Skip to main content
. 2019 Jul 9;9:9939. doi: 10.1038/s41598-019-45852-2

Figure 4.

Figure 4

(A) Representative profile of activity (where 1-min bins with activity > 0 shown as 1) across one light period between sunrise and sunset. The plot below shows reshuffled activity periods of the same recording, where the number and duration of short activity bouts are retained as above, but their timing of occurrence is randomised. (B) Survival analysis of inactivity (rest) episodes for empirical data and simulated activity profiles as above. Note that shorter rest bouts are more likely to occur in the simulated activity time profiles, while longer sustained rest periods occur more frequently in the real data. (C) The difference in the distribution of rest episodes derived from data and simulation, calculated as in (B). An additional interruption criterion is introduced from 0–5 min. Vertical lines depict minimal rest episode duration where the rest episode duration in data is below simulation for rest episodes where no interruption is allowed (black) and where up to 5-min brief awakenings are permitted (magenta). (D) Mean duration of rest episodes derived from empirical and simulated data sets, shown as a function of the interruption criterion. Mean values, n = 7, SEM.