LARGE_INITIALIZE#

invertedai.large.large_initialize(location: str, regions: List[Region], agent_properties: Optional[List[AgentProperties]] = None, agent_states: Optional[List[AgentState]] = None, traffic_light_state_history: Optional[List[Dict[int, TrafficLightState]]] = None, get_infractions: bool = False, random_seed: Optional[int] = None, api_model_version: Optional[str] = None, display_progress_bar: bool = True, return_exact_agents: bool = False) → InitializeResponse[source]#

A utility function to initialize an area larger than 100x100m. This function takes in a list of Region objects on each of which initialize() is run and each initialize response is combined into a single response which is returned. While looping over all regions, if there are agents in other regions that are near enough to the region of interest, they will be passed as conditional to initialize(). initialize() will not be called if no agents are specified in the region. A boolean flag can be used to control failure behaviour if initialize() is unable to produce viable vehicle placements, specifically whether the initialization should continue initializing other regions or raise an exception.

As well, predefined agents may be passed to this function in 2 different ways. If the index of the predefined agents must be preserved, pass these agents’ data into the agent_properties and agent_states parameters. Each agent for which its states is defined MUST have its respective agent properties defined as well but an agent is permitted to be defined by its properties only and initialize() will fill in the state information. If the index of the predefined agents does not matter, they may be placed directly into the region objects or passed into the parameters mentioned previously, but make sure to avoid adding these agents more than once.

Parameters:

location – Please refer to the documentation of initialize() for information on this parameter.
regions – List of regions that contains information about the center, size, agent states and agent properties formatted to align with initialize(). Please refer to the documentation for initialize() for more details.
agent_properties – The properties of the agents that will have their indexes preserved within the response output. Please refer to the documentation of initialize() for information on this parameter.
agent_states – One timestep worth of agent states that will have their indexes preserved within the response output. Please refer to the documentation of initialize() for more information on this parameter.
traffic_light_state_history – Please refer to the documentation of initialize() for information on this parameter.
get_infractions – Please refer to the documentation of initialize() for information on this parameter.
random_seed – Please refer to the documentation of initialize() for information on this parameter.
api_model_version – Please refer to the documentation of initialize() for information on this parameter.
display_progress_bar – If True, a bar is displayed showing the progress of all relevant processes.
return_exact_agents – If set to True, this function will raise an InvertedAIError exception if it cannot fit the requested number of agents in any single region. If set to False, a region that fails to return the number of requested agents will be skipped and only its predefined agents (if any) will be returned with respective RecurrentState’s.