em_tau: Compute probabilities for the observations to belong to each...

In mixAR: Mixture Autoregressive Models

Compute probabilities for the observations to belong to each of the components

Description

Supporting function for EM algorithm. Update the conditional probabilities of the components of the MixAR model (for E-step of EM algorithm).

Usage

em_tau(stdetk, prob, scale, pdf = dnorm)

em_tau_safe(stdetk, prob, scale, pdf = dnorm)

Arguments

stdetk	standardised component residuals, a MixComp object.
prob	current estimates of the probabilities of the components, a numeric vector of length equal to the number of components in the model.
scale	scales (standard deviations) of the noise components, a numeric vector of length equal to the number of components in the model.
pdf	densities of the noise components.

Details

em_tau and em_tau_safe compute the conditional probabilities of the components of the MixAR model (for the E-step of the EM algorithm). The two functions do the same computations but em_tau_safe, in addition, protects agains NaN's and infinite values in argument stdetk or obtained during computations.

Author(s)

Georgi N. Boshnakov

mixAR documentation built on May 3, 2022, 5:08 p.m.