My main night-time coding/writing accompaniment for the last week or so (h/t & © signalrun).

Initial fuckery with field/spiking network hybrid models is over, now it’s time for polishing and paperwork (and maybe, just maybe, sleep).

Conclusions from the last few months:

• When in doubt, read Theoretical Neuroscience, by Dayan, P. and Abbott, L. F. (2001);
• MATLAB is a great tool but its language gets more atrocious the longer one uses it;
• The previous point can be remedied by just writing everything in Julia and not looking back;
• Converting average firing rates to individual spikes in a population of neurons by way of first converting the former into instantaneous firing rates for a group of Poisson “neurons” inserted between the neural field and the spiking population is not cheating, even if it kind of feels like that;
• BLAS/LAPACK is magic (both in the sense that it allows solving complex problems with a few short incantations and in the sense that it is a thing of nightmare and eldritch portent);
• I still can do double and triple all-nighters, fuck yes.

Two hours of sleep, fresh coffee and back to keyboard, whiteboard, and paper. This is the old days part two, revenge of the old days.

Nothing is destroyed until it is replaced

Wiped this place clean, to move from collection of digital lagan to something more coherent and focused. Notes and tidbits from our computational neuroscience research will go to my secondary blog.

Now, confused as to where go next with this one.