discRob
discRob
function.
predict.discRob(object, newdata, doubt=0, digits=.Options$digits)
discRob
object
constructed by the
discRob
function.
newdata
.
doubt
<
(g-1) / g
, where
g
is the number of groups,
defining the minimum difference the maximum posterior
probability of group membership must exceed all others
before assigning the observation to that group. If there
does not exist a posterior probability of a group
membership for an observation that exceeds
doubt + 1/g
, the observation
is assigned to an "unknown" group.
group
factor column followed by the posterior probability
of group membership for each observation in the
newdata
data frame
or the training data.
Predicts group membership for observations in the data
frame
newdata
using the
discriminant function contained in the
discRob
object
object
. It computes the posterior probability
of group membership for each observation, where the
prior probabilities of group membership are given by
the variable
prior
,
and assigns the observation to the group that has
the highest probability. This is the optimal allocation
rule, or Bayes rule. This
predict
method is based on the
predict.discrim
function.