How to publish R packages in good journals

R packages for researchers

Photo by Markus Spiske on Unsplash

In earlier posts, we have discussed the procedure to develop an R package and its importance in computational research. The R packages are a great tool to create reproducible research. Thousands of such R packages are been contributed to the CRAN, nowadays. A similar thing is observed in Python and other languages such as Matlab. On my own experience, I can say, it’s really difficult and challenging to make your package/s popular and raise the number of downloads unless you have a great team and follower who can share it with the targeted audience.

Source: https://www.r-project.org/about.html

I found, publishing such packages in reputed journals is a great option to bring your R package in the traditional research and make researchers try-on. It is true that the R package itself is the citable entity, but making available it in publication format has several benefits, such as:

1. While following the publishing procedure, the package gets reviewed by several reviewers and editors, which ensure the usefulness and reliability of the package.
2. The published version of the package increases the confidence of its users while using it.
3. Out of hundreds of available packages, the published ones are experiencing more downloads and citations.
4. The packages published in journals attract traditional researchers (who are not expert in Coding and developing the package), which eventually increase the users/viewers of the packages.

Now, it is worth noting that publishing an R package is not an easy task, as one can imagine. The traditional journals usually reject to review the package related manuscript with the very common comment that it is out of the scope of the journal. For the R packages, the R community has a great platform journal, the R journal. It is targeted only for the R packages, its introduction, and demonstration, with the minimum mathematical derivations. But still, it’s challenging to publish all kind of R packages in this journal, since it prefers the articles which are related to statistics and machine learning with significant contributions in computational capabilities.

Hence, it is very important to know the possible journals to publish R packages with higher chances of acceptance. The following is the list of journals in several domains:

1. The R Journal
2. Journal of Statistical Software
3. The Journal of Open Source Software
4. SoftwareX
5. Nature Toolbox
6. The Stata Journal
7. PeerJ Computer Science
8. Journal of Machine Learning Research
9. Computing and Software for Big Science
10. Environmental Modelling & Software
11. Bioinformatics
12. Computational Statistics
13. ACM Transactions on Mathematical Software
14. Information and Software Technology
15. Neurocomputing (Software track)
16. Journal of Systems and Software
17. Many more…

Apart from these journals, there are several domain-specific journals which have published several R packages. Some of them are as follows:

1. International Journal of Forecasting
2. IEEE Access
3. Energies
4. Methods in Ecology and Evolution
5. PLoS ONE
6. And, many more…

A good review of such journals for several domains such as Physical Sciences, Geosciences, Engineering, Humanities, Social Sciences, Image Processing, Informatics, Mathematics, Statistics, and Life Sciences is presented here.

It is worth noting that publishing an R package is a very challenging task and the package must hold significant contributions in the domain. If it is to be published in journals like the R Journal or the Journal Of Statistical Software, which are dedicated to software packages, the manuscript draft should be focused on introduction, description, and examples of the package. Whereas, if you are targeting a domain-specific journal with traditional scopes, such as IEEE Access or Energies, it is essential to discuss some case studies in the manuscript which can prove how your package can be useful in that domain (an example). This is very important to impress the editor of such a journal so that he should not reject your manuscript with ‘Out of the Scope’ stamp.

Additionally, it is crucial to provide the link of your R package along with GitHub code folders. It will be helpful to locate your package and can improve the possibility of package reuse. Further, the R packages allow you to provide the citable link for the package, where you can replace the existing CRAN BibTeX with the published journal BibTeX. For example, the R allows you to know the possible citation of an R package with the following syntax.

In default form, it will show you something like the following:

But, if you update the description of the package with a new citation with the revised BibTeX, you can find it here:

This ensures the proper citation for your package’s publication and the better spread of the package. There is another trick, that you can put the published journal article of your package at the Vignette of the package, provided you must be clear with the sharing policies of the journal. There won’t be any issue with the articles published with open access policies.

Enjoy developing and publishing.

Author:

Dr. Neeraj Dhanraj Bokde

Postdoctoral Researcher,

Aarhus University, Denmark

https://www.researchgate.net/profile/Neeraj_Bokde