MLRMPA (version 1.0)

lmcv: LOOCV

Description

The Leave-One-Out (LOO) and Leave-Group-Out (LGO) Cross-Validation for Multiple Linear Regression.

Usage

lmcv(obj, ng = NULL, random = FALSE)

Arguments

obj

obj: the model of MLR

ng

ng: number of group, if missing, do LOO

random

random: logical, if TRUE, do random CV

Value

q.squared

cross-validation relation coefficient

SDEP

standard deviation of error of predictions

Details

The predictive ability of the developed model can be illustrated using leave-one-out (LOO) and leave-group-out (LGO) cross-validation procedures.

See Also

lm

Examples

u <- c(5,10,15,20,30,40,60,80,100)
lot1 <- c(118,58,42,35,27,25,23,20,18)
lot2 <- c(69,35,26,21,18,16,13,12,6)
data <- data.frame(cbind(u,lot1,lot2))
lm <- lm(u~.,data=data)
y.pre<-fitted(lm)
y<-u
Rsquared(y.pre,y)
summary(lm)$r.squared
LOOCV<-lmcv(lm)
LGOCV<-lmcv(lm,5)
print(LOOCV)
print(LGOCV)