[R-sig-ME] Problems with an anova mehtod to test for differences between groups in a lme ()

[Douglas Bates]

On 4/10/08, R.S. Cotter <cotter.rs at gmail.com> wrote:
> Dear all,
>
>  I have problems using anova mehtod to test for differences between groups in
>  a lme () model. As described at page 224 and 225 in Mixed-Effects Models in
>  S and S-PLUS, by Pinhiro and Bates. Could it that this script doesn't work
>  for R?, see below (in yellow).
>
>  Respons: Speed
>  Fixed effects: Fuel, CarMod (1,2&3)
>  Random effects: Place
>
>  > lme1 <- lme(Speed ~  Fuel*Car, random=~1|Place,data=test)
>  > summary (lme1)
>  Linear mixed-effects model fit by REML
>  Data: test
>       AIC      BIC    logLik
>  261.2013 275.6996 -121.6007
>
>  Random effects:
>  Formula: ~1 | Place
>            (Intercept)         Residual
>  StdDev: 0.0003238738   5.013858
>
>  Fixed effects: Speed ~ Fuel + Car + Driver + Fuel * Car
>                       Value        Std.Error     DF   t-value       p-value
>  (Intercept)        -29.33479  12.743084   30   -2.302017   0.0285
>  Fuel                10.04684    1.408789    30    7.131542   0.0000
>  CarMod2         46.55593    14.192029   7     3.280428   0.0135
>  CarMod3          1.65157     18.247158   7     0.090511   0.9304
>  Fuel:CarMod2  -5.53264     1.624159    30   -3.406464   0.0019
>  Fuel:CarMod3  -0.18452      2.010470   30   -0.091779   0.9275

>  Number of Observations: 44
>  Number of Groups: 10

>  > anova( lme1, L = c(Fuel:CarMod2 = 1, Fuel:CarMod3 = -1) )
>  Error: syntax error in "anova( lme1, L = c(Fuel:CarMod2="

The semicolon is not a valid character in an R name.  Try quoting the names.
>
>  Regards R.S
>
>         [[alternative HTML version deleted]]
>
>  _______________________________________________
>  R-sig-mixed-models at r-project.org mailing list
>  https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>