Table 2.

Profiling results for several simulation types, “R” for region level simulations, “S” for surface level.

Sim.	Time (s)	Module/function
R / G2D / 20	11.72	<numpy.core.multiarray.array>
	6.14	numpy.lib.npyio, loadtxt
	5.18	tvb.simulator.simulator, __call__
	3.18	numpy.lib.npyio, pack_items
	2.56	numexpr.necompiler, evaluate
2	11.87	<numpy.core.multiarray.array>
	6.10	numpy.lib.npyio, loadtxt
	5.54	tvb.simulator.simulator, __call__
	3.16	numpy.lib.npyio, pack_items
	2.50	numexpr.necompiler, evaluate
JR / 20	14.21	<numpy.core.multiarray.array>
	9.99	tvb.simulator.simulator, __call__
	7.28	tvb.simulator.models, dfun
	6.20	numpy.lib.npyio, loadtxt
	3.24	numpy.lib.npyio, pack_items
2	14.21	<numpy.core.multiarray.array>
	10.57	tvb.simulator.simulator, __call__
	7.40	tvb.simulator.models, dfun
	6.12	numpy.lib.npyio, loadtxt
	3.25	numpy.lib.npyio, pack_items
S / G2D / 20	126.61	<_csc.csc_matvec>
	57.56	<numpy.core.multiarray.array>
	56.17	<gdist.local_gdist_matrix>
	9.05	<numpy.core._dotblas.dot>
	7.56	numpy.lib.npyio, loadtxt
2	125.95	<_csc.csc_matvec>
	57.75	<numpy.core.multiarray.array>
	56.16	<gdist.local_gdist_matrix>
	12.10	<numpy.core._dotblas.dot>
	7.37	numpy.lib.npyio, loadtxt
JR / 20	126.31	<numpy.core.multiarray.array>
	56.37	<gdist.local_gdist_matrix>
	19.52	<numpy.core._dotblas.dot>
	9.47	tvb.simulator.models, dfun
	8.87	<mtrand.RandomState.normal>
2	126.09	<numpy.core.multiarray.array>
	56.79	<gdist.local_gdist_matrix>
	29.10	<numpy.core._dotblas.dot>
	14.18	<mtrand.RandomState.normal>
	9.57	tvb.simulator.models, dfun

“G2D” signifies the generic two-dimensional oscillator whereas “JR” is the Jansen–Rit model. Finally, either 20 m/s or 2 m/s conduction velocity was used. Time is given as the total time spent in the method or function listed in the right column.