[R] Text Mining in Non English Speaking Countries

Hello All, I am interested in conducting text mining in languages other English. My understanding is the following R packages can analyze alternative (to English) languages: 1. "topicmodels" 2. "snowball" 3. "tm" Can anyone confirm? Specifically, I am interested in Hindi

[R] Trouble Runing R Version 2.13.0

To Whom It May Concern, I have an R script that requires the Windows XP platform, version 2.13.0. When launching R, however, I get the error message below: Error in loadNamespace(name). there is no package called "kernlab." Unable to restore saved data in .RData. Note that I also have version

[R] Problem Installign Packages

To Whom It May Concern, In attempting to install the "e1071" package, I get the message below after selecting the mirror site. -- Please select a CRAN mirror for use in this session --- Warning: unable to access index for repository http://mirror.fcaglp.unlp.edu.ar/CRAN/bin/windows/contrib/2.1

[R] Running R from SAS

Hello All, I am interested in running an R program with several random seeds. One approach is to launch the program from SAS. The recommended approach is to use the X command as shown below: OPTIONS XWAIT XSYNC; X "r.exe --no-save --quiet <""c:\temp\r\program.r"" >""c:\temp\r\program.log""";

[R] Question About rscaleUsage in Bayesm

To Whom It May Concern, In using the Bayesm package titled "rscaleUsage," it does not reports "mudraw," "taudramw," "sigmadraw," "Lambdadraw," and "edraw." However, it does not report the transformed y ratings or latent z scores. Is there any way to infer the value for y and z? Ziad

[R] rscaleUsage Function

To Whom It May Concern, In using the Bayesm package titled "rscaleUsage," it does not reports "mudraw," "taudramw," "sigmadraw," "Lambdadraw," and "edraw." However, it does not report the transformed y ratings or latent z scores. Is there any way to infer the value for y and z? Ziad Elmou

[R] Bayesian Belief Networks in R

Hello All, Is anyone aware of an R package for Bayesian Belief Networks (BBN) that can handle latent (hidden) nodes? I am referring to process learning the structure and the parameters of the model. Based on the documentation, the "bnlearn" package does not provide for such capability. In ad