Atom/Hydrogen or VSCode/Python allows creating a python files split into cells with # %% separators with the ability to run cells via backend Jupyter session and interactively show results back.
More examples: Jupyter Python VSCode examples, Atom/Hydrogen Getting Started.
colab-convert python module converts files: .ipynb to .py and .py to .ipynb.
colab-convert is a fork of the ipynb-py-convert.
conda install -c defaults -c conda-forge colab-convertor
pip install colab-convert- If encoding problems on Windows try using
python>=3.7, settingset PYTHONUTF8=1in Windows console and usecolab-convertfor UTF-8 files only. If using Git-Bash on Windows setting:
export LANG=C.UTF-8
export PYTHONIOENCODING=utf-8
export PYTHONUTF8=1should be enough. Also try setting default Bash settings to UTF-8: [Options] - [Text] - [Locale / Character set] - [C / UTF-8]. It might affect all Bash runs so there would be no need to setting encoding every time.
colab-convert examples/plot.py examples/plot.ipynb
or
colab-convert examples/plot.ipynb examples/plot.py
VSCode
Markdown cells are converted to python multiline strings '''. Code cells are left as is. # %% is used by vscode as the cell marker on which 'Run Cell' action is available.
Metadata is converted from notebooks into .py and vise versa using # !! to denote the meta data lines in the .py files
Jupyter ipynb notebook

