I have produced a dynamic document using knitr. The document makes
extensive use of the package's knit_expand() function for
templates. This is illustrated by the MWE (based on Yihui Xie's own
example for the function).
Main document knit-expand-MWE.Rnw
\documentclass{article}
\title{How to extract code when using\\
knit\_expand() for templates?}%
\author{Knitr User}
\begin{document}
\maketitle
\tableofcontents
\newpage
\section{Write one row of data}
Only the first two sections are evaluated.
<<run-all, include=FALSE>>=
library(knitr)
src = NULL
for (i in 1:3) src = c(src, knit_expand('template.Rnw'))
#
\Sexpr{paste(knit(text = src), collapse = '\n')}
\end{document}
Template template.Rnw called by main document
\subsection{Now i is {{i}}}
This chunk is {{if (i > 2) 'not '}}evaluated.
<<row-{{i}}, eval={{i <= 2}}>>=
# row number {{i}}
iris[{{i}}, ]
#
I now need to extract the corresponding R code. Running purl("knit-expand-MWE.Rnw") outputs knit-expand-MWE.R, which includes the code in the chunk with a reference to the template:
## ----run-all, include=FALSE----------------------------------------------
library(knitr)
src = NULL
for (i in 1:3) src = c(src, knit_expand('template.Rnw'))
What I would like instead is the corresponding "expanded" code (for the benefit of colleagues who do not use knitr), for example:
## ----row-1, eval=TRUE----------------------------------------------
## row number 1
iris[1, ]
## ----row-2, eval=TRUE----------------------------------------------
## row number 2
iris[2, ]
## ----row-3, eval=FALSE----------------------------------------------
## row number 3
iris[3, ]
How can I achieve this?
You can run purl() on src, e. g.
purl(text = src, output = 'knit-expand.R')
Related
I am writing an R Markdown document using the Python engine of {reticulate}. I am quite happy with how it works.
The only thing is, I cannot use r as a Python object name that I'm going to use in multiple chunks.
---
title: "Untitled"
output: html_document
---
## Object name `r`
```{python}
r = 10
print(r) ##> 10
```
```{python}
print(r) ##> <__main__.R object at 0x119ad37d0>
```
I understand r is a good name when we use R objects from within a Python chunk. Since I know I am not going to do that in my project, I would like to use r as a name for my Python object.
Is there any way to change the name, r, for the R object created by reticulate? Or, to tell reticulate not to create r object?
I am aware of the two straightforward workarounds
Don't use r for Python objects.
Write everything in one big chunk and not share r between Python chunks.
but I'd like to have more freedom.
The object name r is special since it is used to communicate between R and python. The same holds for py in R:
---
title: "Untitled"
output: html_document
---
## Object name `r`
```{python}
foo = 10
print(foo) ##> 10
```
```{r}
library(reticulate)
py$foo ##> [1] 10
```
```{r}
foo <- 10
```
```{python}
print(foo) ##> 10
print(r.foo) ##> 10.0
```
Avoiding the use of r as an opject name is therefore the only good possibility I can see.
Still working on the details but I think setting source and output knit hooks may help. My dirty first try looks like this:
---
title: "Untitled"
output: html_document
---
```{r setup, include=FALSE}
library(knitr)
knitr::knit_hooks$set(
source = function(x, options) {
if (!is.null(options$rsub)) x = gsub(options$rsub, 'r', x, fixed = TRUE)
sprintf('<div class="%s"><pre class="knitr %s">%s</pre></div>\n', "source", tolower(options$engine), x)
},
output = function(x, options) {
if (!is.null(options$rsub)) x = gsub(options$rsub, 'r', x, fixed = TRUE)
paste0('<pre><code>', x, '</code></pre>\n')
}
)
```
```{python, rsub='rrr'}
rrr = 10
print(rrr)
```
```{python, rsub='rrr'}
print(rrr)
```
I'm creating a R Markdown document using knitr and am running into trouble using xtable to create a table. My table is very large and I'm trying to reduce the size using the size command in the print statement. The issue I'm running into is that the command seems to add two extra curly braces which show up in the PDF, one before the table and one after.
Does anyone know a way to fix this?
MWE:
---
output:
pdf_document:
keep_tex: yes
tables: true
---
```{r, results='asis', echo=FALSE}
library(xtable)
my.df <- data.frame(matrix(c(1:18),nrow=2))
glossaryprint <- xtable(my.df, caption="Summary of Terms")
print(glossaryprint,
comment=FALSE,
floating=FALSE,
size="footnotesize"
)
```
Note: The issue that is subject of the question and this answer has been resolved in xtable 1.8-2.
Although the question has been self-answered with a workaround, I think for other users some more details might be helpful.
What happens?
To understand what is happening here, we need to take a close look at the conversion steps the document undergoes on its way from RMD to PDF. The steps are:
RMD --> MD --> TEX --> PDF
Let's look at the files in reversed order:
PDF: (generated from TEX by pdflatex)
TEX: (generated from MD by pandoc)
% …
\{\footnotesize
\begin{tabular}{rrrr}
\hline
& X1 & X2 & X3 \\
\hline
1 & 1 & 3 & 5 \\
2 & 2 & 4 & 6 \\
\hline
\end{tabular}
\}
% …
MD (generated from RMD by knitr)
---
output:
pdf_document:
keep_tex: yes
---
{\footnotesize
\begin{tabular}{rrrr}
\hline
& X1 & X2 & X3 \\
\hline
1 & 1 & 3 & 5 \\
2 & 2 & 4 & 6 \\
\hline
\end{tabular}
}
RMD: (source file)
---
output:
pdf_document:
keep_tex: yes
---
```{r, results='asis', echo=FALSE}
library(xtable)
mytable <- xtable(data.frame(matrix(c(1:6), nrow = 2)))
print(mytable,
comment = FALSE,
floating = FALSE,
size = "footnotesize"
)
```
Recall: The problem is, that there are visible curly braces in the PDF. Where do they come from?
They are the result of the escaped curly braces in the TEX file (\{ and \}).
These curly braces also exist in the MD file, but there they are not escaped.
So we know two things by now: We see the curly braces because they are escaped and they are escaped by pandoc.
But why do these curly braces exist at all? print.xtable outputs them when a size is specified. The goal is to create a group and to apply size only within that group. (With floating = TRUE, no grouping by curly braces is required because there is a figure environment whithin which the size is set. The curly braces are printed anyways.)
And why does pandoc escape that pair of curly braces but leaves all the other curly braces (e.g. in \begin{tabular}) unescaped? This is because pandoc is supposed to escape special characters that are meant literally but leave raw LaTeX unescaped. However, pandoc does not know that the outer curly braces are LaTeX commands and not text.
(With floating = TRUE the problem does not occur because the curly braces are within a figure environment which is recognized as LaTeX.)
Solutions
After having understood the problem, what can we do about it? One solution has already been posted by the OP: Abstain from spefifying size in print.xtable and insert the footnotesize command manually:
---
output:
pdf_document:
keep_tex: yes
---
```{r, results='asis', echo=FALSE}
library(xtable)
mytable <- xtable(data.frame(matrix(c(1:6), nrow = 2)))
cat("\\begin{footnotesize}")
print(mytable,
comment = FALSE,
floating = FALSE
)
cat("\\end{footnotesize}")
```
However, on the long run it would be nice if xtable (current version: 1.8-0) generated LaTeX code that survives the pandoc conversion. This is quite simple: print.xtable checks if size is set and if so, inserts { before the size specification and } at the end of the table:
if (is.null(size) || !is.character(size)) {
BSIZE <- ""
ESIZE <- ""
}
else {
if (length(grep("^\\\\", size)) == 0) {
size <- paste("\\", size, sep = "")
}
BSIZE <- paste("{", size, "\n", sep = "")
ESIZE <- "}\n"
}
This small modification replaces { with \begingroup and } with \endgroup:
if (is.null(size) || !is.character(size)) {
BSIZE <- ""
ESIZE <- ""
}
else {
if (length(grep("^\\\\", size)) == 0) {
size <- paste("\\", size, sep = "")
}
BSIZE <- paste("\\begingroup", size, "\n", sep = "")
ESIZE <- "\\endgroup\n"
}
For LaTeX, this makes no difference, but as pandoc recognizes \begingroup (as oppsed to {) it should solve the problem. I reported this as a bug in xtable and hopefully the issue will be fixed in future versions.
I was able to fix this by not including the size parameter in the print statement but rather directly before and after the chunk.
\begin{footnotesize}
#chunk
\end{footnotesize}
Below is an rmarkdown document that can be pasted into rstudio.
My problem is that output from htmlTable is prepended/appended with cruft from the htmlTable attributes.
---
title: "SO_question"
author: "AC"
date: "Wednesday, May 28, 2014"
output:
html_document:
theme: readable
---
My heading
============
This is a few tables. Notice that `htmlTable` prints `[1]"` before each table and `" attr(,“class”) [1] “htmlTable” “character” [1] “` after each table. How can I avoid this?
``` {r html_table, results='asis', echo=FALSE, message=FALSE}
library("htmlTable")
library("reshape2")
#Chick weight example
names(ChickWeight) <- tolower(names(ChickWeight))
chick_m <- melt(ChickWeight, id=2:4, na.rm=TRUE)
for (i in unique(chick_m$diet)) {
diet <- subset(chick_m, diet==i)
table_to_print <- dcast(chick_m, time ~ variable, mean)
print(htmlTable(table_to_print, rgroup=c(""), n.rgroup=nrow(table_to_print)))
}
```
Bonus question: How to format the last row in each table as bold text (suited for a 'total' row)?
Rather than using print on your htmTable, use cat to properly render it.
for (i in unique(chick_m$diet)) {
diet <- subset(chick_m, diet==i)
table_to_print <- dcast(chick_m, time ~ variable, mean)
cat(htmlTable(table_to_print, rgroup=c(""), n.rgroup=nrow(table_to_print)))
}
There is a print.htmlTable function that is called when a print is performed on an object from the htmlTable function. It should automatically call the cat, not sure if this was true May '14 but it works today.
In the 1.1 version of the htmlTable-package (the function was separated from the Gmisc-package) there is a total option:
for (i in unique(chick_m$diet)) {
diet <- subset(chick_m, diet==i)
table_to_print <- dcast(chick_m, time ~ variable, mean)
print(htmlTable(table_to_print, total=TRUE))
}
Note: You do not need to specify the rgroup element if you are not using it.
#author
I have been unable to get external code chunks to work in knitr from a LaTeX (*.Rtex) document.
foo.R contains:
## #knitr foo
#
tx <- seq(0,2*pi,length.out=100)
ty <- cos(tx*3.0)
plot(tx,ty,pch=20,col='blue')
seq(3)
The test1.Rtex document contains:
%% read_chunk('Code/foo.R')
%
%%
%% begin.rcode foo, echo=TRUE, tidy=TRUE, eval=TRUE
%% end.rcode
knit('test1.Rtex') yields:
chunk "foo" is empty or set not to be evaluated
I have a copy of the knitr git repository. Modifying this code with cat(sprint()) statements there is no evidence of read_chunk() being used in LaTeX code.
I've created a for loop that outputs several plots (via ggplot2) from R into a single .tex file using the tikzDevice package. This makes it easier to include multiple diagrams from within a latex document using a single command that points to the .tex file outputted from R (say 'diagrams.tex'): \include{diagrams}.
However, I would also like to wrap each tikzpicture with the \begin{figure} environment, so that I can insert two additional lines into each respective figure: \caption{} and \label{}.
Question: is there a way to include the figure wrapper, caption, and label latex commands directly, for each respective ggplot image (from my R loop), in the outputted .tex file?
Here is reproducible R code that generates a file 'diagrams.tex' containing 3 ggplots:
require(ggplot2)
require(tikzDevice)
## Load example data frame
A1 = as.data.frame(rbind(c(4.0,1.5,6.1),
c(4.0,5.2,3.5),
c(4.0,3.4,4.3),
c(4.0,8.2,7.3),
c(4.0,2.9,6.3),
c(6.0,3.9,6.6),
c(6.0,1.5,6.1),
c(6.0,2.7,5.3),
c(6.0,2.9,7.4),
c(6.0,3.7,6.0),
c(8.0,3.9,4.2),
c(8.0,4.1,3.5),
c(8.0,3.7,5.8),
c(8.0,2.5,7.5),
c(8.0,4.1,3.5)))
names(A1) = c("state","rmaxpay","urate")
i = 2
## name output file
tikz( 'diagrams.tex' )
for (i in 2:4){ #begin LOOP
st = i*2
df = NULL
df = subset(A1, state == st , select = c(2:3))
print( # start print
ggplot(df, aes(rmaxpay,urate)) + geom_point()
) # end print
} #end LOOP
dev.off()
There may be a way to do this with plot hooks but as it is you can do it by using the console option and sink():
require(ggplot2)
require(tikzDevice)
## Load example data frame
A1 = as.data.frame(rbind(c(4.0,1.5,6.1),
c(4.0,5.2,3.5),
c(4.0,3.4,4.3),
c(4.0,8.2,7.3),
c(4.0,2.9,6.3),
c(6.0,3.9,6.6),
c(6.0,1.5,6.1),
c(6.0,2.7,5.3),
c(6.0,2.9,7.4),
c(6.0,3.7,6.0),
c(8.0,3.9,4.2),
c(8.0,4.1,3.5),
c(8.0,3.7,5.8),
c(8.0,2.5,7.5),
c(8.0,4.1,3.5)))
names(A1) = c("state","rmaxpay","urate")
i = 2
fn <- "diagrams.tex"
if(file.exists(fn)) file.remove(fn)
for (i in 2:4){ #begin LOOP
st = i*2
df = NULL
df = subset(A1, state == st , select = c(2:3))
cat("\\begin{figure}\n", file = fn, append=TRUE)
sink(fn, append=TRUE)
tikz(console = TRUE)
print( # start print
ggplot(df, aes(rmaxpay,urate)) + geom_point()
) # end print
dev.off()
sink()
cat(paste("\\caption{figure}\\label{fig:",i,"}\n",sep=""), file = fn, append=TRUE)
cat("\\end{figure}\n", file = fn, append=TRUE)
} #end LOOP