notes

  (

index

)

converting jupyter notebooks to org mode

introduction

use case

tests

It will be helpful to use a relatively simple jupyter notebook. I will use the ipynb for testing nbcorg and the jupytext markdown for testing pandoc conversion.

tools

pandoc

It might make more sense to just use pandoc from the command line; however, there is a package named org-pandoc-import that allows for importing these files directly from emacs. The doom config is

(package! org-pandoc-import
  :recipe (:host github
           :repo "tecosaur/org-pandoc-import"
           :files ("*.el" "filters" "preprocessors")))

After loading that you just open the ipynb and then run the function org-pandoc-import-to-org. This works!

nbcorg for nbconvert

reference :: nbcorg ยท PyPI

Make sure we have the package installed.

echo $HOME $0
pwd

/Users/crs58 /bin/sh
/Users/crs58/projects/notes/org

pip freeze | grep nbcorg
pip install nbcorg

Create directory and copy files for running the test.

PROJECT_DIR=~/projects/convert-jupyter-org
SOURCE_DIR=~/projects/applied-probabilistic-inference
mkdir -p $PROJECT_DIR
cp $SOURCE_DIR/plotting.* $PROJECT_DIR
ls $PROJECT_DIR


sh-3.2$ sh-3.2$ sh-3.2$ plotting.ipynb	plotting.md	plotting.org	plotting.py

echo $PROJECT_DIR
echo $SOURCE_DIR

/Users/crs58/projects/convert-jupyter-org
/Users/crs58/projects/applied-probabilistic-inference

cd $PROJECT_DIR
jupyter nbconvert --to orgmode --OrgmodeExporter.src_block_options=':session plotting :results output replace' --OrgmodeExporter.exclude_output=True plotting.ipynb
# jupyter nbconvert --version

Good to know that this is essentially broken for now due to

File "/Users/crs58/.asdf/installs/python/3.8.2/lib/python3.8/site-packages/jinja2/loaders.py", line 429, in load
  raise TemplateNotFound(name)
jinja2.exceptions.TemplateNotFound: base.tpl