Skip to main content
. 2016 Jul 28;27(5):1293–1305. doi: 10.1007/s11222-016-9687-5

Fig. 1.

Fig. 1

a The number of replications out of 100 in which using NMCD+ with varying NI results in the same results as NMCD without screening. b The number of changepoints detected with with increasing window size NI. c The proportion of true changepoints detected with varying window size NI. d The computational time (s) for NMCD+ with increasing window size NI