Function mult_vector_gelman_rubin (o2scl)

O2scl : Function List

template<class vec_t, class fp_t>
fp_t o2scl::mult_vector_gelman_rubin(std::vector<vec_t> &v, int verbose = 0)

Compute the Gelman-Rubin statistic for a set of vectors.

This function requires at least two vectors and all of the vectors provided must have a length at least greater than two.

An old recommendation by Gelman (2004) was that convergence is indicated by R<1.1, subsequent works found but tighter constraints may be required.

Gelman (2004) is Gelman, A., Carlin, J. B., Stern, H. S., and Rubin, D. B. (2004). Bayesian Data Analysis. Chapman & Hall/CRC, Boca Raton, FL,.

https://arxiv.org/pdf/1812.09384 may provide an interesting alternative.