# Should one cite a LaTeX package in a scientific publication?

R world has the convention of citing heavily used packages in a scientific publication (see for example this). You can get the reference to be used with citation() command. For example:

citation("base")

To cite R in publications use:

R Core Team (2013). R: A language and environment for statistical computing.
R Foundation for Statistical Computing, Vienna, Austria. URL
http://www.R-project.org/.

A BibTeX entry for LaTeX users is

@Manual{,
title = {R: A Language and Environment for Statistical Computing},
author = {{R Core Team}},
organization = {R Foundation for Statistical Computing},
year = {2013},
url = {http://www.R-project.org/},
}

We have invested a lot of time and effort in creating R, please cite it when
packages.


I am not familiar with LaTeX conventions. Should you cite a heavily used LaTeX package in a scientific publication? Say that I have for example used pgfplotstable to make table that would have been impossible/very difficult to make otherwise. If the answer is 'Yes', where can I find the format the author of a package prefers?

-
This is an example (slightly different from citing a package manual: the authors also wrote a paper for TUGboat, hence one can cite the paper). –  Claudio Fiandrino Jan 27 at 13:46
this is a tough one. i'd have to say, in general, no. R is a tool for data analysis, which is what a scientific publication is about. most latex packages are intended for data presentation, so although they may be invaluable, they're not at the same level as an analytical tool. that said, some authors credit (la)tex on the copyright page of books, or in acknowledgments. and if a publication is meta-scientific, say on how to present data effectively, then a credit or citation would be appropriate. –  barbara beeton Jan 27 at 13:50
As @barbarabeeton says, you usually cite things that are required to reproduce stuff, which may include data analysis. Even that depends, as general number crunching is not dependent on a specific program. –  Joseph Wright Jan 27 at 13:57
In a scientific work, if the results depend of how the data are analyzed, the R packages are an important part of the material and methods that must be cited so that anyone could reproduce this work. How are typed is irrelevant from the scientific point of view, so citing a software for this task is only a matter of courtesy, or in the worse case, of license requirements (CC BY, for example). –  Fran Jan 27 at 14:19
@Fran I disagree. Why must R be cited for reproducibility? Taking the same code with the same parameters and the same bugs is not properly reproducing. Describe the mathematics, and whether the maths is implemented in R, Matlab, Python, or your grandfathers abacus, is an irrelevant or potentially dangerous detail. I will acknowledge my tools, but in the acknowledgements, not in the methodology. We used function x in package y is not a scientific way to describe a methodology. But this discussion is off-topic on TeX - LaTeX and belongs on Academia. –  gerrit Jan 28 at 9:57

this is a tough one. i'd have to say, in general, no.

R is a tool for data analysis, which is what a scientific publication is about. most latex packages are intended for data presentation, so although they may be invaluable, they're not at the same level as an analytical tool.

that said, some authors credit (la)tex on the copyright page of books, or in acknowledgments. and if a publication is meta-scientific, say on how to present data effectively, then a credit or citation would be appropriate.

-
Thanks for the help! I approve the logic behind the statement that LaTeX packages should not be cited as long as they do not change the numbers behind the representation (i.e. the work can be reproduced). Following the same logic: you should not cite say ggplot2 unless you use the stat_smooth() function, which does number crunching that might be hard to reproduce. Instead one could consider placing these packages to acknowledgements. –  Mikko Jan 27 at 14:59

Citations and references are for sources that have contributed to the content of a text (which can include the statistical analysis).

While it is not necessary or expected that you cite the document preparation software used, this kind of information is traditionally included in the colophon, a little summary of the typography used for a publication. (Only applicable to books or other stand-alone works, as far as I know). If you're excited about the ease and quality that the use of LaTeX (and/or particular packages) has made possible, you can acknowledge it by adding a little blurb in the back of the book. (Wikipedia says the colophon usually appears in the front nowadays; I guess I'm just old-fashioned).

But short works like articles have no place for a colophon. For those, some people mention LaTeX in the acknowledgements (which, depending on the text, might appear in a footnote or in a short section at the start or end of the article.) Again, there's traditional precedent: Authors used to thank the secretary who typed numberless versions of their manuscript; and you'll still see novelists thank their editor, for example.

-
The problem is that most works don't have a colophon :-/ –  tohecz Jan 27 at 16:56
@tohecz But for books and things, you often can include one. For example, I do this in materials I produce for teaching. Sometimes I also specify my editor and OS. (Whatever makes for a rounded statement.) And then say something about the software being free... –  cfr Jan 28 at 2:11
But yeah, good point. For texts that do not stand alone, there's no place for a colophon. In that case the acknowledgements are sometimes used instead. Added it to the answer. –  alexis Jan 28 at 12:19

Short answer. No, you don't need to do that, but you can.

A bit longer answer. In general, it might be a good idea to mention in which software you draw your figures, so for instance I tend to cite TikZ recently in my articles. There seem to be no exactly standard way of doing it. I've ended with this .bib entry:

@manual {tikz,
key = {Ti\textit{k}Z},
organization = {Till Tantau et al.},
title = {Ti\textit{k}{Z} \& {PGF} (version 2.10)},
note = {\url{http://sourceforge.net/projects/pgf}},
year = {2010},
}


(You might have to remove the key, depending on your bibliography style.) With amsplain this gets displayed as

Till Tantau et al., \emph{Ti\textit{k}{Z} \& {PGF} (version 2.10)}, 2010,
\url{http://sourceforge.net/projects/pgf}.


which is quite fine.

-