Displaying 1 result from an estimated 1 matches for "duhray2".
Did you mean:
duhray
2003 Jul 21
0
Changing the labels on a regression tree (repeat post - with added clarity)
...ed indicated that my post was unclear....
I have a regression tree created in rpart with
tr.logypsx <- rpart(log(YPSX + 1)
~AGE+drugfact+sexfact+as.numeric(OBSX) +WINDLE + EABUSED + PABAU +
positive.par + control.par + lenient.par, xval = 10, method = 'anova',
cp = 0.0001, data = duhray2)
and then
tr.logypsx.pruned <- prune(tr.logypsx, cp = .012)
The labels at the nodes are, naturally enough, the dependent variable
[log(YPSX + 1)]. This DV is better than YPSX because YPSX is extremely
skew. However, the intuitively meaningful variable is YPSX. Is there a
way to change the...