Relative sizes of the different variants of the full model.

Execution speed increases are calculated relative to a direct translation of the original BrainSignals model into the BCMD framework. Thus, some improvements are seen from refactoring and optimisation even with no model reduction. Values were averaged over 1400 simulations for each model (100 repetitions for each of 10 sets of volunteer input data and 4 steady state simulations). Relative sizes of the different variants of the full model.