Visualization

We view visualization as an integral part of Statistical methods and we are inspired by the vision of John Tukey for graphical display and analysis of data. As a fun fact, John Tukey is also universally credited with developing the Fast Fourier Transform (FFT), which is widely used in signal processing and analysis. With modern data sets that are enormous and highly complex there is an increase need for visualization both for data analysis and communication. To understand the enormity of the task consider the problem of visualizing, and understanding, multimodality brain imaging data from a longitudinal study of multiple sclerosis. The dimensionality, complexity, and size of the data make visualization hard, though not impossible. Our group has been developing visualization tools, which are constantly evolving, and we are moving in the direction of web and document embedding of manipulable 4-D images. 

Visualization is probably as close as we get to a form of art, as it is hard to quantify what type of plot will work for communicating information as unbiasedly as possible to the largest possible intended audience. Thus, visualization requires continuous experimentation, excellent computing skills, and a will to not be academically recognized for it. However, as we spend a lot of our time trying to understand and visualize data, "look through an ocean of noise to find the little coral reef", knowing how to look when one does not know where to look becomes increasingly important and, sometimes, essential.

We invite you to explore the many plots on this website, especially in the area of Scientific Areas of Interest. They are all produced by the members of our research group, who feel passionately about their work, the importance of careful plotting, and of integration between visualization, computations, and deep analytical skills.