[ESS] General question aboout literate statistical practice pipeline

[Kevin Coombes]

There are (at least) two options. The more straightforward is to define and
describe the function when and where you need it. This option often starts
with something like "The next chunk defines a function to do
thus-and-such." And then typically follows a few sentences pointing out any
interesting or tricky bits of the implementation.

The other option is to "hide" the display of the function, in a named
chunk, at the point of definition. Then add an appendix where you can refer
to the named chunk to display the function body, after which you can go
into as much detailed documentation as you desire.

On Mon, Feb 3, 2025, 4:05 AM Hüsing, Johannes via ESS-help <
ess-help using r-project.org> wrote:

> Dear all,
> glad to notice that this list has become a bit more lively. I hope I won't
> derail the ongoing discussions by posing another, rather general question.
> The aim of literate programming seems to be to work on code and its
> documentation in parallel. One technique to achive this is to keep code and
> its documentation in close proximity.
> In literal statistical practice (LSP), the results of the statistical
> analysis becomes another part of the output. The output is normally a side
> effect of running the program, while in classical literate programming the
> code needn't necessarily be executed in the tangling process. The term LSP,
> introduced by Anthony Rossini, did not quite catch on but the principle is
> similar
> When I produce an analysis in the form of an LSP document, I tend to
> struggle with the arrangements of function definitions, which are normally
> documented for the underlying assumptions for my choice of model, and
> running the functions, which produces automatically generated
> documentation. When I keep definition and execution separated, the
> definition parts remain silent because nothing is "happening" yet. The
> narrative of what I am doing and why tends to fall apart.
> One could say that putting the analysis into a package would solve my
> problem, because the functions will be documented in separate *.Rd files
> while the actual analysis can be put into a vignette. This would address my
> problem but at the cost of the overhead that package productions brings
> with it (even if I steer clear of devtools::check()).
> Am I the only one around here or did anyone else encounter this problem?
> What is your generic pipeline for analyses done where methods are not
> run-of-the-mill?
>
> Hope I could get my problem across.
>
> Best
>
> Johannes
>
> PS sorry for any unrelated content my employer tacks to the end of this
> mail.
>
> Dr. Johannes Hüsing
> Epidemiologie
>
> Landeskrebsregister NRW gGmbH
> Gesundheitscampus 10
> 44801 Bochum
>
>
>
> T 0234 54509-216
> F 0234 54509-499
> Johannes.Huesing using krebsregister.nrw.de<mailto:
> Johannes.Huesing using krebsregister.nrw.de>
> www.landeskrebsregister.nrw<http://www.landeskrebsregister.nrw>
>
> Das Landeskrebsregister NRW online
> FACEBOOK: https://www.facebook.com/LKRNordrheinWestfalen/
> INSTAGRAM: https://www.instagram.com/landeskrebsregister_nrw
> LINKEDIN: https://www.linkedin.com/company/87452209/admin/feed/posts/
>
> Newsletter LKR NRW - jetzt anmelden unter
>  https://www.landeskrebsregister.nrw/aktuelles/newsletter
>
>
>
>
> Geschäftsführer
> Dr. Andres Schützendübel
> Vorsitzender der Gesellschafterversammlung
> Staatssekretär Matthias Heidmeier
> Sitz der Gesellschaft
> Bochum
> Registergericht
> Amtsgericht Bochum
> HRB 17715
>
> HINWEIS: Diese Nachricht ist nur für den Adressaten bestimmt. Es ist nicht
> erlaubt, diese Nachricht zu kopieren oder Dritten zugänglich zu machen.
> Sollten Sie irrtümlich diese Nachricht erhalten haben, bitte ich um Ihre
> Mitteilung per E-Mail oder unter der oben angegebenen Telefonnummer.
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> ESS-help using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/ess-help
>

	[[alternative HTML version deleted]]