Wondering if energy-dependent convolution code is in scope for pyOpenSci

[pibion]

Hello,

My team and I (especially @Arqu1100) have been working on energy-dependent convolutions for a nuclear physics application, and we are wondering if the code would be suitable for pyOpenSci review: https://github.com/det-lab/energyDependentColvolve.

We don't think this is suitable for JOSS because it is a single-function library.

However, the code is useful within the nuclear physics community and potentially the astrophysics/imaging community. We ran into quite a few issues when writing the code and would like to help out other groups who need to do this "simple" calculation. One of the strengths of this code base is the inclusion of test cases. The only existing library we found that does what we do here is varconvolve, which we haven't been able to verify against our test cases. Other examples of code that does a similar job are referenced on Stack Overflow as being deeply embedded into existing libraries (https://stackoverflow.com/questions/18624005/how-do-i-perform-a-convolution-in-python-with-a-variable-width-gaussian).

[coatless]

@pibion please resubmit following the pre-submission template guidelines.

https://github.com/pyOpenSci/software-submission/issues/new?template=presubmission-inquiry.md

[jonas-eschle]

@pibion did you reach out to domain-specific packages whether they are interested to incorporate the function? This seems a lot more promising to me

[pibion]

@jonas-eschle we have not - we'll do so.

[pibion]

@jonas-eschle we've reached out to AstroPy, Numpy, and SciPy and it looks like the answer is no for each - this isn't quite in scope for AstroPy (and there is an existing package in their community that does this work!). Numpy is very conservative about adding new features, and there wasn't any conclusive response from SciPy.

Would this package as a standalone be suitable for PyOpenSci or is it too small?

[pibion]

@coatless @jonas-eschle also opened a new issue: #248

[jonas-eschle]

@jonas-eschle we've reached out to AstroPy, Numpy, and SciPy and it looks like the answer is no for each - this isn't quite in scope for AstroPy (and there is an existing package in their community that does this work!).

Thanks for reaching out to them! When you say there is a package that does this, the immediate thought would be, what does this package (energyDependentConvolution) add to the community? What can it do that the other package can't?
(please don't take this the wrong way that I am asking critically: software is easily built but hard to maintain. Buildinging is 20%, the rest that comes more like 80%, and if two packages do the same, that means time is lost which could be spent on one. So to make sure a package will survive, it also needs a niche to fit in and not duplicate existing functionality)

Numpy is very conservative about adding new features, and there wasn't any conclusive response from SciPy.

Could have guessed this one, but makes sense.

Would this package as a standalone be suitable for PyOpenSci or is it too small?

I think if it is a standalone package (and fills a niche), there is nothing that would be too small, as long as it is a standalone package and fulfills the requirements (setup, tests, docs, tutorials, etc.). The effort for a little package to set these things up is just proportionally higher.