R help - Jul 2011 - Creating composite factor and changing format from character to factor

Dear Help-list, I have a dataframe containing 6 variables, 4 of which are
factors, 2 numeric.  I want to create another factor variable (SitePos) by
combining 2 existing factors (Site and Position).  I have tried a number of
approaches based on trolling the R FAQs, various R webpages, etc., none of which
work.  One approach e.g. Data1$SitePos <- paste(Data1$Site, Data1$Position)
creates the appropriate "SitePos" values e.g. "CR core" but
as character values not as a factor.  A linear model run on the updated
dataframe works but notes that it coerces "SitePos" from character to
factor e.g. Model.G = lmer(log10(SrCa) ~ SitePos + (1 | Eel), data = Data1) . 
The next step of a multiple comparison test on the output of the linear model:
Model.G.mct = glht(Model.G, linfct = mcp(SitePos = "Tukey")) fails
because the mct does not recognize "SitePos" as a factor and gives
error message: "Error in mcp2matrix (model, linfct = linfct): Variable(s)
'SitePos' of class 'character' is/are not contained as a factor
in 'model'.  My final step is outputting the mct results:  summary
(Model.G.mct, test = adjusted(type = "single-step")).  Any suggestions
as to how to create a composite factor variable in the dataframe that would
permit execution of the analysis code would be much appreciated.  I am using R
version 2.13.0 with packages lme4 and multcomp loaded.  Regards,B Jessop
 		 	   		  
	[[alternative HTML version deleted]]

On Jul 16, 2011, at 10:21 AM, B Jessop wrote:
> Dear Help-list, I have a dataframe containing 6 variables, 4 of  
> which are factors, 2 numeric.  I want to create another factor  
> variable (SitePos) by combining 2 existing factors (Site and  
> Position).  I have tried a number of approaches based on trolling  
> the R FAQs, various R webpages, etc., none of which work.  One  
> approach e.g. Data1$SitePos <- paste(Data1$Site, Data1$Position)  
> creates the appropriate "SitePos" values e.g. "CR core"
but as
> character values not as a factor.
In one line you could have written:

  Data1$SitePos <- factor( paste(Data1$Site, Data1$Position) )

Or you could have created an interaction:

Data1$SitePos <- interaction(Data1$Site, Data1$Position)

>  A linear model run on the updated dataframe works but notes that it  
> coerces "SitePos" from character to factor e.g. Model.G =  
> lmer(log10(SrCa) ~ SitePos + (1 | Eel), data = Data1) .  The next  
> step of a multiple comparison test on the output of the linear  
> model: Model.G.mct = glht(Model.G, linfct = mcp(SitePos =
"Tukey"))
> fails because the mct does not recognize "SitePos" as a factor
and
> gives error message: "Error in mcp2matrix (model, linfct = linfct):  
> Variable(s) 'SitePos' of class 'character' is/are n!
> ot contained as a factor in 'model'.  My final step is outputting  
> the mct results:  summary (Model.G.mct, test = adjusted(type =  
> "single-step")).  Any suggestions as to how to create a composite
> factor variable in the dataframe that would permit execution of the  
> analysis code would be much appreciated.  I am using R version  
> 2.13.0 with packages lme4 and multcomp loaded.  Regards,B Jessop
> 		 	   		
> 	[[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.
David Winsemius, MD
West Hartford, CT