Hello,
David is right, the response is named 'crp', not 'fastMM'. And there is
no need to run the regression twice.
Try
# function to obtain MSE
MSE <- function(data, indices, formula, method) {
d <- data[indices, ] # allows boot to select sample
fit <- lmrob(formula, data = d, method = method)
ypred <- predict(fit)
mean((d[["crp"]] - ypred)^2)
}
Hope this helps,
Rui Barradas
Às 15:56 de 30/03/20, David Winsemius escreveu:
On 3/30/20 2:45 AM, varin sacha wrote:
Many thanks David,
But I am not sure the Dataset and newdata are the real problem. I have
solved, at least I think so, what you were asking. Here below my new R
code but still the same error message. IMHO the problem is with
"boot.ci" function...
# # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # #
bmi=c(23,43,21,23,45,65,45,11,12,13,23,34,NA,NA,34,35,45,65,43,23,12,11,15,43,23,88,78,79,89,89,99,43,21,34,32,45,65,76,56,45,34,23,12,32)
glucose=c(NA,12,23,11,12,21,23,21,23,43,23,12,NA,23,11,12,32,12,14,12,11,10,9,8,9,8,7,90,76,32,12,11,12,23,11,123,32,12,14,34,54,65,76,87)
crp=c(123,212,154,342,123,111,121,765,453,123,213,211,NA,NA,32,123,213,145,143,123,132,143,165,176,181,123,87,567,342,123,143,132,143,234,345,32,123,132,143,345,321,543,231,123)
age=c(67,45,34,56,87,NA,NA,23,18,65,45,87,65,33,23,65,43,23,43,12,132,56,76,87,98,78,76,56,78,54,34,56,76,99,12,45,34,65,76,87,98,97,64,53)
sex=c(0,1,1,0,1,0,1,0,0,1,1,1,NA,NA,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,0,0,1,1,1,0,0,0,1,1,0,0)
Dataset=data.frame( bmi,glucose,crp,age,sex)
newdata=na.omit(Dataset)
#install.packages( "robustbase",dependencies=TRUE )
#install.packages( "boot",dependencies=TRUE )
library(boot)
library(robustbase)
# function to obtain MSE
MSE <- function(data, indices, formula,method) {
d <- data[indices, ] # allows boot to select sample
fastMM <- lmrob( crp ~ bmi+glucose+age+sex,data=newdata)
fit <- lmrob(formula, data = d, method="MM")
ypred <- predict(fit)
mean((d[["fastMM"]]-ypred)^2)
At this point I do not see any reason for assuming there is a column in
the d-object with the name "fastMM".
______________________________________________
R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
