predict.wlda: Classify Multivariate Observations Based on Weighted Linear...
In locClass: Collection of Local Classification Methods

Description Usage Arguments Details Value See Also Examples

Classify multivariate observations in conjunction with wlda.

1
2
3

  ## S3 method for class 'wlda'
 predict(object, newdata,
    prior = object$prior, ...)

`object`	Object of class `"wlda"`.
`newdata`	A `data.frame` of cases to be classified or, if `object` has a `formula`, a `data.frame` with columns of the same names as the variables used. A vector will be interpreted as a row vector. If `newdata` is missing, an attempt will be made to retrieve the data used to fit the `wlda` object.
`prior`	The class prior probabilities. By default the proportions in the training data set.
`...`	Further arguments.

This function is a method for the generic function predict() for class "wlda". It can be invoked by calling predict(x) for an object x of the appropriate class, or directly by calling predict.wlda(x) regardless of the class of the object.

A list with components:

`class`	The predicted class labels (a `factor`).
`posterior`	Matrix of class posterior probabilities.

wlda.

library(mlbench)
data(PimaIndiansDiabetes)

train <- sample(nrow(PimaIndiansDiabetes), 500)

# weighting observations from classes pos and neg according to their
# frequency in the data set:
ws <- as.numeric(1/table(PimaIndiansDiabetes$diabetes)
    [PimaIndiansDiabetes$diabetes])

fit <- wlda(diabetes ~ ., data = PimaIndiansDiabetes, weights = ws,
    subset = train)
pred <- predict(fit, newdata = PimaIndiansDiabetes[-train,])
mean(pred$class != PimaIndiansDiabetes$diabetes[-train])