[R] obtaining p-values for lm.ridge() coefficients (package 'MASS')

[Michael Friendly]

On 8/23/2011 3:35 AM, Liviu Andronic wrote:
[snip]
>
> But how does one obtain the customary 'lm' summary information for the
> model above? I tried supplying the chosen lambda to Design::ols()
> using its 'penalty' argument, but strangely the results differ. See
> below.
>
>> require(Design)
>> Design::ols(Employed ~ GNP.deflator + GNP + Unemployed + Armed.Forces + Population + Year,data=longley,penalty = 0.03,tol=1e-12)
>

First, you should be using rms::ols, as Design is old.
Second, penalty in ols() is not the same as the ridge constant in 
lm.ridge, but rather a penalty on the log likelihood.  The documentation
for ols refers to ?lrm for the description.

Third, other ridge estimators (e.g., ElemStatLearn::simple.ridge also
give slightly different results for the same data due to differences in 
the way scaling is handled.

hth,
-Michael

-- 
Michael Friendly     Email: friendly AT yorku DOT ca
Professor, Psychology Dept.
York University      Voice: 416 736-5115 x66249 Fax: 416 736-5814
4700 Keele Street    Web:   http://www.datavis.ca
Toronto, ONT  M3J 1P3 CANADA