It is always important to visualize data in a manner which tells a story. Often times this means carefully designing a visualization or set of visualizations to which tell the story with the data. Other times, presented in the right manner, data can tell a story on its own when confronted by a curious mind. Written in Python using the bokeh package, the first custom interactive presentation I built uses data from trail camera photo classification information to visualize the activity levels of various species.
This project eventually had a strong influence on future works at the Wisconsin Department of Natural Resources (DNR). While working for the Snapshot Wisconsin project at the DNR I realized there are is a lot of dirty data in this visualization which I didn’t quite notice at the time. Most noticeable are the spikes near noon (12pm). These spikes are partially an artifact of trail cameras resetting their time which will create large spikes near 12am or 12pm. There was also some sort of time conversion issue and many of the points which probably should have been between 0:00 and 1:00 were between 12:00 and 13:00 in the data instead.
Feel free to give it a whirl below!