CNNGNNDataset Class Reference

Dataset for hybrid CNN-GNN training. More...

Inheritance diagram for CNNGNNDataset:

Public Member Functions
	__init__ (self, data_dir, stage, use_comm_map, world_size)
	len (self)
	get (self, idx)

Public Attributes
	stage = stage
	maps = DataLoaderUtils.load_maps(f"{data_dir}/{stage}", use_comm_map)
	dataset_size = self.maps.shape[0]
	targets
	targets_mean
tuple	targets_std
	edge_weights = DataLoaderUtils.load_edge_weights(f"{data_dir}/{stage}")
tuple	robot_positions

Detailed Description

Dataset for hybrid CNN-GNN training.

Definition at line 100 of file loaders.py.

Constructor & Destructor Documentation

__init__()

__init__	(		self,
			data_dir,
			stage,
			use_comm_map,
			world_size )

Definition at line 102 of file loaders.py.

Member Function Documentation

get()

get	(		self,
			idx )

Definition at line 131 of file loaders.py.

len()

len

(

self

)

Definition at line 128 of file loaders.py.

Member Data Documentation

dataset_size

dataset_size = self.maps.shape[0]

Definition at line 108 of file loaders.py.

edge_weights

edge_weights = DataLoaderUtils.load_edge_weights(f"{data_dir}/{stage}")

Definition at line 113 of file loaders.py.

maps

maps = DataLoaderUtils.load_maps(f"{data_dir}/{stage}", use_comm_map)

Definition at line 107 of file loaders.py.

robot_positions

tuple robot_positions

Initial value:

=  DataLoaderUtils.load_robot_positions(
                f"{data_dir}/{stage}"
                )

Definition at line 115 of file loaders.py.

stage

stage = stage

Definition at line 105 of file loaders.py.

targets

targets

Definition at line 110 of file loaders.py.

targets_mean

targets_mean

Definition at line 110 of file loaders.py.

targets_std

tuple targets_std

Initial value:

=  (
                DataLoaderUtils.load_actions(f"{data_dir}/{stage}")
                )

Definition at line 110 of file loaders.py.