[R] cubic spline/smoother with nlme

[Peter Dalgaard]

Suzette Blanchard <suzette at sdac.harvard.edu> writes:

> Greetings,  I would like to use a cubic spline
> or smoother to model the fixed effects within
> nlme.  So far the only smoother I have been able
> to get to run successfully in nlme is smooth().
> 
> I tried smooth.spline:
>    fixed=list(lKa~1,lCL~smooth.spline(BSA, df=3))
>    the error I got was the following.
>    Error in model.frame(formula, rownames, variables, varnames, extras,
>    extranames,  :  invalid variable type
> 
> Can anyone suggest a cubic spline that would work within
> this context?

The fixed-knots ones (ns(), bs()) should work (and did so in at least
one case a couple of years ago...). These are linear, so lme() is used
rather than nlme(), unless of course you have other parts that need to
be modeled nonlinearly.

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907