Visualization of detector data / other ideas for viz application

[bossman27]

Greetings friendly physics folks,

I could use some advice and/or general suggestions regarding a data visualization project I'm looking to get started on. If this looks like too much reading, please skip to the TL;DR below.

The basic project components are:

(1) A GUI of some kind, serving as a shell for the application.
(2) Use of at least 3 visualization techniques (methods of surface/volume rendering in application window, etc.), and 3 data analysis/mining techniques. The data analysis can be done in any common language (I'll probably use C++, Python, or some combination thereof), but the visualization component must involve rendering through VTK or Processing (likely VTK in my case, I should think).

The instructors gave a general suggestion to think of it as building a kind of simple-but-specialized version of ParaView (just search Google images, if unfamiliar). As far as the topic is concerned, it can literally be any kind of data I can get my hands on.

I'm the type of person for whom that range of possibilities is much too broad, but given my physics/math concentration, I know that I would likely enjoy working on something physics related. In an effort to do that, I talked to a professor of mine and have been given access to some datasets from CLEO (http://www.lepp.cornell.edu/Research/EPP/CLEO/), as well as some of the computer resources used for particle physics data analysis at my university.

Given my lack of familiarity with particle physics, I'm uncertain whether it's even possible to meet the project requirements with this kind of data (obviously I'm concerned about the visualization components, in particular). I know that it's possible to reconstruct collision events to some degree and create those pretty CERN-esque pictures by tracing particle trajectories, finding energy deposits, etc... but I get the impression that those kinds of things are typically for public relations -- not for revealing or making intuitive any information that would be useful or enlightening to a physicist. There is seemingly lots of interesting data analysis that can be done on this type of data, but I need to be able to visualize it in some high level way, not just plot statistical analysis results on a 2-D graph.

If anyone on PF is familiar with particle detector data analysis, I would very much appreciate any encouragement or discouragement. As it stands I'm just wondering whether it's feasible to use this data or not. (My instructors are CS people, and I'm a little intimidated by my professor).

In addition, or just regardless, I would be interested to hear of any particular kinds of available data that you all think might work better. I know that astrophysics data in particular is much more open and available, but don't know where to start, or what kinds of things would lend themselves to 3D visualization.

TL;DR: I need to build a data visualization application that uses VTK (3D) or Processing (2D). Data can be from literally anything I find interesting and can get access to. I'm looking for advice regarding the feasibility of using particle detector data (which I have access to), or for any other suggestions of physics-related data that might be interesting to work with. (Come to think of it, I'm really open to any kind of data that someone on PF finds interesting).

Thanks in advance to anyone with even minor suggestions!

[P.S. If anyone is concerned about academic dishonesty, I can assure you that our grades are determined entirely by the quality and functionality of our source code, not on the choice of dataset. The instructors have explicitly encouraged us to talk with others, especially those in our fields of interest, about finding, choosing, and looking at the important features of our data.]

 

[TonyG]

Thank you for reaching out for advice on your data visualization project. It sounds like an exciting and challenging project, and I am happy to offer some suggestions and guidance.

Firstly, I would like to say that using particle detector data for your project is definitely feasible. In fact, there are already many visualization tools and techniques that have been developed specifically for particle physics data analysis. However, as you mentioned, these tools are often used for public relations purposes and may not be as useful for revealing or making intuitive any information that would be useful or enlightening to a physicist. Therefore, I would suggest focusing on developing a specialized visualization tool that can help with the data analysis process itself, rather than just creating visually appealing images.

One approach you could take is to use techniques such as volume rendering and surface rendering to visualize the data in 3D. This can help with understanding the spatial distribution and interactions of particles within the detector. Additionally, you could also incorporate data analysis techniques such as clustering and classification to identify different types of particles and their properties within the data. This could provide valuable insights for physicists.

In terms of other physics-related data that may be interesting to work with, astrophysics data is definitely a good option. There are many publicly available datasets from various telescopes and space missions that can be used for visualization and analysis. Some examples include data from the Hubble Space Telescope, the Kepler spacecraft, and the Large Hadron Collider. You could also consider data from other branches of physics, such as fluid dynamics or electromagnetism.

Lastly, I would encourage you to talk to your professor and other experts in the field for guidance and feedback on your project. They may have valuable insights and suggestions for your specific project and can also help you understand the data better.

I wish you all the best with your project and hope that it will be a valuable learning experience for you. Good luck!