Unformatted text preview: lass tree, some of the tree
functions are used by the rpart methods, e.g. tree.depth. Tree methods that have
not been implemented as of yet include burl.tree, cv.tree, hist.tree, rug.tree, and
Not all of the plotting functions available for tree objects have been implemented for rpart objects. However, many can be accessed by using the as.tree
function, which reformats the components of an rpart object into a tree object.
The resultant value may not work for all operations, however. The tree functions
ignore the method component of the result, instead they assume that
if y is a factor, then classi cation was performed based on the information
51 otherwise, anova splitting regression was done.
Thus the result of as.tree from a Poisson t will work reasonably for some plotting
functions, e.g., hist.tree, but would not make sense for functions that do further
computations such as burl.tree. 12 Source
The software exists in two forms: a stand-alone version, which can be found in
statlib in the `general' section, and an S version, also on stalib, but in the `S' section
of the library. The splitting rules and other computations exist in both versions,
but the S version has been enhanced with several graphics options, most of which
are modeled copied actually on the tree functions. References
1 L. Breiman, J.H. Friedman, R.A. Olshen, , and C.J Stone. Classi cation and
Regression Trees. Wadsworth, Belmont, Ca, 1983.
2 L.A. Clark and D. Pregibon. Tree-based models. In J.M. Chambers and T.J.
Hastie, editors, Statistical Models in S, chapter 9. Wadsworth and Brooks Cole,
Paci c Grove, Ca, 1992.
3 M. LeBlanc and J Crowley. Relative risk trees for censored survival data. Biometrics, 48:411 425, 1992.
4 O. Nativ, Y. Raz, H.Z. Winkler, Y. Hosaka, E.T. Boyle, T.M. Therneau, G.M.
Farrow, R.P. Meyers, H. Zincke, and M.M Lieber. Prognostic value of ow
cytometric nuclear DNA analysis in stage C prostate carcinoma. Surgical Forum,
pages 685 687, 1988.
5 T.M. Therneau. A short introduction to recursive partitioning. Orion Technical
Report 21, Stanford University, Department of Statistics, 1983.
6 T.M. Therneau, Grambsch P.M., and T.R. Fleming. Martingale based residuals
for survival models. Biometrika, 77:147 160, 1990. 52...
View Full Document
- Fall '13
- Regression Analysis, Missing values