igibson.metrics package

Submodules

igibson.metrics.agent module

class igibson.metrics.agent.AgentMetric

Bases: igibson.metrics.metric_base.MetricBase

gather_results()

Produce a dictionary of values for this metric, to be added onto demo information.

step_callback(igbhvr_act_inst, _)

igibson.metrics.disarrangement module

class igibson.metrics.disarrangement.KinematicDisarrangement

Bases: igibson.metrics.metric_base.MetricBase

static calculate_object_disarrangement(obj, prev_state_cache, cur_state_cache)

This function implements logic to check the current index in order to aggregate kinematic state as an object is split/joined. This logic should be shifted into the multiplexer class, as it requires a significant amount of (error prone) handling.

gather_results()

Produce a dictionary of values for this metric, to be added onto demo information.

property relative_disarrangement
step_callback(igbhvr_act_inst, _)
update_state_cache(task)
class igibson.metrics.disarrangement.LogicalDisarrangement

Bases: igibson.metrics.metric_base.MetricBase

static cache_single_object(obj_id, obj, room_floors, task)
compute_logical_disarrangement(object_state_cache_1, object_state_cache_2)
create_object_logical_state_cache(task)
diff_object_states(obj_1_states, obj_2_states)
end_callback(igbhvr_act_inst, _)

When pybullet sleeps objects, getContactPoints is no longer refreshed Setting collision groups on the agent (which happens when users first activate the agent) Wipes active collision groups. To get the logical disarrangement, we must wake up all objects in the scene This can only be done at the end of the scene so as to not affect determinism.

gather_results()

Produce a dictionary of values for this metric, to be added onto demo information.

step_callback(igbhvr_act_inst, _)

igibson.metrics.gaze module

class igibson.metrics.gaze.GazeMetric

Bases: igibson.metrics.metric_base.MetricBase

gather_results()

Produce a dictionary of values for this metric, to be added onto demo information.

start_callback(igbhvr_act_inst, _)
step_callback(igbhvr_act_inst, log_reader)
class igibson.metrics.gaze.GazeVizMarker(s, radius, color=[1, 0, 0])

Bases: object

Spherical visual marker that can be used to visualize gaze. Does not load into PyBullet, so shouldn’t affect determinism.

set_pos(pos)

igibson.metrics.metric_base module

class igibson.metrics.metric_base.MetricBase

Bases: object

end_callback(igbhvr_act_inst, log_reader)
abstract gather_results()

Produce a dictionary of values for this metric, to be added onto demo information.

start_callback(igbhvr_act_inst, log_reader)
step_callback(igbhvr_act_inst, log_reader)

igibson.metrics.task module

class igibson.metrics.task.TaskMetric

Bases: igibson.metrics.metric_base.MetricBase

gather_results()

Produce a dictionary of values for this metric, to be added onto demo information.

start_callback(igbhvr_act_instance, _)
step_callback(igbhvr_act_inst, _)

Module contents