Replace Anaconda

[carschno]

In the current setup instruction, we link to the Anaconda installation page: https://www.anaconda.com/products/distribution This currently redirects to https://www.anaconda.com/download which requires providing an email to get to the software. This forms a privacy-related obstacle. A free/open that approach would be preferable.

Alternative Options

Update: I have implemented a version using pip/env in #513. Given that the installation remains flawless and that it does not require installing additional tools, I find this to be the preferred option.

@colinsauze @psteinb Do you have any opinion on this issue? I personally prefer Poetry, but also see an advantage in using built-in tooling (venv). I do not see a reason for still using Conda, but it is difficult to generalize this. Is there a general discussion about Conda in Carpentries?

Miniconda

Advantages:

• Apart from the Anaconda installation, the installation works in the same way as now
• Does (currently) not require (privacy-related) extra step for downloading

Disadvantages:

• License is the same as for Anaconda (free for non-commercial use)
• Conditions for download and for usage can change in the future

Pip and venv

Advantages:

• Comes with Python, no installation required
• Established practice, probably already known by many learners
• Free and open license

Disadvantages:

• Requires update in the lesson
• Additional cognitive load: requires learners to create an environment and to use it consistently

Poetry

Advantages:

• Easy installation
• Good at resolving dependency issues (in isolated environment)
• Free and open license
• Popular and established, might already be known by some learners

Disadvantages:

• Requires update in the lesson
• Cognitive load: requires learners to install and use new tool Poetry

  
  ### Tasks

• [x] Replace or remove "Installing Python using Anaconda" block
• [x] Adapt "Installing the required packages" block
• [x] Double-check or remove "Python package installation troubleshooting" block
• [x] Check if Jupyter Lab uses the correct environment/Python kernel

[carschno]

I'm looking into the Pip/Venv version for starters.

[carschno]

The pip/venv solution seems very easy, there is no longer a need for additional tools. I have double-checked every step of the MacOS/Linux solution.