# Interpreting rpart output for decision trees?

How do I go about selecting the ideal location to use for pruning the tree here?

Or maybe someone can explain to me in simple language what this output means. I see that rel_error is constantly decreasing when I run this but x_error decreases until the 3rd point at .48 before rising again.

Which point do I use to pick the best length for the tree?

 library(MASS)
library(part)
z=rpart(crim~., data = Boston, method="anova", xval=10, cp=.0005)
printcp(z)
plotcp(z)