Real-time dynamic data display
Displaying data from observations of dynamic phenomena isn’t easy. Normally, we select sequences in discrete time slices and look for trends implied in the static data. A lot is lost in the paired translations of dynamic-to-static and static-to-trends.
This was made clear to me recently while working with the Center for Advanced Manufacturing in Puget Sound (CAMPS) on wind energy applications. Looking into wind data visualization, I ran across an interesting real-time, animated map of wind data.
The work is a product of Fernanda Viegas and Martin Wattenberg at hint.fm and it is fascinating to watch. The complexity of the data is highlighted by the concurrent display of large regional wind patterns and chaos of smaller, local disturbances. By zooming in on the map, you can drill down to local patterns of specific interest.
After exploring the animated map, it is instructional to look at a more traditional display of the same data.
The contrast is surprising. It left me thinking about data reductions and analyses that I have done in the past and how much richer the resulting insights and conclusions might have been if I had looked at continuous data—continuously.