After a visit to the ER, I discovered an ocean of personal data: more than 3 million rows of data about one of my favorite subjects: me. My watch averages almost 5,000 measurements a day. That experience led me to explore the Apple Health Export in more detail. It’s the largest dataset I’ve ever worked with. The variety of sources and the quantity data can be overwhelming. I will describe the basic measurements that are available and some of the things one needs to be on the lookout for. A ggplot of a full day of heart rate data provided me (and my doctor) important context on two occasions when I went to the ER. Fortunately there’s no great medical drama involved, but the plots are pretty neat and actually turned out to be helpful.
Talk materials are available at https://www.johngoldin.com/blog/2022-07-notes-apple-health-export/.
John Goldin spent his career as a data analyst for the Yale University administration. He was the Director of Institutional Research for over 20 years until retirement in 2011. Since then he has maintained some involvement with his old office, although each year he becomes more thoroughly retired. Even in retirement he finds plenty of data just begging to be analyzed. He was a teaching assistant for Lee Wilkinson (the author of The Grammar of Graphics) back in the 1970’s when they were both graduate students in Psychology at Yale.
