[R] contrast package with interactions in gls model

[Frank Harrell]

You did not follow the posting guide.  You did not specify which packages you
were using.  It appears that you are mixing the rms package with some other
functions such as gls.  If you want to use rms, use the Gls function instead
of gls, and type ?contrast.rms to see examples of the use of contrast().
Frank

Marylin Bejarano wrote:
> 
> Hi!
> 
> I try to explain the efffect of (1)  forest where i took samples's soils
> (*
> Lugar*: categorical variable with three levels), (2) nitrogen addition
> treatments (*Tra*: categorical variable with two levels) on total carbon
> concentration's soil samples (*C: *continue* *variable) during four months
> of sampling (*Time:* categorical and ordered variable with four levels).
> 
> I fitted the following final model with gls function:
> 
> var1<-varIdent(form=~ 1| Lugar* factor(Time))
> FINAL<-gls(C ~  Tra+ Lugar+ Time + Time*Tra + Tra*Lugar, data=datos,
> weights=var1, method="REML")
> 
> the summary function resulted in this first data's set (I omit
> correlation's
> matrix):
> 
> Generalized least squares fit by REML
>   Model: C ~ Tra + Lugar + Time + Time * Tra + Tra * Lugar
>   Data: datos
>        AIC      BIC    logLik
>   1129.458 1191.982 -540.7291
> 
> Variance function:
>  Structure: Different standard deviations per stratum
>  Formula: ~1 | Lugar * factor(Time)
>  Parameter estimates:
> Chixchulub*0   Xmatkuil*0    Hobonil*0 Chixchulub*2   Xmatkuil*2
>  Hobonil*2 Chixchulub*3
>    1.0000000    0.7759324    0.5300811    0.9640559    0.8200742
>  0.2966545    0.9553168
>   Xmatkuil*3    Hobonil*3 Chixchulub*4   Xmatkuil*4    Hobonil*4
>    1.7350290    0.3430286    0.6241658    0.9573922    0.4651515
> 
> Coefficients:
>                                      Value             Std.Error   
> t-value
>         p-value
> (Intercept)                 260.48540  16.48991  15.796653  0.0000
> Tra0                          -9.38703      23.74893  -0.395261  0.6935
> LugarChixchulub       -0.15377  19.60260  -0.007845  0.9938
> Lugar Hobonil        -173.21354  15.89736 -10.895741  0.0000
> Time2                    -14.74999  14.55909  -1.013112  0.3135
> Time3                     14.42177  15.64594   0.921758  0.3589
> Time4                     14.77803  16.72367   0.883659  0.3790
> Tra0:Time2             17.93859  20.78257   0.863156  0.3901
> Tra0:Time3            -48.77118  22.17628  -2.199250  0.0302
> Tra0:Time4            -52.63611  23.20192  -2.268610  0.0254
> Tra0:LugarChixchulub   74.43956  28.11275   2.647893  0.0094
> Tra0:Lugar Hobonil     43.03416  23.32391   1.845066  0.0680
> 
> anova function generated this table:
> 
>  enom. DF: 100
>             numDF   F-value p-value
> (Intercept)     1 1693.1234  <.0001
> Tra             1    5.3225  0.0231
> Lugar           2  247.7047  <.0001
>  Time            3    0.4767  0.6992
> Tra:Time        3    6.0531  0.0008
> Tra:Lugar       2    3.5061  0.0338
> 
> I want to detetect differences between levels of Tra:Lugar interaction.
> For
> example:
> 
> 1. Tra0:LugarChixchulub vs Tra1:LugarChixchulub  (between treatment levels
> for same forest) or,
> 2. Tra0:LugarChixchulub vs Tra0:LugarHobonil        (for same treatment
> among forests levels)
> 
> I used function contrast (package contrast) whit following script to probe
> the hypotesis 1.:
> 
> con<-contrast(FINAL, list(Lugar= 'Xmatkuil', Tra=1),
> list(Lugar='Xmatkuil',
> Tra = 0))
> 
> but i found this error message:
> 
> Error en gendata.default(fit = list(modelStruct = list(varStruct =
> c(-0.253689933940456,  :
>   not enough factors
> 
> I would be grateful if somebody tell me I'm doing wrong with my contrast
> function script.
> 
> Thanks in advance,
> 
> 
> Marylin Bejarano
> PHd candidate in Ecology Institute of Mexico's National Autonomous
> University
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 


-----
Frank Harrell
Department of Biostatistics, Vanderbilt University
--
View this message in context: http://r.789695.n4.nabble.com/contrast-package-with-interactions-in-gls-model-tp3751255p3751269.html
Sent from the R help mailing list archive at Nabble.com.