A Python version of AFQ is something that I have been working on for almost as long as I have been at the University of Washington. And yet, we have yet to release even a first version of the software the we have been developing here. One of the reasons for this is the fact that neither of I nor my collaborators have not really needed the software to do the things that we wanted to do over this time. PyAFQ remained more aspiration than actual intention.

But now, with several projects planned for this fall that will require a robust and scalable tractometry pipeline, this summer seems to be a good time to finish this up and start using it. In a fortunate turn of events, John Kruper, who is an undergraduate in the computer science department here at UW, and a trainee the computational neuroscience training program, joined our group and has been plugging away at the software over the last couple of weeks. He has already found ways to improve it, which is great. But for this to be truly effective in the long run, we need to start figuring out what the next few weeks should be devoted to.

In other software developments, our software for modeling retinal prosthetics is undergoing a massive overhaul. I spent a couple of hours early on Wedensday morning reviewing the code that Michael Beyeler has redesigned for the spatial aspect of the model. In the next couple of weeks, this work will shift to focus on the temporal aspect. This will be a first opportunity to integrate XArray into one of my neuroscience projects, which is exciting.