Manipulate xtable() output

3
votes

I am trying to print out this table with xtable(), which is reproduced below:

                   X            B.1            B.2            B.3
1        (Intercept) -1.669 (0.093) -1.701 (0.094) -1.774 (0.121)
2          SEXFemale  -0.46 (0.023)  -0.386 (0.04)  -0.274 (0.17)
3    SEXFemale:BLACK                                0.132 (0.163)
4    SEXFemale:ASIAN                               -0.063 (0.089)
5 SEXFemale:HISPANIC                               -0.128 (0.074)

But I would like to override the way the first column is printed. Is it possible to include latex commands in xtable() or some other function to do this? Specifically, I would like the first column to be :

    X
Male: Female
Female X Black
       X Asian
       X Hispanic

i.e. I would like to write a \phantom{} switch into the column so that the spacing comes out right.

2
rlatex
You should ask this on tex.stackexchange.com instead – Giel
This is possible with print.xtable(..., print.results = FALSE) and then using regular expressions to find the second column and inserting your \phantom{} call. I gotta run now, but I may post a solution later. – Roman Luštrik
Would appreciate that, Roman. Thanks in advance. – user702432
Perhaps you could split the X column into two columns and use ftable for producing a table and memisc::toLatex to create Latex output for it? – ROLO

2 Answers

2
votes

You can use a regular expression to insert the \phantom{} into your table before calling xtable():

Recreate your data:

x <- structure(list(X = structure(c(1L, 2L, 4L, 3L, 5L), .Label = c("(Intercept)", 
                                                                    "SEXFemale", "SEXFemale:ASIAN", "SEXFemale:BLACK", "SEXFemale:HISPANIC"
), class = "factor"), B.1 = structure(c(3L, 2L, 1L, 1L, 1L), .Label = c("", 
                                                                        "-0.46 (0.023)", "-1.669 (0.093)"), class = "factor"), B.2 = structure(c(3L, 
                                                                                                                                                 2L, 1L, 1L, 1L), .Label = c("", "-0.386 (0.04)", "-1.701 (0.094)"
                                                                                                                                                 ), class = "factor"), B.3 = structure(c(4L, 3L, 5L, 1L, 2L), .Label = c("-0.063 (0.089)", 
                                                                                                                                                                                                                         "-0.128 (0.074)", "-0.274 (0.17)", "-1.774 (0.121)", "0.132 (0.163)"
                                                                                                                                                 ), class = "factor")), .Names = c("X", "B.1", "B.2", "B.3"), class = "data.frame", row.names = c(NA, 
                                                                                                                                                                                                                                                  -5L))

The regex:

x$X <- gsub("SEXFemale:", "\\\\phantom{Female} X ", x$X)
x$X <- gsub("SEXFemale", "Female", x$X)


library(xtable)
xx <- print(xtable(x), print.results = FALSE, include.rownames = FALSE,
        sanitize.text.function=function(x)x)
cat(xx)

The resulting text:

% latex table generated in R 2.15.0 by xtable 1.7-0 package
% Thu Jul 26 17:10:05 2012
\begin{table}[ht]
\begin{center}
\begin{tabular}{llll}
  \hline
X & B.1 & B.2 & B.3 \\ 
  \hline
(Intercept) & -1.669 (0.093) & -1.701 (0.094) & -1.774 (0.121) \\ 
  Female & -0.46 (0.023) & -0.386 (0.04) & -0.274 (0.17) \\ 
  \phantom{Female} X BLACK &  &  & 0.132 (0.163) \\ 
  \phantom{Female} X ASIAN &  &  & -0.063 (0.089) \\ 
  \phantom{Female} X HISPANIC &  &  & -0.128 (0.074) \\ 
   \hline
\end{tabular}
\end{center}
\end{table}

And the final output:

enter image description here

0
votes

Ok here's one way of doing this.

# dummy example to get ANOVA table
ctl <- c(4.17,5.58,5.18,6.11,4.50,4.61,5.17,4.53,5.33,5.14)
trt <- c(4.81,4.17,4.41,3.59,5.87,3.83,6.03,4.89,4.32,4.69)
group <- gl(2,10,20, labels=c("Ctl","Trt"))
weight <- c(ctl, trt)
lm.D9 <- lm(weight ~ group)

# print xtable into a character
x <- print(xtable(lm.D9), print.results = FALSE)

What I'm doing here is finding specific words and wrapping them into \emph{}. What you could do is have it replaced with a specific text or wrap in your custom switch.

x <- sub("(\\(Intercept\\))", "\\\\emph{\\1}", x)
x <- sub("(groupTrt)", "\\\\emph{\\1}", x)
cat(x)

% latex table generated in R 2.15.1 by xtable 1.7-0 package
% Thu Jul 26 17:41:16 2012
\begin{table}[ht]
\begin{center}
\begin{tabular}{rrrrr}
  \hline
 & Estimate & Std. Error & t value & Pr($>$$|$t$|$) \\ 
  \hline
\emph{(Intercept)} & 5.0320 & 0.2202 & 22.85 & 0.0000 \\ 
  \emph{groupTrt} & -0.3710 & 0.3114 & -1.19 & 0.2490 \\ 
   \hline
\end{tabular}
\end{center}
\end{table}

enter image description here