[R] NA's in multiple choice summary table in Hmisc

[Erich Studerus]

Thank you so much. I have yet another problem that I could not resolve from
the documentation. I want to get line breaks for long variable names.

Here's an example:
Gender<-sample(c("m","f"),20,replace=TRUE)
Education<-rnorm(20,13)
label(Education)<-"Years of\nEducation"
summary(Gender~Education,method="reverse")
Descriptive Statistics by Gender

Descriptive Statistics by Gender

+---------+--------------+--------------+
|         |f             |m             |
|         |(N=12)        |(N=8)         |
+---------+--------------+--------------+
|Years of |12.5/12.9/13.9|12.5/12.9/13.5|
|Education|              |              |
+---------+--------------+--------------+

As you can see, it works fine for the printing in R, but when I submit this
table to the latex-function the \n gets removed. How can I prevent this?

Regards
Erich



-----Ursprüngliche Nachricht-----
Von: Frank E Harrell Jr [mailto:f.harrell at vanderbilt.edu] 
Gesendet: Donnerstag, 9. Oktober 2008 04:17
An: Erich Studerus, Psychiatrische Uni-Klinik
Cc: r-help at r-project.org
Betreff: Re: [R] NA's in multiple choice summary table in Hmisc

Erich Studerus, Psychiatrische Uni-Klinik wrote:
> Hi,
> 
> I have a set of 30 binary variables measuring side effects after drug 
> treatment. Since each subject can have multiple side effects, I want to 
> display these side effects in a multiple choice table. I'm using the 
> summary and mChoice functions of the Hmisc package, because it produces 
> nicely formatted latex tables. My problem is, that table includes a 
> category for people who have at least one absent symptom and I don't 
> know how to exclude it.
> 
> Here's a very easy reproducible example:
> 
> library(Hmisc)
> Symptom1<-c("Headache","Headache",NA)
> Symptom2<-c(NA,"Anxiety",NA)
> Symptoms<-mChoice(Symptom1,Symptom2)
> summary(~Symptoms,method="reverse")
> Descriptive Statistics  (N=3)
> 
> +-------------------+-------+
> |                   |       |
> +-------------------+-------+
> |Symptom1 : Headache|67% (2)|
> +-------------------+-------+
> |    NA             |67% (2)|
> +-------------------+-------+
> |    Anxiety        |33% (1)|
> +-------------------+-------+
> 
> Now, I want to either completely exclude the NA category or - if 
> included - it should count the number of people who have NA's in all 
> symptoms rather than in a single symptom. In the example above, the NA's 
> should be 1 instead of 2 cases.
> I tried to specify na.include=FALSE but it does not work.
> 
> Any help is highly appreciated.
> 
> Erich

Instead of NA put '' or "" and it should work.
Frank


-- 
Frank E Harrell Jr   Professor and Chair           School of Medicine
                      Department of Biostatistics   Vanderbilt University