jupytext-python.py

# ---
# jupyter:
#   celltoolbar: Slideshow
#   jupytext:
#     cell_metadata_filter: all
#     cell_metadata_json: true
#     formats: py:percent
#     notebook_metadata_filter: all,-language_info,-jupytext.text_representation.jupytext_version
#     text_representation:
#       extension: .py
#       format_name: percent
#       format_version: '1.3'
#   kernelspec:
#     display_name: Python 3
#     language: python
#     name: python3
#   rise:
#     autolaunch: true
# ---

# %% [markdown] {"slideshow": {"slide_type": "slide"}}
# # using jupytext

# %% [markdown]
# if you have jupytext enabled, you can use RISE in a usual manner  
# on notebooks that are stored as `.py` or `.md` (or other extensions, for that matter).

# %% [markdown] {"slideshow": {"slide_type": "slide"}}
# ## exact same behaviour

# %% [markdown]
# there is almost nothing that changes in this case as far as RISE is concerned

# %% [markdown]
# the only notable difference is for locating the notebook-specific CSS file

# %% [markdown]
# of course, and as you might expect, if your notebook   
# is called either `mynotebook.py` or `mynotebook.md`,  
# or, here again, any other extension  
# then it is `mynotebook.css` that is used,  
# if it exists, to load a notebook-specific CSS

# %% [markdown] {"slideshow": {"slide_type": "slide"}}
# ## pros and cons

# %% [markdown]
# jupytext is supercool if you use git a lot, and you don't carre about saving cell outputs  
# no need anymore to run `nbstripout` all the f... time

# %% {"cell_style": "split"}
# you can still embed code of course
def syracuse(n):
    while n != 1: 
        if n % 2 == 0:
            n //= 2
            yield n
        else:
            n = 3*n + 1
            yield n


# %% {"cell_style": "split"}
# but the output is no longer stored
for n in (4, 8, 27):
    print(f'n=${n} :', end=' ')
    for i in syracuse(n):
        print(i, end=' ')
    print()