flotpython exos: data-science

Thierry Parmentelat

this repo contains exercises and TPs that complement the Python MOOC

excluded are the auto-corrected exercises mentioned in the MOOC these are bundled in the main course repo

in the present repo we try to gather all the other, generally more informal, material for practising the Data-Science ecosystem (for stuff about pure Python, see the other repo https://github.com/flotpython/exos-python)

there is no runtime tool here !¶

also note that, as opposed to the auto-corrected exercises mentioned above, there is a deliberate choice to not provide a notebooks infrastructure
this is because we want our students to become autonomous, so it means they are supposed to solve all these problems on their own laptop, where they are expected to have acquired the skills for installing and managing a decent software stack (typically bash + vscode + python + ipython + jupyter)
the fact that most of the material is written as a notebook is mostly a convenience, both for authoring (outputs are up-to-date), and of course in cases where the starting material is a notebook itself

contents¶

the material is organized along these rather vague categories:

• exos: short, simple one-shot assignments

• tps: more elaborate assignments, with several steps, that let students achieve something

• howtos: more for reading than for practising, that can be recipes to achieve some common tasks

as well as, less interesting probably, some low-order categories like samples, reading, quizzes, etc.

historical note: in an older version, this repo contained material about both pure Python and the Data Science tools; it has been split in two to ease its maintenance - github slugs are now flotpython/exos-python and flotpython/exos-ds respectively

formats & jupytext¶

as noted above, most of the contents is written as notebooks; all notebooks are jupytext-encoded using either py:percent or md:myst formats
you will need to pip install jupytext to be able to read those as notebooks
also all notebooks have their filename prefix ending in -nb to help the distinction between notebooks and pure Python or pure markdown

note on autoreload in ipython or notebooks¶

if you use IPython or Jupyter on your laptop, make sure to read
https://intro.info-mines.paris/1-1-installations/#-configuration-de-lautoreload

it’s in French, but gives you the recipe to get IPython and the notebooks to play along if you’re doing module development (otherwise you’ll be bitten by module caching)