Proposal/question/suggestion ?

Symbolics commented 2 years ago

CLJ has solved one of the biggest problems in reproducible research with the notebooks, and I think anyone working in the small intersection of data analysis and common lisp is grateful, so let me say thank you.

The one area I still struggle with is document generation. Something like R Markdown. Ironically, R Markdown (well, knitr) supports quite a few other languages, including Python and Julia, but sadly not Common Lisp.

I thought I'd open this issue here because of all the projects I'm aware of in the common lisp ecosystem, this seems closest to embedding lisp as an 'engine' in a foreign language system, and rather close in terms of function too.

So, I have been wondering:

Could any of the code/techniques used in CLJ be used to add Common Lisp as a custom language engine in Knitr?
Are you aware of any alternative to R-Markdown for statistical document preparation? I believe org-mode might have something, but it seems a shame to 'waste' all the good work that goes into R Markdown.

yitzchak commented 2 years ago

For the first question see my project metys. Its like knitr but can run multiple jupyter kernels. I have not used it for while so beware.

Symbolics commented 2 years ago

Thanks. If you like, I'll leave this open for now and update when I settle on a solution for preparing statistical analysis docs. I like the TeX/LaTeX approach in metys, but worry a bit about it's lack of usage anywhere. At least with knitr, there's no shortage of resources for help.

yitzchak commented 2 years ago

metys isn't really LaTeX specific. You can use it on top of any plain text document like MarkDown.

yitzchak / common-lisp-jupyter

Proposal/question/suggestion ? #90