108 curated models, including AdaptiveExponentialIF, Antidots, ArnoldCatMap, BetaTransformationMap …Models & Dynamics
Specify the dynamical system at each node: equations, parameters, noise, events and units.
Part of the running example, where stage 1 adds the model, the network and the clock.
Every TVB-O simulation starts with a dynamical system: the equations that govern one node, before any network exists. You can write those equations yourself, or reference one of the curated models by its ontology IRI and never type them at all.
Most people start by reusing a curated model and only later declare their own. Both paths use the same specification, so moving from one to the other costs nothing.
Writing the system
Defining dynamical systems
Events & stimulation
Writing a model from scratch, from a damped pendulum to a neural mass.
Initial conditionsSetting the starting state, fixed or sampled from a distribution.
Noise & stochastic integrationAdditive and multiplicative noise, and what turns an ODE into an SDE.
Events & stimulationThreshold events, preset-time events, and external stimuli.
Unit-aware computationsKeeping a model dimensionally consistent.
Solving it
Beyond one averaged unit
A neural mass is one averaged population per region. When that is too coarse, Beyond the mean field covers spiking cells, ion channels and multi-scale models, and Bifurcation & continuation covers finding where a model changes qualitative behaviour.
Once the node is specified, Networks & connectomes wires many of them into a brain.
Reference
Dynamics · StateVariable · Parameter · the model gallery.