Re: [R] User input when running R code in batch mode
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Kaushik Krishnan wrote: Is there any way to make R stop for the user to enter values when running in batch mode either by changing the way I invoke scan() or readLines() or by using any other function? At least on linux this works: be...@t40:~/r-test $ cat test.R readline2 = function(prompt=){ cat(prompt) readLines(stdin, 1) } print(readline2(enter a number: )) be...@t40:~/r-test $ Rscript test.R enter a number: 42 [1] 42 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iD8DBQFK5xUSxT6R4jlFoh0RAkhTAKCB4qUjPlULL9vjUaLLxarTor2z1QCeNGHA 4B2VmkqtyEKMKKGUnt8sjiY= =VYJs -END PGP SIGNATURE- __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] API question (embedding R)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hello list, i want to access R from a Pascal program. Basically i only need to be able to * start the engine / the interpreter (and keep it running) * pass strings to it (R code) that should be parsed and executed * get floats or vectors of floats into R and assign them to R variables * read the contents of variables and get them back into my Pascal program. (I dont need this for all possible data structures, just floats and maybe vectors of floats would be sufficient) At the moment i have something running that does exactly this via Python and the rpy bindings, this works but is horribly complicated to install because of all the needed dependencies. It works like this: the pascal program loads the python interpreter (this API is extremely simple, i can get floats and lists of floats in and out of python variables and have it compile and execute python code), the pascal program then lets python import the rpy bindings, put the data into python variables, from there (by executing python code) move the data into R variables and then finally execute R code (by executing python code that would call into R to execute R code contained in a string literal). Getting the results back goes a similar way. Now i want to make this more straightforward, i need a way to directly access the R interpreter just like i can already access the python interpreter. Using the python interpreter from pascal is a child's play, you really only need to import a handful of functions from the dll and nothing else, you just call Py_Initialize() and then you can call PyRun_SimpleString() as often as you want and thats basically all! No need to convert the whole C header file to pascal, it all fits more or less into 5 lines of pascal declarations. I couldn't find an easy documentation of how i could achieve the same thing with R: load the dll, import only a few needed functions (which ones do i actually need?) and how to use them. Is there a minimal example somewhere? What documents should I read? Maybe i missed something important? An example would actually be the most important thing, an example tells more than thousand words. Thanks in advance, Bernd Kreuss -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iD8DBQFK5ikSxT6R4jlFoh0RAi8LAJwMiDRVWLZBZp2JYynpCtnS48r25gCgzrLF s5hYB20RCr2RuxIPe8EW8ws= =erKc -END PGP SIGNATURE- __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] How to stop an R script when running JGR on a Linux/SuSE system
This issue was addressed in a recent discussion [1]. Liviu [1] http://mailman.rz.uni-augsburg.de/pipermail/stats-rosuda-devel/2009q2/001106.html I would like to add a few points to this list, some of them I personally find even more annoying then some of them mentioned there: - add the keyboard shortcuts shift-del and shift-ins for the actions cut and paste as an alternative to ctrl-x and ctrl-c to make it compliant with CUA and behave like virtually every other editor in existance. (this is my most important concern, i use it all the time when moving rearranging lines, being used to eclipse, SciTE, Kate, even MS-Word, OpenOffice or notepad on windows, everytime i try to cut/paste I already have a finger of the left hand on the shift-key (from selecting the text) and it takes one or two seconds to realize that the intended shortcut is missing and then move both hands to find the other two keys on the keyboard) - change the keyboard shortcuts for block-indent and block-dedent to the more commonly used tab and shift-tab - remove the restriction of a minimal window size for the console window, allow the user to resize it as small as he wants. - add a shortcut for run all, this is the only menu item without shortcut. F5 would be a good choice as some other widely used editor/shell combinations for other languages have similar functionality on that key too (SciTE, IDLE, etc.), it seems to be some kind of standard too. Bernd __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] How to stop an R script when running JGR on a Linux/SuSE system
Bernd Kreuss wrote: I would like to add a few points to this list [...] I would even make the changes on my own (i probably would already have done it) and supply patches if i only could find any hint on how to build JGR from sources. (where to place the source files, what command to start the build process and where to copy which files after successful built to test them) So this is another point on my list: write a small howto on how to participate in development. Bernd __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] How to stop an R script when running JGR on a Linux/SuSE system
Liviu Andronic wrote: other missing dependencies), fetch the source archive [1] from CRAN, I tried this already but this source archive only contains (besides some C code for the starter) only one .jar file without any Java sources in it, only .class files. Also i tried to check out the project directly from svn but there is nothing java related at all. The source code must be elsewhere. I'm sure that JGR would make much faster progress if people were able to just download a tar.gz with all relevant files in it together with a readme.txt and a makefile or the java equivalent of it. If something like this already exists it would be nice to put a link to it directly on the JGR homepage, I couldn't find anything. Bernd __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] How to stop an R script when running JGR on a Linux/SuSE system
mau...@alice.it wrote: I wonder whether there is a more gentle way to stop an R script running on top of JGR aother than ... unplugging the power cord. there must be a bug in JGR on Lunux. Clicking the stop button should stop the script, clicking it here on my linux machine will immediately crash R together with JGR altogether. Unfortunately JGR still seems to be the best (and only) available shell/editor combination available. When running R in an ordinary terminal window execution can be terminated with CTRL-C I hope they will soon fix this (and a small handfull of other bugs/flaws, mostly missing/wrong keyboard shortcuts) and JGR would make a huge step from very good to excellent Bernd __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] How to stop an R script when running JGR on a Linux/SuSE system
sorry for the eventual double posting, but i got a strange error from a versatel(???) server about not enough quota when replying to the message __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Forumla format?
Steve Lianoglou wrote: Is it possible to build up your formula as a string, and then convert to formula w/ as.formula? what about simply using it this way instead: svm(label ~ ., data=mydata[,-c(22,23,25,31)]) __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.