[R-sig-QA] [Fwd: create groups based upon different levels of 4 factors] update 1

[Douglas Bates]

On Fri, Nov 28, 2008 at 12:06 PM, Mary A. Marion <mmstat at comcast.net> wrote:
> Hello to the R gurus,
>
> My data is bound in a dataframe.
> data2<-cbind.data.frame(digrp,gender, agegrp, bmigrp,wt)
> data2[1:5,]
>
> I want to create groups based upon different levels of 4 factors
> ind<- as.logical ( (digrp=1) & (gender=1) & (agegrp="[20-35)") &
> (bmigrp="[0,18.5)" ) );

Use "==" not "=" for comparisons.

ind <- (digrp == 1) & (gender == 1) & ...

There are many other comments that could be made regarding your code
such as converting gender to a factor with meaningful labels and not
calling a S3 method like cbind.data.frame directly but that should get
you started.

> Error in (digrp = 1) & (gender = 1) & (agegrp = "[20-35)") :
> operations are possible only for numeric or logical types
>>
> I keep getting an error message.  How do I see that the above statement is
> in correct type format?
> #operations are possible only for numeric or logical types.   why does it
> not include bmigrp?
>
> Data was created as follows:
>
> #BMI CLASSIFICATION
> breaks<-c(0,18.5,25,30,200)
> bmigrp<-cut(bmi,breaks=breaks,right=FALSE)
>
> #AGE GROUPS
> breaks<-c(1,35,60,90)
> agegrp<-cut(age,breaks=breaks,right=FALSE)
>
> #DI GROUPS
> #1=yes has diabetes, 2=no does not have
> digrp<-as.factor(join5[,"DIQ010"])
>
> #GENDER GROUPS
> #1=male, 2=female
> gender<-as.factor(join5[,"RIAGENDR"])
>
> #FORM DATA FRAME TO ANALYZE
> data2<-cbind.data.frame(digrp,gender, agegrp, bmigrp,wt)
> data2[1:5,]
>
> Thanks.
>
> Sincerely,
> Mary A. Marion
>
> _______________________________________________
> R-sig-QA at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-qa
>