iViz: The Immersive, Multi-Dimensional Visualization Platform

Jan 13, 2014

(It's actually dramatically simpler than it sounds)

Paper, Presentation

I spent my summer at Caltech working with Unity3D, a technology that is used to make games, except no games were made. I lead the development to extend and further iViz, a program used to plot points in 3D space with up to ten layers of information.

Information is encoded in these ways:

  • XYZ – coordinate positions
  • RGBA – red, green, blue, alpha, color levels
  • Size – radius of the point
  • shape – sphere, cube, pyramid, cone, cylinder, wedge, and torus are all shapes that a point could be
  • texture – an image that can be applied to the surface of the shape

The purpose of this is to allow the user to control a camera that flies through 3D space to explore data visually. Rather than write computer algorithms to try to cluster data, iViz relies on one of the oldest and most reliable clustering algorithms: the human brain. Trends and relationships can be gathered by exploring the data. In the program, it’s easy to switch certain properties to be reflected by a different dimension to improve experimentation.

People may call today the age of collecting data but the next era is the age of understanding data. iViz can hopefully help with this goal. There are still features to add and interfaces such as the Oculus Rift to experiment with. There are still some problems such as large data sets simply looking incoherent, or the ability of outliers to skew the data. Hopefully these can be addressed in the future, but I’m proud of what’s been done so far.

Credit is due to Professor Djorgovski at the California Institute of Technology for guiding and initiating this idea and Ciro Donalek for advising me on this research project. I am honored to have worked on it.