prior_conversion_sgp {BayesGP}R Documentation

Construct prior based on d-step prediction SD (for sgp)

Description

Construct prior based on d-step prediction SD (for sgp)

Usage

prior_conversion_sgp(d, prior, freq, period, a, m = 1)

Arguments

d

A numeric value for the prediction step.

prior

A list that contains alpha and u. This specifies the target prior on the d-step SD \sigma(d), such that P(\sigma(d) > u) = alpha.

freq

The frequency of the sgp, ignored if a is provided.

period

The period of the sgp, ignored if a or freq is provided.

a

The frequency parameter of the sgp.

m

The number of harmonics that should be considered, by default m = 1 represents only the sgp.

Value

A list that contains alpha and u. The prior for the smoothness parameter \sigma such that P(\sigma > u) = alpha, that yields the ideal prior on the d-step SD.


[Package BayesGP version 0.1.3 Index]