dimension reduction

projection.

(x1,x2,x3,x4,x5,...xn) -> (x1,x2,x3)

You're welcome.

If the loss of information bothers you, you could first perform principal component analysis on the data set, which would give you a rotation matrix that would allow you to rotate your data set in such a manner that the component with the greatest variance lies along the first axis, and the next greatest variance along the next axis and so on. Then when you truncate by projection you keep the greatest variance in your final data. You still lose information but that is kind of in the nature of reduction of dimensionality.

Impossible to say if PCA makes sense for what you are doing since you were rather coy about giving even the merest hint about what you are trying to do.

marlin314a at 2007-7-14 22:05:07 >