Activity: Working with python environments

Background

We have learned in the lecture that python environments allow us to maintain different python installations on the same system that contain all the python packages that we need for our work, while also maintaining potential conflicts between packages.

It is good practice to have one environment per project or task and to share this environment with your code to ensure reproducibility of your work.

Learning Goals

After this practice you should be able to

Create a new virtual environment
Activate an environment
Install a package into a pre-existing python environment
Run python code using a specific environment
Update the environment using a .yml file.

Task

Create a new environment named ISAT_420_S26
1. I recommend trying this on the command line.
  - For Windows: Go the the start menu and open the Anaconda Powershell Prompt
  - For Apple: Open the Terminal in launcher and enter the neccesary commands
Activate the environment and install at least the following packages: jupyter, xarray, netcdf4, matplotlib
- Note: For some Anaconda installations (i.e. the Windows Version), the direct installation of these packages fails. This is because, they are distributed through the conda-forge channel.
  You can check which channels are available like this:
```
conda config --show channels
```
  If not already added you can add the conda-forge channel like this:
```
conda config --add default_channels conda-forge
```
Launch a jupyter notebook to test whether your installation works by running the notebook in the Code directory: w7_environment_test.ipynb
(Optional): Use the provided environment.yml file to update your environment.
- Note: This requires you to have the environment.yml file in your current working directory.
  - Here is a quick guide on how to navigate directories within the Terminal: Linux Command - Shell Tutorial.
    - This is a bit too much information, but Windows Powershell which is running in the windows Terminal and bash which is running on Mac and Linux are two flavours of different kinds of interactive command interpreters (also known as Command Line Interface). They differ in how commands are written but have sufficient overlap in the basics, that it won’t matter for us.
1. If you try to update the environment, conda will be thinking for a while and try to find a combination of packages without conflicts in their dependencies. This might fail!
2. If it fails, you can try to update the environment using the environment_installed_packages.yml. This environment file contains a list of the packages that I directly installed without all their dependencies. This means it will take longer to update and you won’t have exactly the same versions for each module, but it should work.
3. If that also fails, you can manually install all the packages listed in the environment_installed_packages.yml. You can install multiple packages together. For example the command below will install numpy and pandas:
```
conda install numpy pandas
```

Learning Checklist

I understand that using environments is beneficial on the long term, because it allows me to experiment with additional packages, without being afraid of breaking anything
I am able to create new conda environments, activate them, and switch between them.
I understand that packages installed with conda install ... are always installed in the environment which is currently active. (base) is like any other environment: it is simply the default one.
I am able to install new packages using conda.
I have used the environment.yml file to create a new environment for this course that contains all the packages that I need right now.
I know that I need to execute all code for this class using this newly created environment