def __init__(self, nfeat, nhid, nclass, dropout, nlayer=3):
super(HyperGraphX, self).__init__()
self.conv1 = HypergraphConv(nfeat, nhid)
self.conv2 = HypergraphConv(nhid, nclass)
self.convx = nn.ModuleList(
[HypergraphConv(nhid, nhid) for _ in range(nlayer - 2)])
self.dropout_p = dropout
def test_hypergraph_conv():
in_channels, out_channels = (16, 32)
hyper_index = torch.tensor([[0, 0, 0, 1, 2, 3], [1, 2, 3, 0, 0, 0]])
hyper_weight = torch.tensor([1, 0.5, 0.3, 0.7])
num_nodes = hyper_index.max().item() + 1
x = torch.randn((num_nodes, in_channels))
conv = HypergraphConv(in_channels, out_channels)
assert conv.__repr__() == 'HypergraphConv(16, 32)'
out = conv(x, hyper_index, hyper_weight)
assert out.size() == (num_nodes, out_channels)
def test_hypergraph_conv_with_more_edges_than_nodes():
in_channels, out_channels = (16, 32)
hyperedge_index = torch.tensor([[0, 0, 1, 1, 2, 3, 3, 3, 2, 1, 2],
[0, 1, 2, 1, 2, 1, 0, 3, 3, 4, 4]])
hyperedge_weight = torch.tensor([1.0, 0.5, 0.8, 0.2, 0.7])
num_nodes = hyperedge_index[0].max().item() + 1
x = torch.randn((num_nodes, in_channels))
conv = HypergraphConv(in_channels, out_channels)
assert conv.__repr__() == 'HypergraphConv(16, 32)'
out = conv(x, hyperedge_index)
assert out.size() == (num_nodes, out_channels)
out = conv(x, hyperedge_index, hyperedge_weight)
assert out.size() == (num_nodes, out_channels)
def test_hypergraph_conv():
in_channels, out_channels = (16, 32)
hyperedge_index = torch.tensor([[0, 0, 1, 1, 2, 3], [0, 1, 0, 1, 0, 1]])
hyperedge_weight = torch.tensor([1, 0.5])
num_nodes = hyperedge_index[0].max().item() + 1
x = torch.randn((num_nodes, in_channels))
conv = HypergraphConv(in_channels, out_channels)
assert conv.__repr__() == 'HypergraphConv(16, 32)'
out = conv(x, hyperedge_index)
assert out.size() == (num_nodes, out_channels)
out = conv(x, hyperedge_index, hyperedge_weight)
assert out.size() == (num_nodes, out_channels)
conv = HypergraphConv(in_channels,
out_channels,
use_attention=True,
heads=2)
out = conv(x, hyperedge_index)
assert out.size() == (num_nodes, 2 * out_channels)
out = conv(x, hyperedge_index, hyperedge_weight)
assert out.size() == (num_nodes, 2 * out_channels)
conv = HypergraphConv(in_channels,
out_channels,
use_attention=True,
heads=2,
concat=False,
dropout=0.5)
out = conv(x, hyperedge_index, hyperedge_weight)
assert out.size() == (num_nodes, out_channels)
def __init__(self, dim_in, dim_out, bias=False, **kwargs):
super(HypergraphConvGG, self).__init__()
self.model = HypergraphConv(dim_in, dim_out, bias=bias)
def __init__(self, nfeat, nhid, nclass, dropout, nlayer=1):
super(HyperGraph1, self).__init__()
self.conv1 = HypergraphConv(nfeat, nclass)
self.dropout_p = dropout