I conat figure out what are the knows and what you want to vary for your extremem principle. :I know there exists LSF algorithms that minimize criterion: : :\sum^{n}_{i=1}w_{i}[a_{i}-t_{i}]^{2} : Is this supposed to be the standard linear regresion critera? If so it should be more like: sum[ ( y_i(x_i) - t(x_i) )^2 ] where t(x) is linear t(x)=a x + b Kind of looks like what you have, but what is w? Some kind of weight function? -- george jefferson : [EMAIL PROTECTED] to reply simply press "r" -- I hate editing addresses more than I hate the spam! =========================================================================== This list is open to everyone. Occasionally, people lacking respect for other members of the list send messages that are inappropriate or unrelated to the list's discussion topics. Please just delete the offensive email. For information concerning the list, please see the following web page: http://jse.stat.ncsu.edu/ ===========================================================================