package gpr

  1. Overview
  2. Docs
Legend:
Library
Module
Module type
Parameter
Class
Class type

(Untrained) model - does not require targets

type t

Type of models

type co_variance_coeffs

Type of covariance coefficients

val calc : Inputs.t -> sigma2:float -> t

calc inputs ~sigma2

  • returns

    model given inputs and noise level sigma2 (= variance, i.e. squared standard deviation).

val update_sigma2 : t -> float -> t

update_sigma2 model sigma2

  • returns

    model by updating model with new noise level sigma2.

val calc_log_evidence : t -> float

calc_log_evidence model

  • returns

    the contribution to the log evidence (= log marginal likelihood) of model.

val calc_co_variance_coeffs : t -> co_variance_coeffs

calc_co_variance_coeffs model

  • returns

    the coefficients required for computing posterior (co-)variances for model.

val get_kernel : t -> Spec.Kernel.t

get_kernel model

  • returns

    the kernel associated with model.

val get_sigma2 : t -> float

get_sigma2 model

  • returns

    the noise level associated with model.

val get_inputs : t -> Inputs.t

get_inputs model

  • returns

    the inputs associated with model.

val get_inducing : t -> Inducing.t

get_inputs model

  • returns

    the inducing inputs associated with model.