Fig. 5From: Optimal choice of word length when comparing two Markov sequences using a χ 2-statisticThe power loss of \(S_{k}^{R}\) based on different methods for estimating the order of Markov sequences based on NGS short reads. Panels are the same as in Fig. 4. a homogeneous sampling without errors, b homogeneous sampling with errors, c heterogeneous sampling without errors, and d heterogeneous sampling with errorsBack to article page