starling_sim.basemodel.schedule.scheduler

This module is dedicated to discrete event scheduling. It allows the representation of a time flow inside the simulation.

Module Contents

Classes

Scheduler

This class manages a discrete event schedule using the SimPy library.

class starling_sim.basemodel.schedule.scheduler.Scheduler

This class manages a discrete event schedule using the SimPy library.

Along with containing a SimPy environment, this class also provides methods for the generation of SimPy Event objects (Process objects are events too). Yielding these events will suspend the current activity (e.g agent loop), which will resume once the event occurs (or before if the process is interrupted).

This way, agents can yield processes corresponding to an action (move, wait, etc) and resume to their loop after the action has finished.

interrupt

run(run_time)

Update the runtime limit attribute and run the environment.

Parameters:

run_time – duration of the simulation run

now()

Get the current simulation step.

Returns:

Current time of the simulation (arbitrary unit)

timeout(duration='inf')

Create a SimPy timeout event with the given duration

Parameters:

duration – time to pass, in simulation units. Defaults to ‘inf’, ie end at the simulation time limit. Must be positive and integer-like

Returns:

SimPy timeout event

new_event_object()

Create and return a new Event object

Returns:

Event object

new_process(generator)

Create a SimPy process from the given generator

Parameters:

generator – Python generator (function containing the yield keyword)

Returns:

SimPy Process object

static interrupt(process, message='')

Interrupt the given SimPy process, e.g. an agent waiting in a queue

Parameters:

• process – simPy Process object

• message – interruption message, to be displayed in the agent’s trace