Package org.ddogleg.chainrule.ops
package org.ddogleg.chainrule.ops
-
ClassesClassDescriptionThis is a
Huber Lossfunction adapted to return the error and is triggered based on the F-Norm of the error vector instead of individual elements.Computes the error between a known prior and an estimated value.Huber Loss is a robust loss function that is less sensitive to outliers than the squared error loss.General purpose operator for blending 4 knots into an N-D output, each knot stores N-coefficients for a yet to be defined function.Implementation ofCROperatorBilinearNDwhere the knots are general polynomials.Implementation ofCROperatorBilinearNDwhere the knots are also splines which use linear interpolation.Holds a value at or above a floor:output = max(input, minimum).Concatenates its inputs into one output, in the order they were declared.Element-wise division of two vectorsDivides a vector by a scalar value.Divides a vector by a damped scalar:a[i]/(beta + b).Turns a pixel into the 4 bilinear weights of the cell it lands in, ready to be blended byCROperatorBilinearPolyND.Mean ofcountvalues read at a fixed stride from one input vector.Element-wise multiplication of two vectorsMultiplies a vector by a scalar value.Adds two arrays of equal length together.Adds two vectors with scalar coefficients: 'output = alphaV0 + betaV1'Remaps parameters from an input format into an output format.Penalizes a contiguous run of parameters, one residual per parameter, eachweight*value.