dgl.sparse.identity

dgl.sparse.identity(shape: Tuple[int, int], d: int | None = None, dtype: dtype | None = None, device: device | None = None) → SparseMatrix[source]

Creates a sparse matrix with ones on the diagonal and zeros elsewhere.

Parameters:

shape (tuple[int, int]) – Shape of the matrix.
d (int, optional) – If None, the diagonal entries will be scaler 1. Otherwise, the diagonal entries will be a 1-valued tensor of shape (d).
dtype (torch.dtype, optional) – The data type of the matrix
device (torch.device, optional) – The device of the matrix

Returns:

Sparse matrix

Return type:

SparseMatrix

Examples

Case1: 3-by-3 matrix with scaler diagonal values

[[1, 0, 0],
 [0, 1, 0],
 [0, 0, 1]]

>>> dglsp.identity(shape=(3, 3))
SparseMatrix(indices=tensor([[0, 1, 2],
                             [0, 1, 2]]),
             values=tensor([1., 1., 1.]),
             shape=(3, 3), nnz=3)

Case2: 3-by-5 matrix with scaler diagonal values

[[1, 0, 0, 0, 0],
 [0, 1, 0, 0, 0],
 [0, 0, 1, 0, 0]]

>>> dglsp.identity(shape=(3, 5))
SparseMatrix(indices=tensor([[0, 1, 2],
                             [0, 1, 2]]),
             values=tensor([1., 1., 1.]),
             shape=(3, 5), nnz=3)

Case3: 3-by-3 matrix with vector diagonal values

>>> dglsp.identity(shape=(3, 3), d=2)
SparseMatrix(indices=tensor([[0, 1, 2],
                             [0, 1, 2]]),
             values=tensor([[1., 1.],
                            [1., 1.],
                            [1., 1.]]),
             shape=(3, 3), nnz=3, val_size=(2,))