Prediction Method for class discRob

DESCRIPTION:

Predict group membership ob observations using a discriminant function estimated by the discRob function.

USAGE:

predict.discRob(object, newdata, doubt=0, digits=.Options$digits)

REQUIRED ARGUMENTS:

object
the discRob object constructed by the discRob function.

OPTIONAL ARGUMENTS:

newdata
a data frame containing the values at which predictions are required. This argument can be missing, in which case predictions are made for the training data used to compute the discriminant function. Only those variables used to compute the discriminant function need be present by name in newdata.
doubt
a numerical variable, 0 <= 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.
digits
the number of significant digits to retain in the posterior probability. Rounding the posterior probabilities is odne mostly for printing purposes.

VALUE:

an n x (g+1) data frame, where n is the number of observations and g is the number of groups, containing a group factor column followed by the posterior probability of group membership for each observation in the newdata data frame or the training data.

DETAILS:

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.