Hi, After getting the PLS model, if one needs to know, the F-statistic and the estimate of std dev of the model, how could one go about it? Especially, when no of independent variables are greater than the no of observations. In normal Multiple Linear Regression models, F-stat will be (r^2(n-p-1))/(p*(1-r^2)) est of std dev will be sqrt(sum(Ycalc - Yobs)^2/(n-p-1)) My question is what will be p in this case, becoz if p > n, there is no valid answer for it?
thanks. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
