v_function: Normalized v-function EM
In pneff93/HMM: Fitting a Hidden Markov Model

This function calculates the v-function which is part of the estimation step. Each element in v aligns with the conditional expectation to reach state k from the previous state j given your datapoint x(t). Due to the normalizing of the factors we also need to include the corresponding weights into our function.

1	v_function(m, N, beta, p, weight, Gamma, set, u)

`m`	number of likelihoods
`N`	length of the supplied dataset
`beta`	beta matrix
`p`	vector of likelihood probabilities of dataset
`weight`	weights provided by the alpha calculation
`Gamma`	Gamma matrix
`set`	index vector to align the p vector
`u`	output matrix of the u_function

returns the coresponding values, but due to further calculation steps the m x m matrix for each timepoint of v-values is returned as one row in the output matrix.

pneff93/HMM documentation built on Oct. 26, 2019, 8:16 a.m.