find_change_points_JSD: Jensen-Shannon divergence algorithm

In aaaaana/changepoints:

Description

Recursive function. In a top down manner, it finds one change point for the whole sequence and then performs the same operation on the two resulting subsequences. It stops according to a model selection criterion, or when the smallest size of the sequence is reached.

Usage

find_change_points_JSD(chain, cutoff = 15, t_in = 1, t_fin = NULL)

Arguments

chain	= the (supposedly) piecewise Markov Chain on which to run the algorithm
cutoff	= smaller possible size of a chunk
t_in	= the starting position of the considered subsequence in the current call of the function. In the first call, 't_in = 1“ if you want to perform the algorithm on the whole sequence
t_fin	= the final position of the considered subsequence in the current call of the function. In the first call, 't_fin = NULL' if you want to perform the algorithm on the whole sequence. This forces t_fin = length(chain) in the function

Value

a 1-row matrix containing vector of change points. The first element of the matrix is always '1“, due to the recursive structure

aaaaana/changepoints documentation built on May 26, 2019, 1:35 p.m.