# Copyright 2022 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ============================================================================
"""SGConv Layer."""
import math
import mindspore as ms
from mindspore.common.initializer import XavierUniform
from mindspore_gl import Graph
from .. import GNNCell
[文档]class SGConv(GNNCell):
r"""
Simplified Graph convolutional layer.
From the paper `Simplifying Graph Convolutional Networks <https://arxiv.org/pdf/1902.07153.pdf>`_ .
.. math::
H^{K} = (\tilde{D}^{-1/2} \tilde{A} \tilde{D}^{-1/2})^K X \Theta
Where :math:`\tilde{A}=A+I`.
..Note:
PYNATIVE mode only now.
Args:
in_feat_size (int): Input node feature size.
out_feat_size (int): Output node feature size.
num_hops (int, optional): Number of hops. Default: ``1``.
cached (bool, optional): Whether to use cached. Default: ``True``.
bias (bool, optional): Whether to use bias. Default: ``True``.
norm (Cell, optional): Normalization function Cell. Default: ``None``.
Inputs:
- **x** (Tensor) - The input node features. The shape is :math:`(N, D_{in})`
where :math:`N` is the number of nodes,
and :math:`D_{in}` should be equal to `in_feat_size` in `Args`.
- **in_deg** (Tensor) - In degree for nodes. The shape is :math:`(N, )` where :math:`N` is the number of nodes.
- **out_deg** (Tensor) - Out degree for nodes. The shape is :math:`(N, )`
where :math:`N` is the number of nodes.
- **g** (Graph) - The input graph.
Outputs:
- Tensor, output node features with shape of :math:`(N, D_{out})`, where :math:`(D_{out})` should be the same as
`out_feat_size` in `Args`.
Raises:
TypeError: If `in_feat_size` or `out_feat_size` or `num_hops` is not an int.
TypeError: If `bias` or `cached` is not a bool.
TypeError: If `norm` is not a `Cell`.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore as ms
>>> import mindspore.context as context
>>> from mindspore_gl.nn import SGConv
>>> from mindspore_gl import GraphField
>>> context.set_context(device_target="GPU", mode=context.PYNATIVE_MODE)
>>> n_nodes = 4
>>> n_edges = 8
>>> src_idx = ms.Tensor([0, 0, 0, 1, 1, 1, 2, 3], ms.int32)
>>> dst_idx = ms.Tensor([0, 1, 3, 1, 2, 3, 3, 2], ms.int32)
>>> in_deg = ms.Tensor([1, 2, 2, 3], ms.int32)
>>> out_deg = ms.Tensor([3, 3, 1, 1], ms.int32)
>>> feat_size = 4
>>> in_feat_size = feat_size
>>> nh = ms.ops.Ones()((n_nodes, feat_size), ms.float32)
>>> eh = ms.ops.Ones()((n_edges, feat_size), ms.float32)
>>> g = GraphField(src_idx, dst_idx, n_nodes, n_edges)
>>> in_deg = in_deg
>>> out_deg = out_deg
>>> sgconv = SGConv(in_feat_size, feat_size)
>>> res = sgconv(nh, in_deg, out_deg, *g.get_graph())
>>> print(res.shape)
(4, 4)
"""
def __init__(self,
in_feat_size: int,
out_feat_size: int,
num_hops: int = 1,
cached: bool = True,
bias: bool = True,
norm=None):
super().__init__()
if (not isinstance(in_feat_size, int)) or in_feat_size <= 0:
raise ValueError("in_feat_size must be positive int")
if (not isinstance(out_feat_size, int)) or out_feat_size <= 0:
raise ValueError("out_feat_size must be positive int")
if (not isinstance(num_hops, int)) or num_hops <= 0:
raise ValueError("num_hops must be positive int")
if not isinstance(bias, bool):
raise ValueError("bias must be bool")
if not isinstance(cached, bool):
raise ValueError("cached must be bool")
self.in_feat_size = in_feat_size
self.out_feat_size = out_feat_size
self.num_hops = num_hops
self.bias = bias
self.cached = cached
if norm is not None and not isinstance(norm, Cell):
raise TypeError(f"For '{self.cls_name}', the 'activation' must a mindspore.nn.Cell, but got "
f"{type(norm).__name__}.")
self.dense = ms.nn.Dense(self.in_feat_size, self.out_feat_size, has_bias=self.bias,
weight_init=XavierUniform(math.sqrt(2)))
self.cached_h = None
self.norm = norm
self.min_clip = ms.Tensor(1, ms.int32)
self.max_clip = ms.Tensor(100000000, ms.int32)
# pylint: disable=arguments-differ
def construct(self, x, in_deg, out_deg, g: Graph):
"""
Construct function for SGConv.
"""
feat = x
if self.cached_h:
feat = self.cached_h
else:
in_deg = ms.ops.clip_by_value(in_deg, self.min_clip, self.max_clip)
in_deg = ms.ops.Reshape()(ms.ops.Pow()(in_deg, -0.5), ms.ops.Shape()(in_deg) + (1,))
out_deg = ms.ops.clip_by_value(out_deg, self.min_clip, self.max_clip)
out_deg = ms.ops.Reshape()(ms.ops.Pow()(out_deg, -0.5), ms.ops.Shape()(out_deg) + (1,))
for _ in range(self.num_hops):
feat = feat * out_deg
g.set_vertex_attr({"h": feat})
for v in g.dst_vertex:
v.h = g.sum([u.h for u in v.innbs])
feat = [v.h for v in g.dst_vertex] * in_deg
if self.norm is not None:
feat = self.norm(feat)
if self.cached:
self.cached_h = feat
return self.dense(feat)