davestanley-cogrhythms / a1_delta_gamma_model Goto Github PK

kramer_IB needs to be re-ordered to work better with function mode.

I will re-order to separate "simulation master parameters" (e.g. simulation-wide parameters that control parameters of the actual model) from regular parameters.

Right now, LTS cells do the appropriate behavior on the aperiodic spike based on being driven by noise.

This is because RS cells always fire distinctly ahead of FS cells, which prevents them from properly canceling each other. Due to this lack of cancelation, the most "opportune" firing window is after RS fire and right before FS fire.

I'm going to try to make the RS-FS network more noisy and therefore the spike times will be more jittered and overlapping. Thus, EPSPs and IPSPs from RS and FS cells respectively might be better at cancelling each other out, allowing the "unbalanced" pulse (e.g. the 11th pulse, immediately after the AP pulse) to be the one that triggers LTS firing.

This would be more robust than our current noise-driven LTS firing on the aperiodic pulse.

With new version of code, specifying study_dir breaks compiler mode. This is because it saves the compiled files in a different folder on every sim (named based on the current date), thus preventing re-using the mex files from previous simulations

Apparently monitor functions can bog DynaSim down by an order of 5-50x. I'm going to experiment removing this and seeing how it affects sim speed