predict.enetLTS: make predictions from the '"enetLTS"' object.

In enetLTS: Robust and Sparse Methods for High Dimensional Linear and Binary and Multinomial Regression

make predictions from the "enetLTS" object.

Description

Similar to other predict methods, this function predicts fitted values, logits, coefficients and nonzero coefficients from a fitted "enetLTS" object.

Usage

## S3 method for class 'enetLTS'
predict(object,newX,vers=c("reweighted","raw"),
    type=c("link","response","coefficients","nonzero","class"),...)

Arguments

object	the model fit from which to make predictions.
newX	new values for the predictor matrix X. Must be a matrix; can be sparse as in Matrix package. This argument is not used for type=c("coefficients","nonzero").
vers	a character string denoting which fit to use for the predictions. Possible values are "reweighted" (the default) for predicting values from the reweighted fit, "raw" for predicting values from the raw fit.
type	type of prediction required. type="link" gives the link function. type="response" gives the fitted probabilities for "binomial" and gives the fitted values for "gaussian". type="coefficients" computes the coefficients from the fitted model. type="nonzero" returns a list of the indices of the nonzero coefficients. type="class" is available only for "binomial" model, and produces the class label corresponding to the maximum probability.
...	additional arguments from the enetLTS object if needed.

Details

The newdata argument defaults to the matrix of predictors used to fit the model such that the fitted values are computed.

coef.enetLTS(...) is equivalent to predict.enetLTS(object,newX,type="coefficients",...), where newX argument is the matrix as in enetLTS.

Value

The requested predicted values are returned.

Author(s)

Fatma Sevinc KURNAZ, Irene HOFFMANN, Peter FILZMOSER
Maintainer: Fatma Sevinc KURNAZ <[e-mail]>;<[e-mail]>

See Also

enetLTS, coef.enetLTS, nonzeroCoef.enetLTS

Examples

## for gaussian

set.seed(86)
n <- 100; p <- 25                             # number of observations and variables
beta <- rep(0,p); beta[1:6] <- 1              # 10% nonzero coefficients
sigma <- 0.5                                  # controls signal-to-noise ratio
x <- matrix(rnorm(n*p, sigma),nrow=n)
e <- rnorm(n,0,1)                             # error terms
eps <- 0.1                                    # contamination level
m <- ceiling(eps*n)                           # observations to be contaminated
eout <- e; eout[1:m] <- eout[1:m] + 10        # vertical outliers
yout <- c(x %*% beta + sigma * eout)        # response
xout <- x; xout[1:m,] <- xout[1:m,] + 10      # bad leverage points


fit1 <- enetLTS(xout,yout)
predict(fit1,newX=xout)
predict(fit1,newX=xout,type="coefficients")
predict(fit1,newX=xout,type="nonzero",vers="raw")
# provide new X matrix
newX <- matrix(rnorm(n*p, sigma),nrow=n)
predict(fit1,newX=newX,type="response")
predict(fit1,newX=newX,type="coefficients")
predict(fit1,newX=newX,type="nonzero")

## for binomial

eps <-0.05                                     # %10 contamination to only class 0
m <- ceiling(eps*n)
y <- sample(0:1,n,replace=TRUE)
xout <- x
xout[y==0,][1:m,] <- xout[1:m,] + 10;          # class 0
yout <- y                                      # wrong classification for vertical outliers


fit2 <- enetLTS(xout,yout,family="binomial")
predict(fit2,newX=xout)
predict(fit2,newX=xout,type="coefficients")
predict(fit2,newX=xout,type="nonzero",vers="raw")
predict(fit2,newX=newX,type="response")
predict(fit2,newX=newX,type="class")
predict(fit2,newX=newX,type="coefficients",vers="raw")
predict(fit2,newX=newX,type="nonzero")



## for multinomial

n <- 120; p <- 15
NC <- 3
X <- matrix(rnorm(n * p), n, p)
betas <- matrix(1:NC, ncol=NC, nrow=p, byrow=TRUE)
betas[(p-5):p,]=0; betas <- rbind(rep(0,NC),betas)
lv <- cbind(1,X) %*% betas
probs <- exp(lv)/apply(exp(lv),1,sum)
y <- apply(probs,1,function(prob){sample(1:NC, 1, TRUE, prob)})
xout <- X
eps <-0.05                          # %10 contamination to only class 0
m <- ceiling(eps*n)
xout[1:m,] <- xout[1:m,] + 10       # bad leverage points
yout <- y


fit3    <- enetLTS(xout,yout,family="multinomial")
predict(fit3,newX=xout)
predict(fit3,newX=xout,type="coefficients")
predict(fit3,newX=xout,type="nonzero",vers="raw")
predict(fit3,newX=xout,type="response")
predict(fit3,newX=xout,type="class")
predict(fit3,newX=xout,type="coefficients",vers="raw")
predict(fit3,newX=xout,type="nonzero")

enetLTS documentation built on May 22, 2022, 1:05 a.m.