pyg_spectral.nn

• pyg_spectral.nn.conv

pyg_spectral.nn.conv.ACMConv	Convolutional layer of FBGNN & ACMGNN(I & II).
pyg_spectral.nn.conv.AdjConv	Linear filter using the normalized adjacency matrix for propagation.
pyg_spectral.nn.conv.AdjDiffConv	Linear filter using the normalized adjacency matrix for propagation.
pyg_spectral.nn.conv.AdjResConv	Iterative linear filter with residual connection.
pyg_spectral.nn.conv.AdjSkip2Conv	Iterative linear filter with 2-hop propagation and skip connection.
pyg_spectral.nn.conv.AdjSkipConv	Iterative linear filter with skip connection.
pyg_spectral.nn.conv.Adji2Conv	Iterative linear filter using the 2-hop normalized adjacency matrix for augmented propagation.
pyg_spectral.nn.conv.AdjiConv	Iterative linear filter using the normalized adjacency matrix for augmented propagation.
pyg_spectral.nn.conv.BaseMP	Base filter layer structure.
pyg_spectral.nn.conv.BernConv	Convolutional layer with Bernstein Polynomials.
pyg_spectral.nn.conv.ChebConv	Convolutional layer with Chebyshev Polynomials.
pyg_spectral.nn.conv.ChebIIConv	Convolutional layer with Chebyshev-II Polynomials.
pyg_spectral.nn.conv.ClenshawConv	Convolutional layer with Chebyshev Polynomials and explicit residual.
pyg_spectral.nn.conv.FavardConv	Convolutional layer with basis in Favard's Theorem.
pyg_spectral.nn.conv.HornerConv	Convolutional layer with adjacency propagation and explicit residual.
pyg_spectral.nn.conv.JacobiConv	Convolutional layer with Jacobi Polynomials.
pyg_spectral.nn.conv.LapiConv	Iterative linear filter using the normalized adjacency matrix.
pyg_spectral.nn.conv.LegendreConv	Convolutional layer with Legendre Polynomials.
pyg_spectral.nn.conv.OptBasisConv	Convolutional layer with optimal adaptive basis.

• pyg_spectral.nn.models

pyg_spectral.nn.models.ACMGNN	Iterative structure for ACM conv.
pyg_spectral.nn.models.ACMGNNDec	Decoupled structure for ACM conv.
pyg_spectral.nn.models.AdaGNN	Decoupled structure with diag transformation each hop of propagation.
pyg_spectral.nn.models.BaseNN	Base NN structure with MLP before and after convolution layers.
pyg_spectral.nn.models.BaseNNCompose	Base NN structure with multiple conv channels.
pyg_spectral.nn.models.CppCompFixed	Decoupled structure with C++ propagation precomputation.
pyg_spectral.nn.models.DecoupledFixed	Decoupled structure without matrix transformation during propagation.
pyg_spectral.nn.models.DecoupledFixedCompose	Decoupled structure without matrix transformation during propagation.
pyg_spectral.nn.models.DecoupledVar	Decoupled structure without matrix transformation during propagation.
pyg_spectral.nn.models.DecoupledVarCompose	Decoupled structure without matrix transformation during propagation.
pyg_spectral.nn.models.Iterative	Iterative structure with matrix transformation each hop of propagation.
pyg_spectral.nn.models.IterativeCompose	Iterative structure with matrix transformation each hop of propagation.
pyg_spectral.nn.models.PrecomputedFixed	Decoupled structure with precomputation separating propagation from transformation.
pyg_spectral.nn.models.PrecomputedFixedCompose	Decoupled structure with precomputation separating propagation from transformation.
pyg_spectral.nn.models.PrecomputedVar	Decoupled structure with precomputation separating propagation from transformation.
pyg_spectral.nn.models.PrecomputedVarCompose	Decoupled structure with precomputation separating propagation from transformation.

• pyg_spectral.nn.norm

pyg_spectral.nn.norm.TensorStandardScaler

Applies standard Gaussian normalization to \(\mathcal{N}(0, 1)\).