Romain Francois, Professional R Enthusiast

This is a really simple code snippet that shows how to get the list of CRAN packages and their titles from the html page html page (toulouse mirror in this example).

...

Note that R has the available.packages function, but it does not give the titles of the packages

This post is motivated by this question on R-help. This is a simple java class that sends files through Rserve using the classes RFileInputStream and RFileOutputStream

Then we create a simple file on the client machine:

$ cat > testfile.txt
bla bla
^C

And we are good to go:

$ javac -cp .:REngine.jar:Rserve.jar RserveWire.java
$ java -cp .:REngine.jar:Rserve.jar RserveWire testfile.txt serverfile.txt
/tmp/Rserv/conn6
writing the client file 'testfile.txt' to the server as 'serverfile.txt'
writing the server file 'file.txt' to the client as 'file.txt' 

Now in the directory /tmp/Rserv/conn6 of the server, there are the files "serverfile.txt" and "file.txt"

$ cat serverfile.txt 
bla bla
$ cat file.txt
 [1] -1.16541741 -0.55857285  2.19752036 -0.78432188  1.40739981 -0.87252966
 [7] -0.11545651 -0.36735874 -2.75736666  0.29798096 -0.86836355 -0.03416198
[13] -0.44344089  0.88976360  0.58821334 -0.10354205 -0.88760475 -0.64608338
[19]  0.96552319 -1.57166441 -0.19010633 -1.42239696  0.49363257  0.06167547
[25]  0.34801546 -0.41211734 -0.20320050 -1.45370497  1.34383425 -0.89461504

and on the client there is also the "file.txt"

$ cat file.txt
 [1] -1.16541741 -0.55857285  2.19752036 -0.78432188  1.40739981 -0.87252966
 [7] -0.11545651 -0.36735874 -2.75736666  0.29798096 -0.86836355 -0.03416198
[13] -0.44344089  0.88976360  0.58821334 -0.10354205 -0.88760475 -0.64608338
[19]  0.96552319 -1.57166441 -0.19010633 -1.42239696  0.49363257  0.06167547
[25]  0.34801546 -0.41211734 -0.20320050 -1.45370497  1.34383425 -0.89461504

I pushed 3 more graphics from Biecek Przemyslaw to the graphics gallery

A list of popular names for colors from packages RColorBrewer, colorRamps, grDevices

A set of examples of few graphical low-level parameters lend, ljoin, xpd, adj, legend(), axis, expressions, mtext, srt etc

Examples for different settings of:

• the type and width for lines
• and the type and size for points

Przemyslaw Biecek has submitted this graph (and also others I will add later) to the graphics gallery

A list of examples for the atomic functions polygon(), segments(), symbols(), arrows(), curve(), abline(), points(), lines(). this figure is taken from the book Przewodnik po pakiecie R

I've pushed to CRAN the initial version of the RImageJ package. ImageJ is a java based image processing and analysis platform

This version of the package creates an instance of the IJ class as the IJ object in R, so that many methods of this class can be called using the dollar expansion of rJava.

Here is a simple example that uses the package:

download.file( "http://www.google.fr/intl/en_en/images/logo.gif", 
    dest = "google.gif" )

image = IJ$openImage( "google.gif" )
image$show()
IJ$run( "8-bit" )
IJ$run( "Invert" )
IJ$save( "bw-google.gif" )
image$close()

This downloads the google logo, convert it to black and white and inverts it

• integration of imagej and javagd
• integration of the imagej macro recorder so that it creates R code instead of code in the imagej macro language

I've now updated the completion popups used in the advanced editor plugin to get a better display of the available information

For completion items that are related to functions with a help page, we now get this popup

When completing arguments of functions, the text of the argument is grabbed from the help page (if available), and displayed as such:

For functions that do not have help page, such as functions in the global environment, a deparse of the function is displayed

It is probably better to build the plugin from source, but otherwise I have posted a build here

I am currently working on ways to bring this completion mechanism to the console as well

On this post, I presented the xterm256 package for R, allowing to have text in background and foreground color in the R console

The drawback of relying on xterm escape sequences is that the package needs to be used within a terminal that supports this escape sequence protocol (basically some linux consoles)

Here, I am proposing a patch to the biocep workbench that emulates support for these colors directly in the biocep R console, see the screenshot below:

Combined with the syntax highlighter I am working on, this allows syntax highlighting of R code (pretty close to the way R sees the code) directly in the code and is in my view both visually pleasing and very useful

The functionality requires minor modifications of the source code of biocep that I have written and I can send a patch to interested users, but I unfortunately cannot commit the modifications to the biocep project because it is read-only at the moment

# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
# 
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
# 
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

#' search on gmane.org mailing list archives
#'
#' @param string query
#' @param group search on which gmane group
#' @param author author of the email
#' @param sort sort order criteria
#' @param op how to combine words
#' @param r.group R related group to use, the default "*" means all r related groups
gmaneSearch <- function( string, 
    group = paste( "gmane.comp.lang.r.", r.group, sep = ""), 
    author = "", sort = c("relevance", "date", "revdate"), 
    op = c("and", "or"), 
    r.group = "*" ){
    
    sort <- match.arg(sort)
    op <- match.arg( op )
    
    url <- sprintf( 
        'http://search.gmane.org/?query=%s&author=%s&group=%s&sort=%s&DEFAULTOP=%s', 
            gsub( ' +', '+', string),  author,  group,  sort, op )
    url <- URLencode( url )
    browseURL( url )
}

#' retrieves the list of gmane groups associated with a prefix
#' 
#' @param prefix group prefix
gmaneGroups <- function( prefix = "gmane.comp.lang.r." ){
    url <- URLencode( sprintf( "http://dir.gmane.org/index.php?prefix=%s", prefix) )
    txt <- grep( '^<tr.*<td align=right.*<a', readLines( url ), value = TRUE )
    
    rx <- '^.*?<a href="(.*?)">(.*?)</a>.*<td>(.*?)</td>.*$'
    out <- data.frame( 
        url = gsub( rx, "\\1", txt ), 
        group = gsub( rx, "\\2", txt ),
        description = gsub( rx, "\\3", txt ), 
        stringsAsFactors = FALSE
        )
    out$group <- sub( "...", ".*", out$group, fixed = TRUE )
    out
}

So for example:

R> gmaneSearch ("browser prompt", author="romain" )

This is the sequel of this post where we look at ways to build javadoc documentation as part of an R package

Ant has the javadoc task which we will use to build html files from java sources, leading to this modified build script with the new javadoc target:

   1 <project name="Hello Java World" basedir="." default="build" >
   2 
   3   <property name="target.dir" value="../inst/java" />
   4   <property name="javadoc.dir" value="../inst/javadoc" />
   5   
   6   <target name="clean">
   7     <delete dir="bin" />
   8     <delete dir="${javadoc.dir}" />
   9   </target>
  10   
  11   <target name="compile">
  12     <mkdir dir="bin"/>
  13     <javac srcdir="src" destdir="bin" />
  14   </target>
  15   
  16   <target name="javadoc">
  17     <mkdir dir="${javadoc.dir}" />
  18     <javadoc access="public" destdir="${javadoc.dir}"
  19       author="true" version="true" use="true" 
  20       windowtitle="helloJavaWorld - Java API">
  21         
  22        <fileset dir="src" defaultexcludes="yes">
  23             <include name="**/*.java"/>
  24         </fileset>
  25           
  26     </javadoc>
  27   </target>
  28   
  29   <target name="build" depends="compile,javadoc">
  30     <jar jarfile="${target.dir}/hellojavaworld.jar">
  31       <fileset dir="bin" />
  32     </jar>
  33   </target>
  34   
  35 </project>
  36 

so that when we run "R CMD build", the directory inst/javadoc gets created and filled with documentation generated from the java sources

The next thing we need is a way to access this documentation from R documentation, so we will just add a link to it from the Rd file, I have added this into the helloJavaWorld.Rd file in the seealso section:

18 \seealso{\link[helloJavaWorld:../javadoc/index]{The Java API of the package}}

This is not a perfect solution but gives a starting point. Potential issues, things to improve :

• It would be better to have a link in the 00Index.html file, but there currently is no way to do something like that
• A way to jump back to the R documentation from the java documentation is needed

I have posted a copy of the built package here, and you can download the source of this version here

Next step, use the junitreport ant task to make a report of unit tests of the java code