Convert jupyter notebooks to python files

[larshinueber]

This PR extends and updates the create_scripts.py utility to fix #22

Major changes include:

• Use custom nbconvert template to simplify (and largely remove) conversion logic from python script. Fixes Python script creation removes import statements #53 and most issues in Automated conversion between .py and .ipynb #22
• Make create_scripts.py a CLI app, without requests to the user in the command line during the conversion
• Add plt.show() if it is not added in the notebook, fixes Show plots in auto-generated Python files #52
• Add workflow documentation in README

The script was applied to all notebooks in the repository, no python file was modified afterwards. No more syntax errors are present.

Before merging

Please check if the generated python files are as desired :)
If you want to try the script yourself, make sure to use the tudat-examples conda environment from environment.yaml, specifically the nbconvert version specified there, as there were some changes that are not backwards compatible.

[DominicDirkx]

This all looks good to me. For the environment.yaml, is it important for all the package versions here:
https://github.com/tudat-team/tudatpy-examples/blob/master/environment.yaml
to be fixed? Ideally, we would have the conda environment we use for the examples to just be the same one people use for their tudat applications (linked to here: https://docs.tudat.space/en/latest/_src_getting_started/installation.html).

[larshinueber]

Great :)

Most of the package versions do not need to be fixed, I added them such that the example outputs are reproducible. Of course, with a new tudatpy version these would then have to be updated.
The only inherently important versions to be fixed are the boost-cpp and pygmo versions (will open a new issue from the Brightspace post of P&O), the nbconvert version due to the breaking changes across versions and the astroquery version, as per tudat-team/tudat-bundle#22.

I agree that having one single environment would be nice. On the other hand you could argue that the tudatpy-examples toolchain has some dependencies that a regular user probably won't be using, like nbconvert and tqdm (although tqdm is more a nice-to-have than a must), so including them in the main tudat-space environment.yaml seems a bit out of place?

[DominicDirkx]

Looks good, thanks. Doing the conversion between .ipynb and .py indeed shouldn't be that relevant for the average user.