is a software-company which offers various tools to perform data science tasks on a single machine. Their popular open-source Anaconda Individual Edition distribution includes

• Python and R,
• the package manager conda to install packages
• as well as over 250 preinstalled packages.

Note that Anaconda is quite large -the base install requires around 2GB- and we usually only use a small fraction of the preinstalled packages. Therefore, we will install the lightweight version of Anaconda, called Miniconda.

Miniconda is a free minimal installer for conda. It is a small, bootstrap version of Anaconda that includes only conda, Python, the packages they depend on, and a small number of other useful packages.

• You should be familiar with the shell. If not, take a look at this tutorial.
• If you have an old Anaconda version installed on your system, you may uninstall it first to avoid compatibility problems.

Uninstall Anaconda

You can skip this section if you don't already have Anaconda installed on your machine.

Windows

1. Open the file explorer.
2. Delete your environment (anaconda3\envs) and package (anaconda3\pkgs) folders in your user folder.
3. Open Add or remove programs and uninstall your Anaconda installation.

macOS

Open your terminal (learn how to open your terminal)

Anaconda is usually located in your home directory inside a folder called "opt". You can remove your entire Anaconda directory by typing the following command in your terminal:

rm -rf ~/opt/anaconda3

Windows

• Go to the site latest Miniconda installer links:
• Choose the appropriate Windows version (usually 64-bit) and download the file
• Go to your Downloads folder, right-click on the file and choose "as Administrator"
• Install the software
• During the installation process, choose: "Just Me"

MacOS with Intel

• Download the Graphical Installer and install the software: Miniconda3 macOS Intel x86 64-bit pkg
• During the installation process, choose: "Just Me"

MacOS with M1

• Download the software: Miniconda3 macOS Apple M1 ARM 64-bit bash
• Open a terminal and cd into the directory where the file is located.
• In your terminal:

sh ./Miniconda3-py310_23.1.0-1-MacOSX-arm64.sh

• Scroll to the bottom and type yesto install the software.
• If asked, agree to initialize Miniconda

During the installation process, Miniconda created the so called base environment. Let`s take a look at this environment:

• On Windows open the Start menu and open "Anaconda Powershell Prompt" or "Anaconda Prompt".
• On macOS or Linux open a terminal window.

Usually, the base environment is already activated (and you can see the word base in your shell).

Update your version of conda by running:

conda update -n base -c defaults conda

When conda asks you to proceed (proceed ([y]/n)?), type y.

The conda team, from Anaconda, Inc., provides various packages to all users free of charge in their defaults channel. However, to get more up-to-date modules, it's better to use the popular community-led alternative conda-forge.

Type this in your shell to add the conda-forge channel to conda:

conda config --add channels conda-forge

Then make conda-forge the priority channel:

conda config --set channel_priority strict

This ensures that conda will always try to install packages from the conda-forge channel first instead of other channels.

In the next step, we install some Python module with conda.

Anaconda's package manager conda makes it easy to manage multiple data environments that can be maintained and run separately without interference from each other in so called virtual environments. A virtual environment in Python is a separate and isolated environment where you can install packages and dependencies without affecting other Python installations or the system's environment.

Installing Python modules in a virtual environment is recommended for several reasons:

• Isolation: By using virtual environments, you can isolate your project's dependencies from the system-wide Python installation. This means that you can have different versions of the same library for different projects without causing conflicts.
• Reproducibility: By specifying dependencies in a virtual environment, you can ensure that the project will always run with the same versions of libraries, regardless of system-wide installations or updates.
• Compatibility: Different projects may have different requirements, and installing modules in a virtual environment prevents compatibility issues between projects.
• Easy Management: Virtual environments make it easier to manage dependencies, as you can easily install, update or remove them without affecting other projects or the system.

If you install a new module with conda, it analyses the current environment including everything currently installed, and, together with any version limitations specified (e.g. the user may wish to have a package in version 2.0 or higher), works out how to install a compatible set of dependencies, and shows a warning if this cannot be done.

You can take a look at all available packages on this website: https://anaconda.org/.

• Let's create a new environment:
  • Call the new environment ds (short for data science)
  • Use Python 3.10
  • Install some packages (pandas etc.)

conda create --name ds python=3.10 pandas scikit-learn altair vega vega_datasets

When conda asks you to proceed (proceed ([y]/n)?), type y.

• You can activate the new environment as follows:

conda activate ds

You are now inside your virtual environment ds (you should notice that base is replaced by ds in your shell)

• Take a look at all the modules in your ds environment:

conda list

You should see a list of modules with their name, version, build (more detaild information about the package) and channel (from which the packages were installed).

• If you want to switch back to your base environment simply type:

conda deactivate

If you want to update a specific Anaconda environment (this will update all packages in the selected environment to the latest version but will not update Python), use this command: conda update --all

Make sure to activate the right environment first. In our example, we use the base environment so we don't have to activate it

• Update all

conda update --all

• If you only want to update specific modules, use conda update and the name of the module (e.g., scikit-learn)

conda update scikit-learn

• Let's make sure that you use the machine learning module scikit-learn in version 1.2.1 or higher:

conda list scikit-learn

• It's also possible to install a specific version of a module:

conda install -c anaconda scikit-learn=1.2.1

Congratulations! You have completed the tutorial and learned how to:

✅ install Miniconda

✅ use conda-forge

✅ create a virtual environment

✅ install modules

✅ update modules

Thank you for participating in this tutorial. If you found any issues along the way I'd appreciate it if you'd raise them by clicking the Report a mistake button at the bottom left of this site.

Copyright: Jan Kirenz (2023), kirenz.com, Creative Commons Attribution-NonCommercial 2.0 Generic (CC BY-NC 2.0) License