Using a timeline to correlate data on events

One of the beefs I have with Apple’s Health app whether the data is displayed on my iPhone or iPad is the lack of any correlation between one set of data points and another, e.g. medication taken and incidents of some known side effect. There is too much data to investigate this manually; my most recent export of all data from the Health App was 42MB covering many years and the export of similar data from my Withings Blood Pressure Monitor was 9MB covering just the last two weeks or so.

It crossed my mind that a simple solution might be plotting these different sets of data points along a timeline. Which leads me to ponder whether anyone has ever tried that idea perhaps using a tool like Aeon Timeline, but it needs granualarity down to a minute, or by hacking up some script in R and then outputing the timeline from a dataset. As a fan of Edward Tufte’s wonder if his sparklines Edward Tufte forum: Sparkline theory and practice Edward Tufte might provide the first hints on the correlates.

One additional challenge for visualization might be that your data isn’t all at the same resolution, e.g. minute (or higher) for heart rate, daily (or weekly, or monthly) for blood pressure.

Also, are you looking for “realtime” effects (like blood sugar rising after a meal), or long-term trends (e.g. how certain medications do or don’t affect blood pressure)?

Finding relations between variables in time series - Personal Science Wiki There is an easier version on the wiki too.

1 Like