DataLoaderUtils Class Reference

Class to provide utility functions to load tensors and configuration files. More...

Static Public Member Functions
torch.tensor	load_maps (str path, bool use_comm_map=False)
	Function to load maps stored as tensors.
tuple[torch.tensor, torch.tensor, torch.tensor]	load_features (str path, int output_dim=None)
	Function to load normalized features.
tuple[torch.tensor, torch.tensor, torch.tensor]	load_actions (str path)
	Function to load normalized actions.
torch.tensor	load_robot_positions (str path)
	Function to load robot positions.
torch.tensor	load_edge_weights (str path)
	Function to load edge weights.
torch_geometric.data.Data	to_torch_geometric_data (torch.tensor feature, torch.tensor edge_weights, torch.tensor pos=None)
	The function converts the feature, edge_weights and pos to a torch_geometric.data.Data object This is essential for using the data with the PyTorch Geometric library.

Detailed Description

Class to provide utility functions to load tensors and configuration files.

Definition at line 33 of file data_loader_utils.py.

Member Function Documentation

load_actions()

tuple[torch.tensor, torch.tensor, torch.tensor] load_actions ( str path )

static

Function to load normalized actions.

   The actions are stored as tensors in the following format:
   - {path}/normalized_actions.pt: Normalized actions
   - {path}/../actions_mean.pt: Mean of the actions
   - {path}/../actions_std.pt: Standard deviation of the actions

Parameters

path	Path to the directory containing the actions

Returns

actions The loaded actions

actions_mean Mean of the actions

actions_std Standard deviation of the actions

Definition at line 128 of file data_loader_utils.py.

load_edge_weights()

torch.tensor load_edge_weights ( str path )

static

Function to load edge weights.

   The edge weights are stored as tensors in the following format:
   - {path}/edge_weights.pt: Edge weights

Parameters

path	Path to the directory containing the edge weights

Returns

edge_weights The loaded edge weights

Definition at line 168 of file data_loader_utils.py.

load_features()

tuple[torch.tensor, torch.tensor, torch.tensor] load_features ( str path, int output_dim = None )

static

Function to load normalized features.

   The features are stored as tensors in the following format:
   - {path}/normalized_coverage_features.pt: Normalized coverage features
   - {path}/../coverage_features_mean.pt: Mean of the coverage features
   - {path}/../coverage_features_std.pt: Standard deviation of the coverage features

Parameters

path	Path to the directory containing the features
output_dim	Output dimension of the features

Returns

features The loaded features

features_mean Mean of the features

features_std Standard deviation of the features

Definition at line 86 of file data_loader_utils.py.

load_maps()

torch.tensor load_maps ( str path, bool use_comm_map = False )

static

Function to load maps stored as tensors.

   The maps are stored as tensors in the following format:
   - {path}/local_maps.pt: Local maps
   - {path}/obstacle_maps.pt: Obstacle maps
   - {path}/comm_maps.pt: Communication maps (if use_comm_map is True)

Parameters

path	Path to the directory containing the maps
use_comm_map	Whether to load the communication map

Returns

maps The loaded maps

Definition at line 52 of file data_loader_utils.py.

load_robot_positions()

torch.tensor load_robot_positions ( str path )

static

Function to load robot positions.

   The robot positions are stored as tensors in the following format:
   - {path}/robot_positions.pt: Robot positions

Parameters

path	Path to the directory containing the robot positions

Returns

robot_positions The loaded robot positions

Definition at line 149 of file data_loader_utils.py.

to_torch_geometric_data()

torch_geometric.data.Data to_torch_geometric_data ( torch.tensor feature, torch.tensor edge_weights, torch.tensor pos = None )

static

The function converts the feature, edge_weights and pos to a torch_geometric.data.Data object This is essential for using the data with the PyTorch Geometric library.

Parameters

feature	The feature tensor
edge_weights	The edge weights tensor
pos	The position tensor

Returns

data The torch_geometric.data.Data object

data The feature tensor

data The edge index tensor

data The edge weight tensor

data The position tensor (if pos is not None)

Definition at line 192 of file data_loader_utils.py.