Get ampt, bro!
This project aims to provide some instructions and examples for building animated visualizations of spatiotemporal data.
A Day of Tweets
Video: http://youtu.be/WpuYzFmEPrE?hd=1
Code: https://github.com/jaredwinick/ampt/tree/master/examples/ampt-Tweets
2011 Tornado and Hail Activity
Video: http://youtu.be/R3-4e9a56qM?hd=1
Code: https://github.com/jaredwinick/ampt/tree/master/examples/ampt-Weather-2011
A Week of Tweets - New York City
Video: http://www.youtube.com/watch?v=bzWK4BgVYtQ&hd=1
These instructions need more detail, but hopefully combined with the examples, it is enough to get you started.
- Find some data that has both temporal and spatial features.
- Format your data in a CSV file with the a header identifying the columns. The temporal feature should be formatted as a unix timestamp and given a column header of timestamp. The spatial features should be a latitude and longitude with headers lat and lon respectively.
- Create a project in Tilemill and add your CSV file as a layer.
- Design your map.
- Copy your Tilemill project directory
- Copy the get_ampt.py script template to your new project directory
- Fill in your values for all the constants in get_ampt.py
- Update the copied style.mss to add in conditioning that will enable only data within a given timeframe to be shown for a video frame. Add [timestamp>=%d][timestamp<%d] next to the id controlling the styling for your data layer.
- Edit the project.mml and change style.mss to style_modified.mss