rpart.plot (version 3.1.0)

rpart.rules: Print an rpart model as a set of rules.


Print an rpart model as a set of rules.


rpart.rules(x = stop("no 'x' argument"),
            style = "wide", cover = FALSE, nn = FALSE,
            roundint = TRUE, clip.facs = FALSE,
            varorder = NULL, ...)

# S3 method for rpart.rules print(x = stop("no 'x' argument"), style = attr(x, "style"), ...)



An rpart object. The only required argument.


One of: "wide" (default) One rule per line. May require a lot of horizontal space. "tall" One split per line. "tallw" Like "tall" but with more horizontal white space for readability.


Default FALSE. If TRUE, also print the percentage of cases covered by each rule.


Default FALSE. If TRUE, also print the leaf node number for each rule.


If roundint=TRUE (default) and all values of a predictor in the training data are integers, then splits for that predictor are rounded to integer. For example, display nsiblings < 3 instead of nsiblings < 2.5. Identical to the argument of the same name in rpart.plot, see there for details.


Default FALSE. If TRUE, print splits on factors as female instead of sex = female; the variable name and equals is dropped.

Identical to the argument of the same name in rpart.plot.


By default, the variables in the rules are ordered left to right on importance, where the ``importance'' of a variable here is the number of rules it appears in. Use varorder to force variables to appear first in the rules. For example varorder="sex" or varorder=c("sex", "pclass") will put the specified variables first. Partial matching of variable names is supported.


The following can be passed as dot arguments. See prp for details on these arguments.

argument default
extra = "auto" a subset of the legal values for prp are supported
digits = 2 default is two digits of accuracy, increase if necessary
varlen = 0 default displays full variable names
faclen = 0 default displays full factor names
trace = 0
facsep = " or "
eq = " is "
lt = " < "
ge = " >= "
and = " & "
when = " when "
because = " because " used only by rpart.predict
null.model = "null model" for root-only models (no splits)
response.name = NULL the response name printed before the rules (NULL means automatic)


A data.frame of class c("rpart.rules", "data.frame") with some attached attributes which are passed on to print.rpart.rules. Note that print.rpart.rules temporarily increases options(width).

See Also



model <- rpart(survived ~ ., data = ptitanic, cp = .02)
# }