Re: [R] singular.ok in lm

2008-01-24 Thread Prof Brian Ripley
On Thu, 24 Jan 2008, Martin Eklund wrote: > Dear R users, > > I'm wondering about the singular.ok option in lm. If singular.ok is > set to TRUE does it mean that we allow the generalized inverse of > (X'X)^-1 (where X are the independent variables and ' denotes the > transpose)? Or is the singular

[R] singular.ok in lm

2008-01-24 Thread Martin Eklund
Dear R users, I'm wondering about the singular.ok option in lm. If singular.ok is set to TRUE does it mean that we allow the generalized inverse of (X'X)^-1 (where X are the independent variables and ' denotes the transpose)? Or is the singularity handled in some other way? Thank you! Best