While investigating "jags" for Gibbs sampling it seemed relevant to learn "R", the language. Since I already know and use about 250 language sets it makes it easier to absorb the syntax and relationships. An interesting thing is that on debian it is possible to:
sudo apt-get install littler sudo apt-get install r-base aptitude search "^r-"
"Littler" is a different implementation of R and it is invoked with the lower case r. Big R which is invoked with upper case R seems to be a little more user friendly. Once at the R command prompt it is possible to get instant gratification with:
demo(persp)
It asks to save the workspace on exit and saves it as hidden files ".RData" and ".Rhistory" in the current working directory. There really isn't a whole lot that a language can offer in the way of execution as it must be some constructed set of what is possible with a CPU. Syntax, libraries, nuance, organization, speed, and support. On a normal day it might be necessary to make a shell script to perform some task with the operating system, some quick test involving Python, extension of some C examples, or use of any number of web languages like HTML or css.
Statistical analysis and application is the real goal in this particular instance and so the name of the vessel is unimportant.
0 comments:
Post a Comment