Re: [R] Scripting in R -- pattern matching, logic, system calls, the works!

2008-09-16 Thread Dan Davison
Instead of writing some long, ugly, "script", the way to use R is to break problems down into distinct tasks. Reading data is one task, and performing regressions on the data, plotting & summarising are different tasks. Write functions to do each task in general, and then use those functions. So o

Re: [R] Scripting in R -- pattern matching, logic, system calls, the works!

2008-09-16 Thread bioinformatics_guy
Don, Excellent advice. I've gone back and done a bit of coding and wanted to see what you think and possibly "shore up" some of the technical stuff I am still having a bit of difficulty with. I'll past the code I have to date with any important annotations: topdir="~" library(gmodels) setwd(to