On Mon, 2010-12-13 at 01:55 -0800, jagdeesh_mn wrote: > > Prof Brian Ripley wrote: <snip /> > > Thanks Mr. Brian. That kind of answers my query. > > On the same note I would like to ask few other questions. Sorry if you find > them naive, I am a novice in this subject and am trying to get a grip on > things. > > 1. I am using R package using my code and the fitted object looks like this > : > > The Model representation : > n= 60 > > node), split, n, deviance, yval > * denotes terminal node > > 1) root 60 983551500 12615.670 > 2) dataFrame[, 6]='Small' 13 21804710 7682.385 * > 3) dataFrame[, 6]='Compact','Large','Medium','Sporty','Van' 47 557851600 > 13980.190 > 6) dataFrame[, 3]='Japan/USA','Korea','USA' 29 131050000 12673.030 > 12) dataFrame[, 6]='Compact','Sporty' 14 11426050 11055.570 * > 13) dataFrame[, 6]='Large','Medium','Van' 15 48812470 14182.670 * > 7) dataFrame[, 3]='France','Germany','Japan','Sweden' 18 297418200 > 16086.170 * > > What does the term deviance here stand for?
At this point, go an read up on the theory of classification and regression trees. Depending on how you fitted your tree (what options used, what type of response modelled) the deviance could be computed in different ways. In short it is a measure of how impure each node of the tree is. See the References section of ?rpart HTH G > 2. Could you also suggest me some readings on the topic of CnR trees > specific to R with case studies? > > Regards, > Jagdeesh -- %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% Dr. Gavin Simpson [t] +44 (0)20 7679 0522 ECRC, UCL Geography, [f] +44 (0)20 7679 0565 Pearson Building, [e] gavin.simpsonATNOSPAMucl.ac.uk Gower Street, London [w] http://www.ucl.ac.uk/~ucfagls/ UK. WC1E 6BT. [w] http://www.freshwaters.org.uk %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.