sup3r.utilities.loss_metrics

Content loss metrics for Sup3r

Functions

gaussian_kernel(x1, x2[, sigma])

Gaussian kernel for mmd content loss

Classes

CoarseMseLoss([reduction, name])	Loss class for coarse mse on spatial average of 5D tensor
ExpLoss([reduction, name])	Loss class for squared exponential difference
LowResLoss([s_enhance, t_enhance, t_method, ...])	Content loss that is calculated by coarsening the synthetic and true high-resolution data pairs and then performing the pointwise content loss on the low-resolution fields
MaterialDerivativeLoss([reduction, name])	Loss class for the material derivative.
MmdLoss([reduction, name])	Loss class for max mean discrepancy loss
MmdMseLoss([reduction, name])	Loss class for MMD + MSE
MseExpLoss([reduction, name])	Loss class for mse + squared exponential difference
SpatialExtremesLoss([weight])	Loss class that encourages accuracy of the min/max values in the spatial domain
SpatialExtremesOnlyLoss([reduction, name])	Loss class that encourages accuracy of the min/max values in the spatial domain.
SpatialFftOnlyLoss([reduction, name])	Loss class that encourages accuracy of the spatial frequency spectrum
SpatiotemporalExtremesLoss([spatial_weight, ...])	Loss class that encourages accuracy of the min/max values across both space and time
SpatiotemporalFftOnlyLoss([reduction, name])	Loss class that encourages accuracy of the spatiotemporal frequency spectrum
StExtremesFftLoss([spatial_weight, ...])	Loss class that encourages accuracy of the min/max values across both space and time as well as frequency domain accuracy.
TemporalExtremesLoss([weight])	Loss class that encourages accuracy of the min/max values in the timeseries
TemporalExtremesOnlyLoss([reduction, name])	Loss class that encourages accuracy of the min/max values in the timeseries.