本文整理汇总了Python中torch_sparse.spmm方法的典型用法代码示例。如果您正苦于以下问题：Python torch_sparse.spmm方法的具体用法？Python torch_sparse.spmm怎么用？Python torch_sparse.spmm使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类torch_sparse的用法示例。
在下文中一共展示了torch_sparse.spmm方法的5个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: forward
# 需要导入模块: import torch_sparse [as 别名]
# 或者: from torch_sparse import spmm [as 别名]
def forward(self, normalized_adjacency_matrix, features):
"""
Doing a forward pass.
:param normalized_adjacency_matrix: Normalized adjacency matrix.
:param features: Feature matrix.
:return base_features: Convolved features.
"""
base_features = torch.mm(features, self.weight_matrix)
base_features = torch.nn.functional.dropout(base_features,
p=self.dropout_rate,
training=self.training)
for _ in range(self.iterations-1):
base_features = spmm(normalized_adjacency_matrix["indices"],
normalized_adjacency_matrix["values"],
base_features.shape[0],
base_features.shape[0],
base_features)
base_features = base_features + self.bias
return base_features 示例2: forward
# 需要导入模块: import torch_sparse [as 别名]
# 或者: from torch_sparse import spmm [as 别名]
def forward(self, feature_indices, feature_values):
"""
Making a forward pass.
:param feature_indices: Non zero value indices.
:param feature_values: Matrix values.
:return filtered_features: Output features.
"""
number_of_nodes = torch.max(feature_indices[0]).item()+1
number_of_features = torch.max(feature_indices[1]).item()+1
filtered_features = spmm(index = feature_indices,
value = feature_values,
m = number_of_nodes,
n = number_of_features,
matrix = self.weight_matrix)
filtered_features = filtered_features + self.bias
return filtered_features 示例3: get_feat
# 需要导入模块: import torch_sparse [as 别名]
# 或者: from torch_sparse import spmm [as 别名]
def get_feat(self, graph_list):
if self.feat_mode == 'dense':
dense_feat = self.get_fp(graph_list)
else:
sp_indices, vals = self.get_fp(graph_list)
w = self.input_linear.weight
b = self.input_linear.bias
dense_feat = spmm(sp_indices, vals, len(graph_list), w.transpose(0, 1)) + b
return self.mlp(dense_feat), None 示例4: forward
# 需要导入模块: import torch_sparse [as 别名]
# 或者: from torch_sparse import spmm [as 别名]
def forward(self, phi_indices, phi_values, phi_inverse_indices, phi_inverse_values, features):
"""
Forward propagation pass.
:param phi_indices: Sparse wavelet matrix index pairs.
:param phi_values: Sparse wavelet matrix values.
:param phi_inverse_indices: Inverse wavelet matrix index pairs.
:param phi_inverse_values: Inverse wavelet matrix values.
:param features: Feature matrix.
:return localized_features: Filtered feature matrix extracted.
"""
rescaled_phi_indices, rescaled_phi_values = spspmm(phi_indices,
phi_values,
self.diagonal_weight_indices,
self.diagonal_weight_filter.view(-1),
self.ncount,
self.ncount,
self.ncount)
phi_product_indices, phi_product_values = spspmm(rescaled_phi_indices,
rescaled_phi_values,
phi_inverse_indices,
phi_inverse_values,
self.ncount,
self.ncount,
self.ncount)
filtered_features = torch.mm(features, self.weight_matrix)
localized_features = spmm(phi_product_indices,
phi_product_values,
self.ncount,
self.ncount,
filtered_features)
return localized_features 示例5: test_spmm
# 需要导入模块: import torch_sparse [as 别名]
# 或者: from torch_sparse import spmm [as 别名]
def test_spmm(dtype, device):
row = torch.tensor([0, 0, 1, 2, 2], device=device)
col = torch.tensor([0, 2, 1, 0, 1], device=device)
index = torch.stack([row, col], dim=0)
value = tensor([1, 2, 4, 1, 3], dtype, device)
x = tensor([[1, 4], [2, 5], [3, 6]], dtype, device)
out = spmm(index, value, 3, 3, x)
assert out.tolist() == [[7, 16], [8, 20], [7, 19]]