pubmed

Bases: graph_dataloader

A dataloader class for the PubMed dataset.

This class extends graph_dataloader to handle the PubMed graph dataset, a large-scale dataset used for graph-based learning.

Attributes:

Name	Type	Description
data_profile	dict	Metadata and download links specific to the PubMed dataset.
graph	graph	The loaded graph structure for the PubMed dataset.

Methods:

Name	Description
__init__	Initializes the dataloader for the PubMed dataset.
get_train_test_idx	Generates train and test indices for the PubMed dataset.

Source code in tinybig/data/graph_dataloader.py

class pubmed(graph_dataloader):
    """
    A dataloader class for the PubMed dataset.

    This class extends `graph_dataloader` to handle the PubMed graph dataset,
    a large-scale dataset used for graph-based learning.

    Attributes
    ----------
    data_profile: dict
        Metadata and download links specific to the PubMed dataset.
    graph: graph_class
        The loaded graph structure for the PubMed dataset.

    Methods
    -------
    __init__(name: str = 'pubmed', train_batch_size: int = 64, test_batch_size: int = 64, ...)
        Initializes the dataloader for the PubMed dataset.
    get_train_test_idx(X: torch.Tensor = None, y: torch.Tensor = None, ...)
        Generates train and test indices for the PubMed dataset.
    """
    def __init__(self, name: str = 'pubmed', train_batch_size: int = 64, test_batch_size: int = 64, *args, **kwargs):
        """
        Initializes the dataloader for the PubMed dataset.

        Parameters
        ----------
        name: str, default = 'pubmed'
            The name of the dataset.
        train_batch_size: int, default = 64
            Batch size for the training dataset.
        test_batch_size: int, default = 64
            Batch size for the testing dataset.

        Returns
        -------
        None
        """
        super().__init__(data_profile=PUBMED_DATA_PROFILE, name=name, train_batch_size=train_batch_size, test_batch_size=test_batch_size)

    def get_train_test_idx(self, X: torch.Tensor = None, y: torch.Tensor = None, *args, **kwargs):
        """
        Generates train and test indices for the PubMed dataset.

        Parameters
        ----------
        X: torch.Tensor, optional
            Node features (not used in this method).
        y: torch.Tensor, optional
            Labels (not used in this method).

        Returns
        -------
        tuple
            Train indices (`torch.LongTensor`) and test indices (`torch.LongTensor`).

        Notes
        -----
        The train indices are predefined as the first 60 nodes.
        The test indices are predefined as nodes 6300 to 7299.
        """
        train_idx = torch.LongTensor(range(60))
        test_idx = torch.LongTensor(range(6300, 7300))
        return train_idx, test_idx

__init__(name='pubmed', train_batch_size=64, test_batch_size=64, *args, **kwargs)

Initializes the dataloader for the PubMed dataset.

Parameters:

Name	Type	Description	Default
name	str	The name of the dataset.	'pubmed'
train_batch_size	int	Batch size for the training dataset.	64
test_batch_size	int	Batch size for the testing dataset.	64

Returns:

Type	Description
None

Source code in tinybig/data/graph_dataloader.py

def __init__(self, name: str = 'pubmed', train_batch_size: int = 64, test_batch_size: int = 64, *args, **kwargs):
    """
    Initializes the dataloader for the PubMed dataset.

    Parameters
    ----------
    name: str, default = 'pubmed'
        The name of the dataset.
    train_batch_size: int, default = 64
        Batch size for the training dataset.
    test_batch_size: int, default = 64
        Batch size for the testing dataset.

    Returns
    -------
    None
    """
    super().__init__(data_profile=PUBMED_DATA_PROFILE, name=name, train_batch_size=train_batch_size, test_batch_size=test_batch_size)

get_train_test_idx(X=None, y=None, *args, **kwargs)

Generates train and test indices for the PubMed dataset.

Parameters:

Name	Type	Description	Default
X	Tensor	Node features (not used in this method).	None
y	Tensor	Labels (not used in this method).	None

Returns:

Type	Description
tuple	Train indices (torch.LongTensor) and test indices (torch.LongTensor).

Notes

The train indices are predefined as the first 60 nodes. The test indices are predefined as nodes 6300 to 7299.

Source code in tinybig/data/graph_dataloader.py

def get_train_test_idx(self, X: torch.Tensor = None, y: torch.Tensor = None, *args, **kwargs):
    """
    Generates train and test indices for the PubMed dataset.

    Parameters
    ----------
    X: torch.Tensor, optional
        Node features (not used in this method).
    y: torch.Tensor, optional
        Labels (not used in this method).

    Returns
    -------
    tuple
        Train indices (`torch.LongTensor`) and test indices (`torch.LongTensor`).

    Notes
    -----
    The train indices are predefined as the first 60 nodes.
    The test indices are predefined as nodes 6300 to 7299.
    """
    train_idx = torch.LongTensor(range(60))
    test_idx = torch.LongTensor(range(6300, 7300))
    return train_idx, test_idx