chore: import upstream snapshot with attribution

python/dgl/init.py

@@ -0,0 +1,67 @@
+"""Module for common feature initializers."""
+from __future__ import absolute_import
+
+from . import backend as F
+
+__all__ = ["base_initializer", "zero_initializer"]
+
+
+def base_initializer(
+    shape, dtype, ctx, id_range
+):  # pylint: disable=unused-argument
+    """The function signature for feature initializer.
+
+    Any customized feature initializer should follow this signature (see
+    example below).
+
+    Parameters
+    ----------
+    shape : tuple of int
+        The shape of the result features. The first dimension
+        is the batch dimension.
+    dtype : data type object
+        The data type of the returned features.
+    ctx : context object
+        The device context of the returned features.
+    id_range : slice
+        The start id and the end id of the features to be initialized.
+        The id could be node or edge id depending on the scenario.
+        Note that the step is always None.
+
+    Examples
+    --------
+    If PyTorch is used as backend, the following code defines an feature
+    initializer that initializes tensor value to 1
+
+    >>> import torch
+    >>> import dgl
+    >>> def initializer(shape, dtype, ctx, id_range):
+    >>>     return torch.ones(shape, dtype=dtype, device=ctx)
+    >>> g = dgl.DGLGraph()
+    >>> g.set_n_initializer(initializer)
+
+    See Also
+    --------
+    dgl.DGLGraph.set_n_initializer
+    dgl.DGLGraph.set_e_initializer
+    """
+    raise NotImplementedError
+
+
+def zero_initializer(
+    shape, dtype, ctx, id_range
+):  # pylint: disable=unused-argument
+    """Zero feature initializer
+
+    Examples
+    --------
+    >>> import dgl
+    >>> g = dgl.DGLGraph()
+    >>> g.set_n_initializer(dgl.init.zero_initializer)
+
+    See Also
+    --------
+    dgl.DGLGraph.set_n_initializer
+    dgl.DGLGraph.set_e_initializer
+    """
+    return F.zeros(shape, dtype, ctx)