Published on

Daytime Nap Prediction Using Wearable Data

(in progress)

What makes me nap during the day?

In 2016, in order to keep up with school work and study for the secondary school national examinations, I forced myself to sleep less than 6.5 hours a day. Despite trying to sleep more in high school and college, I realized that I would always wake up at around the 6-hour mark. Even if I managed to sleep 7 hours or more, I always woke up feeling more exhausted than the day before; since I had control over my time, this sleepiness also led me to nap during the day.

I finally got a smartwatch in January 2024 to investigate. This repository contains my analysis, visualization, and interpretation of my sleep data using Python.

Data Description

data/shr_v_temp_phase.csv

Heart rate (HR), Heart rate variability (HRV)

I've worn my Samsung Galaxy Watch 6 every night since I got it in January 2024. However, it wasn't until the Samsung Health app introduced "energy scores" on December 17, 2024 that I got easy access to my sleeping HR and HRV data.

Period & Basal Body Temperature data

I only have basal body temperature data from August 2024, and I used Samsung Health's ovulation predictions and manually-tracked period dates to determine the menstrual cycle phase for each day (follicular VS luteal).

data/daytime_naps.csv

Contains the daytime nap sessions: sleep sessions that 1) start between 6am and 6pm, and 2) last for more than 30mins), recorded by Samsung Galazy Watch 6.

data/nighttime_sleep.csv

Contains the nighttime sleep sessions recorded by Samsung Galazy Watch 6.