Re: [R] declaring constants in an Sweave / LaTeX document

From: Duncan Murdoch <murdoch_at_stats.uwo.ca>
Date: Sat, 22 Nov 2008 10:45:13 -0500

On 22/11/2008 2:01 AM, Kyle Matoba wrote:
> List,
>
> I would like to set a variable to hold, say, the size of my plots in a
> Sweave document. i.e. something like the following in my '.Rnw' file:
>
> ==============================================================================
> smallPlotSize = 4
>
> <<fig1, echo=false, results=hide, height=smallPlotSize, width=smallPlotSize,
> fig=true>>=
> http://www.stanford.edu/~xing/statfinbook/_BookData/Chap05/q_us_gdp.txt",
>
> GDP <- ts(data=dat\$VALUE, start=c(1947,1), frequency=4)
>
> acf(GDP, type="correlation", main= "ACF of US GDP")
> @
> ==============================================================================

I don't think that is possible, but you can do things like

\SweaveOpts{height=3.5, width=7}

anywhere in your document to set the default size for subsequent plots.

You can also set LaTeX macros to change the displayed size, e.g.


Here you're talking to LaTeX, not to R, so those are setting options to the \includegraphics{} calls. Unfortunately, you can't put \SweaveOpts in the macro to also set the R width and height, because Sweave handles all the \SweaveOpts calls before calling LaTeX.

Duncan Murdoch

> This way I can tweak things if I have to include a bunch of graphics of the
> same size in a file. I am not great with LaTeX and still very new to using
> Sweave, so I was wondering what the preferred method of doing something like
> this is between LaTeX and Sweave. I have seen some people doing some tricky
> stuff from within R such that when their code is preprocessed it is compiled
> as LaTeX source, but this seems like an ugly hack for something so simple.
>
> Please forgive me if this has been addressed elsewhere; I have found
> documentation on Sweave to be rather sparse.
>
> Best,
>
> Kyle
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help_at_r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help