arbor
object.
predict.arbor(object, newdata=list(), type=c("vector", "matrix", "tree", "class", "prob"))
arbor
.
This is assumed to be the result of some function that produces an object
with the same named components as that returned
by the
arbor
function.
formula(object)
must
be present by name in
newdata
. If missing, the fitted values are returned.
matrix
for longitudinal data and
vector
for all other data.
type="vector"
:
vector of predicted responses.
if
type="matrix"
:
either a matrix of predicted class probabilities and class counts
(for classification problem),
or number of events at a node (for poisson or exponential methods),
along with the predicted responses
or predicted responses for longitudinal data.
If the input object does not have a yval2 (i.e. anova method was used)
then the vector of predicted responses is returned.
if
type="tree"
:
an object of class
arbor
with new values for
frame$n
and
frame$yval
(and
frame$yprob
if it exists). This options
does not currently work.
if
type="class"
:
vector of predicted factor responses, if method is classification.
if
type="prob"
:
returns a vector or a matrix (as appropriate) of class probabilities.
This function is a method for the generic function predict for class
arbor
. It can be invoked by calling predict for an object of the
appropriate class, or directly by calling
predict.arbor
regardless of
the class of the object.
The new object is obtained by dropping newdata down object.
For factor predictors,
if an observation contains a level not used to grow the tree,
it is left at the deepest possible node
and
frame$yval
at the node is the prediction.
z.auto <- arbor(Mileage ~ Weight, car.test.frame) predict(z.auto) # To obtain the response as factors: fit1 <- arbor(Kyphosis ~ Age + Number + Start, data=kyphosis) ylevels <- attr(fit1, 'ylevels') predFactors <- factor(ylevels[fit1$frame$yval[fit1$where]], levels=ylevels) names(predFactors) <- names(fit1$where) predFactors # this will be the same as predict(fit1, type='class')