publish in R Journal

[RobinHankin]

In the useR conference at Salzburg I caught up with Mark Van Der Loo and I told him about my frustration when trying to publish the spray package in R Journal and later JSS. I emailed him:

Mark

It was great to meet you in Salzburg the other day. I have been
pondering our conversation and have decided to take your advice and
consider submitting some work to the R Journal.

The first package I want to consider is the spray package. In the
interests of full disclosure, I should say that I submitted a
manuscript to R Journal back in 2017, when John Verzani was editor.
The manuscript was rejected; John seemed to indicate that he would
consider a resubmission. One of the reviewers suggested submitting
the work to JSS, which I did, and it eventually got rejected there for
different reasons.

Anyway, a lot has happened since then: spray is now a dependency for
two of my newer packages (weyl and stokes), and I also have new and
more interesting use-cases.

Can you advise whether submitting an R Journal article on spray would
likely be a good idea?

Best wishes

His reply:

Hi Robin.

I've been catching up on work since I came back from Salzburg, hence the slow reply.

I think that a package that allows for manipulaiton of multivariate polynomials is interesting for sure. I went through the vignette, but a paper would have to be a bit more systematic.

[I have some personal interest in this, as I have wriiten code to derive a Grobner basis in a ring of polynomials 15 years ago or so (its an 2^(2^n) algorithm, so it barely works with n=1 :-)). Back when I was studying Cox, Little and O'Shea's book on Ideals, varieties and algorithms, as well as some work of Sturmfels and Diaconis.]

In any case, I think it would be nice if the package would support the basic operations in the ring of polynomials, but also polynomial division where you get the result and the rest after division, because that would allow for a rich set of operations that people could reuse. Also, the package seems to have two purposes: sparse array computations and polynomial representation. I would choose one of those as a focus (probably the latter). Also, section 3.3 in the vignette seems to suggest that not all polynomials can be added, and that there are problems representing the null polynomial. That is something that would have to be fixed I guess.

Hope this helps!
Cheers,
mark