Scientific Visualization

Imagine that you want to simulate a fluid mixture that is phase-separating. For scientific reasons you may be interested in some statistical properties of this process, but how can you be sure that the program is really doing what you expect it to do? And what do you do if you want to find out what the effect of different parameters is? There will often be quite a few different parameters that you need to control.

My vision for this library is that it will help scientists to transform their simulations which are performing their instruction in a way that is hard to observe, into a transparent object where you can look at all times at the relevant variables and where you can change the parameters to see what will happen. This changes the nature of computational research: suddenly you are able to see and manipulate the system you simulate directly. You will now be able to see what is going on in the system, and if something does behave differently than you expect you have probably found a bug in your program and you can eliminate it faster. And if it isn't a bug you have just learned something interesting about the system. And once in a while you will be the first to notice and you have made a discovery that you might not have made without being able to observe the system so closely (this is what happened to me a lot of the time).

This is not to suggest that you give up the more traditional approach of defining curves for which you can make analytical predictions or which you can compare to experiments. But I feel that it is a very valuable extension of this approach.