discrepancy_models (pmrf.discrepancy_models)

Models that cater for the discrepancy between a physical model and data.

Classes

Kernel(*[, name])	Abstract base class for kernel functions enabling kernel algebra.
SumKernel(k1, k2, *[, name])	Kernel representing the sum of two kernels.
ProductKernel(k1, k2, *[, name])	Kernel representing the product of two kernels.
ConstantKernel([variance, name])	Kernel that returns a constant variance.
RBFKernel([length_scale, name])	Radial Basis Function (Squared Exponential) kernel.
WhiteNoiseKernel([variance, name])	Kernel representing independent Gaussian noise.
GaussianProcessDiscrepancy(kernel[, jitter, ...])	Maps model predictions to a Gaussian Process distribution over frequency.