R help - Oct 2011 - Proc Mixed to R

Hi All,

I'm working with some SAS code to analyze an experiment set up as follows:
66 subjects (colonies) treated with a random treatment (1-8) and measured at
three time points.
The data structure looks like:

            input colony tmt y1 y2 y3;

            y=y1; date=*1*; output;

            y=y2; date=*2*; output;

            y=y3; date=*3*; output;

            datalines;

1          3          6725   6750   925

2          8          6950   5800   11275

3          4          4200   6100   6475


Procedure:

*proc* *mixed* data=Nosema method=ml covtest;

            class colony tmt;

            model y=tmt date tmt*date / s;

            repeated / type=un subject=colony;

            random colony;

            lsmeans tmt/cl adjust=tukey;


I am able to get something close by running aov on it, even closer by using
Anova{car} and calling type=3.  The problem I'm having is running the tukey
and/or getting something similar to SAS's "Solution for Fixed
Effects"
table.  Any idea what to do?


Thanks,


Molly

	[[alternative HTML version deleted]]

You need to use either the lme4 or nlme packages for mixed models. 
(There are some other possibilities as well). See
http://glmm.wikidot.com/faq for MUCH more detail


On 10/27/2011 7:19 PM, Molly Hanlon wrote:
> Hi All,
>
> I'm working with some SAS code to analyze an experiment set up as
follows:
> 66 subjects (colonies) treated with a random treatment (1-8) and measured
at
> three time points.
> The data structure looks like:
>
>              input colony tmt y1 y2 y3;
>
>              y=y1; date=*1*; output;
>
>              y=y2; date=*2*; output;
>
>              y=y3; date=*3*; output;
>
>              datalines;
>
> 1          3          6725   6750   925
>
> 2          8          6950   5800   11275
>
> 3          4          4200   6100   6475
>
>
> Procedure:
>
> *proc* *mixed* data=Nosema method=ml covtest;
>
>              class colony tmt;
>
>              model y=tmt date tmt*date / s;
>
>              repeated / type=un subject=colony;
>
>              random colony;
>
>              lsmeans tmt/cl adjust=tukey;
>
>
> I am able to get something close by running aov on it, even closer by using
> Anova{car} and calling type=3.  The problem I'm having is running the
tukey
> and/or getting something similar to SAS's "Solution for Fixed
Effects"
> table.  Any idea what to do?
>
>
> Thanks,
>
>
> Molly
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@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.
	[[alternative HTML version deleted]]