timestep function

(Shortest import: from brian2.core.functions import timestep)

brian2.core.functions.timestep(t, dt)[source]

Converts a given time to an integer time step. This function slightly shifts the time before dividing it by dt to make sure that multiples of dt do not end up in the preceding time step due to floating point issues. This function is used in the refractoriness calculation.

New in version 2.1.3.


t : np.ndarray, float, Quantity

The time to convert.

dt : float or Quantity

The length of a simulation time step.


ts : np.ndarray, int

The time step corresponding to the given time.


This function can handle infinity values, it will return a value equal to half of the maximal integer value. This assures that an expression such as timestep(t) - timestep(lastspike) will result in a reasonable value even if lastspike is -inf. If timestep(lastspike) were equal to the minimal representable integer value, this expression would overflow.