R: lts regression: $ operator is invalid for atomic vectors

Question

I'm trying to do a variable selection using LTS regression but I encounter with this error.

sigma.full<-summary(ltsreg(y~x1+x2+x3+x4+x5))$scale

Error in summary(ltsreg(y~x1+x2+x3+x4+x5))$scale: $ operator is invalid for atomic vectors

Can someone help me to resolve this error?

reproducible example please? What package are you using (and what version)? Why do you think that the result of `summary(ltsreg(...))` is a list with a `scale` element? — Ben Bolker, Dec 28 '14 at 20:18

Rentrop · Answer 1 · 2014-12-28T20:23:10.080

3

Assuming you are using the function MASS::ltsreg

Use

sigma.full<-ltsreg(y~x1+x2+x3+x4+x5)$scale

Have a look at ?ltsreg.

The result of ltsreg(y~x1+x2+x3+x4+x5) itself has the scale-information. The summary of it does not have a $scale element.

Have a look at str(summary(ltsreg(...))) as Roland suggests

edited Dec 28 '14 at 20:23

answered Dec 28 '14 at 20:18

Rentrop

1 Answers1