I have developed a routine in R for longitudinal analysis of networks. I was wondering whether it can be published on an academic journal as a dedicated article?
-
6What do you mean by "whether it can be published"? No mainstream academic journal would publish an article that consists of literally nothing but R code (with no accompanying text at all), so you can't just publish the program. On the other hand, there's no reason why you can't publish an article that describes your methods and results and is accompanied by the code as supplementary information, assuming you've done something innovative and interesting.– Anonymous MathematicianMay 30, 2014 at 2:32
-
1Yes, I meant with text describing it and an example of application.– DanieleMay 30, 2014 at 3:18
-
1Possibly helpful: scicomp.stackexchange.com/questions/660/…– David KetchesonMay 30, 2014 at 4:13
-
2It'd be easier to publish your code if you have access to a time machine: The Incomplete Gamma Integral– Mad JackMay 30, 2014 at 4:14
-
1Yes, but that doesn't mean it should be. See comments here: plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0085047– mankoffMay 30, 2014 at 11:33
|
Show 1 more comment
2 Answers
Depending on the extent of your work, you may be able to get a paper describing your package and methods in PLoS One or a similar journal, or the Journal of Statistical Software.
Alternately, The R Journal is a peer-reviewed journal covering R software, which may be an appropriate venue if you don't have enough methodological material to build out a full paper for another journal
Finally, many journals accept code supplements for papers describing the use of your method, so you may be able to publish your routine's in a paper about the actual work you are doing.
-
2Link to the Journal of statistical software (which is what I assume you are referring to) jstatsoft.org. They publish "code snippets". Of course some exposition will be required. May 30, 2014 at 5:51
-
3Be wary of publishing code in PLoS One. They might accept it: plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0085047– mankoffMay 30, 2014 at 11:34
-
1@mankoff Not sure what your point is. Is it that the article in question is not good? May 30, 2014 at 12:17
-
+1, This question has a related set of advice, Is it ethical to publish a paper about porting an existing software package to another programming language? (although IMO is not a duplicate). Methodology papers are perfectly acceptable in field specific journals as well. For social scientists I'm familiar with the journal Social Networks, although there are surely many others to chose from.– Andy WMay 30, 2014 at 12:58
-
1@FaheemMitha The PLoS One article linked to might as well be called "Reinventing the Wheel". It takes a known, common example and repackages it as new.– FomiteMay 30, 2014 at 14:52
Publication of an article about R code (in package form) is definitely a possibility. Whether your particular code is publishable is something for the reviewers to judge. There is always The R Journal, which is peer-reviewed and read by most serious R programmers. Journal of Statistical Software is a more general outlet that is not R-specific.
I see a lot of projects where a substantive article is published in a disciplinary journal (perhaps describing the algorithm or using software in a particular application) and then an accompanying piece describing the software specifically is published in JSS or The R Journal.
-
Definitely publishable is too strong (I would use the word potentially in this circumstance given the limited amount of information the OP provided). It still needs to make a unique contribution that is necessary of all research articles, and uploading to the CRAN servers is not a measure of such a unique contribution.– Andy WMay 30, 2014 at 13:00
-