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"), ...)
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).
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) |
rpart.plot
data(ptitanic)
model <- rpart(survived ~ ., data = ptitanic, cp = .02)
rpart.plot(model)
rpart.rules(model)
Run the code above in your browser using DataLab