Sleep Regularity Index
This project contains a Jupyter notebook with code for calculating a Sleep Regularity Index (SRI) from a series of sleep records consisting of timestamps and durations (e.g. exported from Fitbit via Zenobase).
The calculated SRI values range from 0 (no overlap in sleep and wake times from one day to another) to 100 (the same sleep times every day). This method handles multiple intervals of sleep per day.
- There's no way to distinguish between missing data and "not sleeping".
- There's no special accommodation for daylight savings or timezone changes.