Eureqa!

[mchargue]

Folks;

I recently read that you folks are compiling a database of experimental results. I suspect that you are looking to determine how variations in the conditions from shot to shot affect the quality of the result.

If that’s the case, I’d like to at least make you aware of a free program, put out by Cornell Creative Machines, that may be able to assist you with discovering how changes in conditions affect the outcome: Eureqa.

http://creativemachines.cornell.edu/eureqa

You can read about the program, its application, and other results at their web site. (above)

As I understand it, the program attempts to analyze experimental results in order to fit a function, with minimal variables, whose evaluation can predict the outcome of an experiment. In so doing, the program may tease out unexpected relationships between an experiments setup, and its results. It may also give a sense of what variables are most important in achieving a desired result.

FYI

Patrick

[Brian H]

Very interesting! I think Occam would have approved.

[pulser]

JMP 9(by SAS) is a commercial data analysis package. I use it daily and it is good for DOE design, DOE analysis, multi-variate linear and non-linear modeling and much more. At least with so many users and more than 15 years in the field many resources have been put behind it to assure validity and functionality and they are always adding more capabilities. Works on a PC, Mac, or Linux.

[Author]

Posts