[R] better way to iterate matrix multiplication?

[Ben Bolker]

Eik Vettorazzi <E.Vettorazzi <at> uke.uni-hamburg.de> writes:

> 
> if you have a homogeneous mc (= a constant transition matrix), your
> state at time 10 is given by (chapman-kolmogorov)
> p10=p0 %*% tm^(10)
> so you need a matrix power function.

  There are matrix exponential functions in the Matrix and expm
packages ... don't know about their speed