JOSS pre-submission enquiry

[JoshKarpel]

Project repository URL: https://github.com/JoshKarpel/idesolver

My software is a Python package for solving integro-differential equations of a fairly general form (see the proto-documentation in the repositories for more details). The algorithm is based on a technique described here. This kind of equation tends to pop up when dealing with systems whose current behavior depends on the history of their behavior. This package was written when my own research (in physics, on numerical quantum mechanics) needed the numerical solution of such an equation. It uses tools from numpy/scipy to implement the technique described in the paper above.

I want to check whether I satisfy the "Should be a significant contribution to the available open source software..." requirement. In particular, I'm concerned that my own contribution is not significant enough, since the technique is not novel. My own contributions are that I

• Wrote the Python package based on the technique from the paper linked above, which provided a sketch of how to implement the technique in MATLAB.
• Extended it to work with complex-valued integro-differential equations. This was a implementation improvement, not a technique improvement.
• Tweaked parts of the algorithm to give more control to the user and use an error bound that I prefer. Again, an implementation change/improvement over the linked paper, not a technique improvement.

Thanks in advance for any guidance you can provide!

[kyleniemeyer]

Hi @JoshKarpel, I definitely think your code is a substantial contribution, so please feel free to submit.

I took a quick look at your paper—it's on the right track, but could be a bit longer with some example uses, a bit more description of the method, etc.

[JoshKarpel]

Thanks for the guidance!