Venable's short course -- materials

Atention: the download of the course materials below DO NO WORK WITH INTERNET EXPLORER. Use another browser, as for instance Mozilla Firefox

1. ASOR package (already installed in the C3SL servers)
   • .tar.gz file (for linux install)
     NOTE: the WIKI system makes all files lower case(!) so, before installing this in your LINUX box you need to rename the file:

     $ mv asor_1.0.tar.gz ASOR_1.0.tar.gz
     $ sudo R CMD INSTALL ASOR_1.0.tar.gz

   • .zip file (for windows install)
2. Text: Exegeses in Linear Models
3. Additional scripts:
   • Boston housing price -- setting training and test data for tree model using ''rpart()'' and ''randomForest''
   • Menarche data -- GLM for binomial counts
   • the plinear method for non linear models

At UFPR the course will be run in the LM (Laboratório Multiterminal) run by C3SL (Contro de computação científica e Softaware Livre). The LM provides LINUX terminals connected with servers. The user working area is common meaning you can user whatever terminal and/or server to get access to your account.

In the following instructions the LINUX prompt is $ and the R prompt is >.

1. Open a Linux terminal window for issuing LINUX commands
2. create a working directory called, for example, Rmodelling for use in this course move to it:

   $ mkdir Rmodelling
   $ cd Rmodelling

3. download the zipped files above (csv.zip, scripts.zip, slides.zip) copying them to the working directory
4. unzip the files with the commands below. This will create the subdirectories CSV, Scripts and Slides:

   Rmodelling$ unzip csv.zip 
   Rmodelling$ unzip scripts.zip 
   Rmodelling$ unzip slides.zip

5. copy one file from the CSV directory to the working directory:

   Rmodelling$ cp CSV/makeInputData.r .

6. Start R inside the Rcourse directory — this will make Rcourse your working directory. You can do this in two different ways:
   • Running R directly from the LINUX terminal simply by typing R:

     Rmodelling$ R

   • running from inside the editor xemacs using the ESS (Emacs Speaks Statistics) package. To do so open xemacs and stard R from it with ESC SHIFT-X SHIFT-R

     Rmodelling$ xemacs makeInputData.R &

7. From inside R run the makeInputData.r script

   > source("makeInputData.r")

8. The script will automatically generate another script file called inputData.R which reads all the data needed for the course. This script file should be run with:

   > source("inputData.R")