Writing notebooks that explore Moves

[gedankenstuecke]

-

Moves is a passive GPS application for iOS and Android that records you movement history. Besides logging your location it will also try to classify your mode of transportation. Around 100 people have already connected their Moves accounts to Open Humans and we have some public data sets as well.

Some ideas for notebooks that use Moves data:

• Visualize the movement patterns over time (ggplot2 for R makes this rather easy, but any other language works too!)
• Analyze as a time series:
  • Do certain movement types (driving a car, public transport, walking, …) fluctuate over time?
  • Are there seasonal changes to these modes of movement? Differences between weekday/weekends?
• advanced: Use the GPS data to get weather information from some 3rd party API and try to correlate if the workout frequency etc. is tied to the weather

[gedankenstuecke]

I wrote a first example in #11 that uses the Moves data to correlate the local weather with daily activity (as measured by Moves).

I find the Moves step data pretty poor though. It would be nice to use the superior step counts of Fitbit and/or Apple Health where possible.

[gedankenstuecke]

14 has a small try to make a cross-over between Moves and Twitter: Using the Weather as requested through the Moves Geo locations to estimate whether the weather impacts the twitter frequency & sentiment.