I'm a Romanian undergraduate in Mathematics at Louisiana State University where I also work as a research undergrad student in the Center for Computation & Technology. In the summers of 2006 and 2007 I had two internships at Lawrence Berkeley National Laboratory. There I was in the Visualization group working with Wes Bethel and Dr. Cecilia Aragon.
At the end of both summer programs I had to give two presentations. Below I have the abstracts and the links to my presentations.
2007 - Abstract for:
Information Visualization for Supernova Spectra
Scientific visualization, the study of graphical representation of complex physical phenomena, plays an important role in feature detection and in providing insight into the meaning of scientific results. Information visualization, on the other hand, focuses on graphically displaying abstract data. The combination of the two techniques can be a powerful tool to aid scientists in browsing and analyzing large scientific datasets.
Information visualization techniques such as overview (context) and detail information (focus) are best performed simultaneously in order to facilitate human understanding of large amounts of information displayed on a computer screen of limited size.
For this summer project, we have designed and implemented an interactive visual interface that uses fisheye visualization -- a "focus plus context" technique -- to display and browse supernova data. The fisheye lens approach to the visual browser provides the scientists with a tool that aids in the exploration and analysis of spectral data, while combining standard information visualization techniques such as filtering, highlighting, and a zoomable user interface.
The visual browser was developed using Java and Piccolo, a toolkit providing support for 2D graphics and the development of zoomable user interfaces.
Download the final presentation.
Download the "minutes" of my presentation.
A summary of my information visualization project
Download the tarball of the demo.
Official website of the SpectraVis project.
2006 - Abstract for:
Visualization and Data Analysis with VisIt and Matlab
Scientific visualization plays an important role in feature detection and in providing insight about the meaning of the scientific results. Visualization can also be used as a tool for finding and diagnosing error in computer applications.
We present some methods to visualize and analyze combustion data on unstructured meshes and of spectra from supernova simulations. For this we have exploited the features of two interactive software applications: VisIt and Matlab.
VisIt is a free, parallel, visualization tool used to visualize and_ analyze data defined on two or three dimensional structured and unstructured meshes; it provides a varied set of visualization operators and plot types. For data analysis of supernova simulations we used Matlab, a useful commercial tool for matrix computations.
We have designed and implemented VisIt plugins to analyze and create images of combustion data on unstructured meshes. We have investigated methods and algorithms to make a comparative analysis between templates and observational data of supernovae using both applications. We made a performance test to evaluate the speedup gained by using the parallel engine provided by VisIt.
In addition we developed code to process the data and prepare it for visualization.
Finally we present results of this work showing images, animations and histograms obtained from different data sets.
Download the final presentation
The 2006 talk was delivered both at the Lawrence Berkeley Lab and the NERSC facility in Oakland.