LaTeX table through xtable

Question

I wonder how to get LATEX table using xtable function from the following R code.

Block <- gl(8, 4)
A <- factor(c(0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,
          0,1,0,1,0,1,0,1,0,1,0,1))
B <- factor(c(0,0,1,1,0,0,1,1,0,1,0,1,1,0,1,0,0,0,1,1,
          0,0,1,1,0,0,1,1,0,0,1,1))
C <- factor(c(0,1,1,0,1,0,0,1,0,0,1,1,0,0,1,1,0,1,0,1,
          1,0,1,0,0,0,1,1,1,1,0,0))
Yield <- c(101, 373, 398, 291, 312, 106, 265, 450, 106, 306, 324, 449,
       272, 89, 407, 338, 87, 324, 279, 471, 323, 128, 423, 334,
       131, 103, 445, 437, 324, 361, 302, 272)
aovdat <- data.frame(Block, A, B, C, Yield)

summary(aov(Yield~Block+A*B+Error(A*Block), data=aovdat))


Error: A
  Df Sum Sq Mean Sq
A  1 3465.3  3465.3

Error: Block
      Df Sum Sq Mean Sq
Block  7   4499  642.71

Error: A:Block
          Df Sum Sq Mean Sq F value Pr(>F)
B          1  41616   41616  3.5354 0.1091
Residuals  6  70628   11771               

Error: Within
          Df Sum Sq Mean Sq F value  Pr(>F)  
B          1 119568  119568  7.3751 0.01673 *
A:B        1     28      28  0.0017 0.96734  
Residuals 14 226975   16213                  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Using the following code

xtable(summary(aov(Yield~Block+A*B+Error(A*Block), data=aovdat)))

producing the following error

Error in rbind(deparse.level, ...) : 
  numbers of columns of arguments do not match

Is it possible to get a LaTeX table like this?

SOV       Df Sum Sq Mean Sq F value  Pr(>F)  
A          1 3465.3  3465.3
Block      7   4499  642.71
B          1  41616   41616  3.5354 0.1091
Residuals  6  70628   11771               
B          1 119568  119568  7.3751 0.01673 *
A:B        1     28      28  0.0017 0.96734  
Residuals 14 226975   16213                  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Answer 1

summary is producing four tables in a list, so using lapply like so:

mod <- aov(Yield ~ Block + A * B + Error(A * Block), data=aovdat)
lapply(summary(mod), xtable)

will produce a list of LaTeX tables.

A for loop with print will produce the LaTeX code only, with the table names in the LaTeX caption.

for (i in 1:4) print(xtable(summary(mod)[[i]], names(summary(mod)[i])))

If you want all the tables as one, you need to do a bit of work because rbind doesn't like tables with different numbers of columns. rbind.fill in the plyr package will do the job

library(plyr)

aovtab <- do.call(rbind.fill, lapply(1:length(summary(mod)), 
  function(x) unclass(summary(mod)[[x]])[[1]]))

but we have to manually add the rownames back as data because a) rbind.fill drops them, and b) they were duplicated anyway, which xtable wouldn't like.

SOV <- unlist(lapply(1:length(summary(mod)), 
  function(x) rownames(unclass(summary(mod)[[x]])[[1]]))

print(xtable(cbind(SOV, aovtab), include.rownames=F)

Answer 2

After playing with this for a little while, the following seems to produce results in LaTeX form.

xtable(as.data.frame(unlist(coef(anova.mod))))

This gives a LaTeX table of the blocks and their effects, as below.

\begin{table}[ht]
\begin{center}
\begin{tabular}{rr}
  \hline
 & unlist(coef(anova.mod)) \\ 
  \hline
(Intercept).(Intercept) & 291.59 \\ 
  A.A1 & 20.81 \\ 
  Block.Block2 & -7.50 \\ 
  Block.Block3 & 5.50 \\ 
  Block.Block4 & -14.25 \\ 
  Block.Block5 & -0.50 \\ 
  Block.Block6 & 11.25 \\ 
  Block.Block7 & -11.75 \\ 
  Block.Block8 & 24.00 \\ 
  A:Block.B1 & 144.25 \\ 
  Within.B1 & 139.00 \\ 
  Within.A1:B1 & 4.33 \\ 
   \hline
\end{tabular}
\end{center}
\end{table}

Is this sufficient for your purposes?

In general, xtable prefers matrix or dataframe input, and R often returns lists, so what I normally do is extract the parts I am interested in and then convert to a matrix or a dataframe.

Apologies for the ugly TeX, there's no support for it here.