A guide to reproducible research papers

A guide to reproducible research papers

[Ludovic Courtès]

Hello!

At long last, we’ve written a guide explaining why and more importantly
how (with step-by-step instructions) to provide reproducible
computational experiments accompanying research articles:

  https://hpc.guix.info/blog/2023/06/a-guide-to-reproducible-research-papers/

Please share!  :-)

Ludo’.

Re: A guide to reproducible research papers

[Giovanni Biscuolo]

Hello!

Ludovic Courtès <ludovic.courtes@inria.fr> writes:

> At long last, we’ve written a guide explaining why and more importantly
> how (with step-by-step instructions) to provide reproducible
> computational experiments accompanying research articles:
>
>   https://hpc.guix.info/blog/2023/06/a-guide-to-reproducible-research-papers/
>
> Please share!  :-)

thank you all for this great article, I'll share it à gogò!

You mentioned this citation from Jon Claerbout:

 Published documents are merely the advertisement of scholarship whereas
 the computer programs, input data, parameter values, etc. embody the
 scholarship itself.

and I was very curious about the source: I found that's a citation from
the abstract of «Making scientc computations reproducible» (Matthias
Schwab , Martin Karrenbach, Jon Claerbout, Published 2000) [1]

Nevertheless, the first occurrence of a similar statement by Jon
Claerbout is in a talk named «Seventeen years of super computing and
other problems in seismology» dated Oct 2 1994, precisely in the section
about "Technology transfer and research reproducibility" [2]

--8<---------------cut here---------------start------------->8---

In engineering, a published paper is an advertisement of scholarship but
the electronic document can be the scholarship itself. Forty years ago
data were "pencil marks on paper" and theory was some Greek
symbols. Then paper documents were adequate. No more. Now we need
electronic documents.

--8<---------------cut here---------------end--------------->8---

Just to add a little bit of history of computational reproducibility of
research.

Happy hacking! Gio'


[1] https://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.10.3712
complete PDF here: https://citeseerx.ist.psu.edu/doc/10.1.1.10.3712

[2] https://sepwww.stanford.edu/sep/jon/nrc.html#Technology%20transfer%20and%20research%20reproducibility

-- 
Giovanni Biscuolo

«Si può sperare
 Che il mondo torni a quote più normali».

Re: A guide to reproducible research papers

[Ludovic Courtès]

Ciao Giovanni,

Giovanni Biscuolo <giovanni@biscuolo.net> skribis:

> You mentioned this citation from Jon Claerbout:
>
>  Published documents are merely the advertisement of scholarship whereas
>  the computer programs, input data, parameter values, etc. embody the
>  scholarship itself.
>
> and I was very curious about the source: I found that's a citation from
> the abstract of «Making scientc computations reproducible» (Matthias
> Schwab , Martin Karrenbach, Jon Claerbout, Published 2000) [1]

Right, I found this one but couldn’t find precisely the original source
(I spent quite a while looking for it and eventually gave up).

> Nevertheless, the first occurrence of a similar statement by Jon
> Claerbout is in a talk named «Seventeen years of super computing and
> other problems in seismology» dated Oct 2 1994, precisely in the section
> about "Technology transfer and research reproducibility" [2]
>
>
> In engineering, a published paper is an advertisement of scholarship but
> the electronic document can be the scholarship itself. Forty years ago
> data were "pencil marks on paper" and theory was some Greek
> symbols. Then paper documents were adequate. No more. Now we need
> electronic documents.
>
> Just to add a little bit of history of computational reproducibility of
> research.

[...]

> [2] https://sepwww.stanford.edu/sep/jon/nrc.html#Technology%20transfer%20and%20research%20reproducibility

Excellent, thank you for sharing!

The takeaway here is that there’s nothing new: we’re probably going
further in implementing these ideas and adapting them to current
practices, but the core issue was already well documented 25+ years ago.

Ludo’.