This is a really self-selected data set. If you look at the New York map you'd think people didn't exist in most of Brooklyn and Queens.
Ditto Vancouver where I originally come from - lots of activity downtown and in rich, young neighborhoods, but nearly zero activity in middle class or older neighborhoods.
It's not just selecting for iPhone users, it's selecting for wealthy, young, iPhone users. On a more meta level these maps are an interesting proxy for wealth, race, and age.
The data is coming from the subset of people who have iPhones. So you can have 30 million people walking, but if only a few are rich enough to have an iPhone, then they are invisible.
Ditto Vancouver where I originally come from - lots of activity downtown and in rich, young neighborhoods, but nearly zero activity in middle class or older neighborhoods.
It's not just selecting for iPhone users, it's selecting for wealthy, young, iPhone users. On a more meta level these maps are an interesting proxy for wealth, race, and age.