R help - Jul 2013 - Does a general latex table-making function exist?

Our Hmisc package summary.formula function and its latex methods can 
make some fairly advanced tables.  But the tables have to be regular. 
For example, all rows of the tables are based on the same data frame. 
I'm thinking that what is needed is a ggplot2-like set of functions for 
building a table row-by-row or row-by-block of rows.  Different row 
blocks could have different denominators, e.g., the first part of the 
table might be on everyone and a latter block of rows be for females, 
with different summary statistics computed.

Has anyone already written functions creating LaTeX markup with such 
functionality?

Thanks
Frank

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

On 13-07-31 4:03 PM, Frank Harrell wrote:
> Our Hmisc package summary.formula function and its latex methods can
> make some fairly advanced tables.  But the tables have to be regular.
> For example, all rows of the tables are based on the same data frame.
> I'm thinking that what is needed is a ggplot2-like set of functions for
> building a table row-by-row or row-by-block of rows.  Different row
> blocks could have different denominators, e.g., the first part of the
> table might be on everyone and a latter block of rows be for females,
> with different summary statistics computed.
>
> Has anyone already written functions creating LaTeX markup with such
> functionality?
My tables package does some of what you are asking for; I'm not sure if 
it does everything.

Duncan Murdoch

Duncan,

I had read your excellent tables package vignette at 
http://cran.r-project.org/web/packages/tables/vignettes/tables.pdf when 
it first came out.  It is extremely impressive.  I'm glad to be reminded 
to give it another look.

Is there a way to make the special symbols n and 1 refer to the number 
of non-missing observations rather than the length of a vector?

Do you feel like taking on this challenge?  An example of an irregular 
table I'm thinking of is the following

                Females              Males
              Q1 Med Q3   (n)       Q1 Med Q3   (n)
Age          25  49 63 (1016)     26  50 64  (1767)

Canadians
  Weight (kg) 57  63 74 ( 243)     67  73 90  ( 401)

Canadians could mean country=='Canada'.

Thanks!
Frank

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