R Sweave not interpreted correctly due to writing style?

Question

This case won't get correctly interpreted. I doubt it must be to do something with my writing style but I cannot notice what it is. You can run it with R CMD Sweave Test.Rnw; pdflatex Test.tex produces the Test.tex file and then compiling to Test.pdf.

What is causing the running to result into gibberish? Why are the Sweave blocks not properly interpreted?

Test.Rnw causing gibberish

\documentclass{article}
\usepackage{subcaption}
\usepackage{graphicx}
\usepackage{pdfpages}

\usepackage{Sweave}
\begin{document}
.
\begin{figure}
\centering
   \begin{subfigure}[b]{0.55\textwidth}
   <<>>=
    all.lab <- c("Arr./A", "Sin", "Digox")
    all.dat <- c(2274959, 1531001, 2406739)
    barplot(all.dat,
            names.arg=all.lab,
            col="darkblue",
            ylab="Average byte size",
            xlab="Groups")
    grid()
   @
   %\includepdf{fig1.pdf}
   \caption{a}
   \label{fig:Ng1}
\end{subfigure}
\begin{subfigure}[b]{0.55\textwidth}
   <<>>=
    age.lab <- c("Arr. +65", "Arr. [45,65]", "Arr", "Sin", "Sin +33")
    age.dat <- c(2274959, 1481397, 773624, 874208, 1087411)
    barplot(age.dat,
            names.arg=age.lab,
            col="darkblue",
            ylab="Average byte size",
            xlab="Groups")
    grid()
   @
   %\includepdf{fig2.pdf}
   \caption{b}
   \label{fig:Ng2}
\end{subfigure}
\begin{subfigure}[b]{0.55\textwidth}
   %\includepdf{fig3.pdf}
   <<>>=
    gender.lab <- c("Arr. f", "A m", "Sin f", "Sinu", "Digox f", "Digon m")
    gender.dat <- c(1416043, 2448017, 1421385, 537783, 1256545, 1181350)
    barplot(gender.dat,
            names.arg=gender.lab,
            col="darkblue",
            ylab="Average byte size",
            xlab="Groups")
    grid()
   @
   \caption{c}
   \label{fig:Ng2}
\end{subfigure}
\caption{MAIN CAPTION}
\end{figure}
\end{document}

Answer 1

The error is that you have indented the Sweave block, you must unindent the block. Sweave seems to be sensitive to indentation.

\documentclass{article}
\usepackage{subcaption}
\usepackage{graphicx}
\usepackage{pdfpages}

\usepackage{Sweave}
\begin{document}
\begin{figure}
\centering
\begin{subfigure}[b]{0.55\textwidth}
<<fig=TRUE>>=
all.lab <- c("Arr./A", "Sin", "Digox")
all.dat <- c(2274959, 1531001, 2406739)
barplot(all.dat,
    names.arg=all.lab,
    col="darkblue",
    ylab="Average byte size",
    xlab="Groups")
grid()
@
\caption{a}
\label{fig:Ng1}
\end{subfigure}
\begin{subfigure}[b]{0.55\textwidth}
<<fig=TRUE>>=
age.lab <- c("Arr. +65", "Arr. [45,65]", "Arr", "Sin", "Sin +33")
age.dat <- c(2274959, 1481397, 773624, 874208, 1087411)
barplot(age.dat,
    names.arg=age.lab,
    col="darkblue",
    ylab="Average byte size",
    xlab="Groups")
grid()
@
\caption{b}
\label{fig:Ng2}
\end{subfigure}
\begin{subfigure}[b]{0.55\textwidth}
<<fig=TRUE>>=
gender.lab <- c("Arr. f", "A m", "Sin f", "Sinu", "Digox f", "Digon m")
gender.dat <- c(1416043, 2448017, 1421385, 537783, 1256545, 1181350)
barplot(gender.dat,
    names.arg=gender.lab,
    col="darkblue",
    ylab="Average byte size",
    xlab="Groups")
grid()
@
\caption{c}
\label{fig:Ng2}
\end{subfigure}
\caption{MAIN CAPTION}
\end{figure}
\end{document}

Minimal Working Examples

The way I solved the issue was to encapsulate the issue, I noticed that there seems to be some conflicting character, block or space in the code. I provide below small working examples that made me realise that the error must be in the indentation.

Example 1. OP's subcase working with Sweave

\documentclass{article}
\usepackage{graphicx}
\usepackage{Sweave}

\begin{document}
\begin{figure}
<<fig=TRUE>>=
library(ggplot2)
all.lab <- c('Arr./A', 'Sin', 'Digox')
all.dat <- c(2274959, 1531001, 2406739)
ggplot() + geom_bar(aes(x=all.lab, y=all.dat), stat='identity')
@
\end{figure}
\end{document}

Example 2. Ggplot working with Sweave

\documentclass{article}
\usepackage{graphicx}
\usepackage{Sweave}
\begin{document}
\SweaveOpts{concordance=TRUE}

\begin{figure}
\centering
<<fig=TRUE>>=
library(ggplot2)
p <- ggplot(diamonds, aes(x=carat))
p <- p + geom_histogram()
print(p)
@
\end{figure}
\end{document}

Example 3. Base commands working with Sweave

\documentclass{article}
\usepackage{graphicx}         
\usepackage{Sweave}
\begin{document}
\begin{figure}
\centering
<<fig=TRUE>>=
plot(1:10)
@
\end{figure}
\end{document}

Run

1. R CMD Sweave test.Rnw
2. pdflatex test.txt
3. open test.pdf

Answer 2

With Rstudio using knitr, your code as is can be compiled showing the plots, but you have to add echo=FALSE to hide the R code and fit the three images in a page. Of course, the \usepackage{Sweave} is not necessary (nor using Sweave with Rstudio). You can omit also any text outside the float.

If you need to stay with Sweave then you should care of the chunk indentation, fig=TRUE option, and of course, at least for a float with three subfigures, omit the echo.

\documentclass{article}
\usepackage{subcaption}
\usepackage{graphicx}
\usepackage{pdfpages}
\begin{document}
\begin{figure}
\centering
   \begin{subfigure}[b]{0.55\textwidth}
   <<echo=F>>=
    all.lab <- c("Arr./A", "Sin", "Digox")
    all.dat <- c(2274959, 1531001, 2406739)
    barplot(all.dat,
            names.arg=all.lab,
            col="darkblue",
            ylab="Average byte size",
            xlab="Groups")
    grid()
   @
   %\includepdf{fig1.pdf}
   \caption{a}
   \label{fig:Ng1}
\end{subfigure}
\begin{subfigure}[b]{0.55\textwidth}
   <<echo=F>>=
    age.lab <- c("Arr. +65", "Arr. [45,65]", "Arr", "Sin", "Sin +33")
    age.dat <- c(2274959, 1481397, 773624, 874208, 1087411)
    barplot(age.dat,
            names.arg=age.lab,
            col="darkblue",
            ylab="Average byte size",
            xlab="Groups")
    grid()
   @
   %\includepdf{fig2.pdf}
   \caption{b}
   \label{fig:Ng2}
\end{subfigure}
\begin{subfigure}[b]{0.55\textwidth}
   %\includepdf{fig3.pdf}
   <<echo=F>>=
    gender.lab <- c("Arr. f", "A m", "Sin f", "Sinu", "Digox f", "Digon m")
    gender.dat <- c(1416043, 2448017, 1421385, 537783, 1256545, 1181350)
    barplot(gender.dat,
            names.arg=gender.lab,
            col="darkblue",
            ylab="Average byte size",
            xlab="Groups")
    grid()
   @
   \caption{c}
   \label{fig:Ng2}
\end{subfigure}
\caption{MAIN CAPTION}
\end{figure}
\end{document}

Moreover, knitr is more powerful: Just consider you "example 2" with knitr:

\documentclass{article}
\begin{document}
\begin{figure}
    <<>>=
        library(ggplot2)
        p <- ggplot(diamonds, aes(x=carat))
        p <- p + geom_histogram(bins=30)
        print(p)
    @
\end{figure}
\end{document}