R is a cool programming language for statistical computations. I like everything in it except it’s name. If I type into a search engine, say, “compute autocorrelation with R”, it gives me search results for “compute autocorrelation” and thinks that I typed “R” by mistake.
Well, here’s lots of R code: http://www.wessa.net/. The site consists of statistical and forecasting calculators that are free for non-commercial use. The calculators are in R with the web interface. I tested it with a few time series, and liked it.
However, some fields for entering data are too short. If the field “Number of observations” is four digits long, I can’t enter a number greater than 9999, yet all time series with which I am working are longer. This is one of inherent problems with any GUI. It introduces additional restrictions. There is virtually no limit of how many characters you can type into a command line.
The sweetest thing in that web-site is that they provide R code for their routines. You can test how it works on the web, copy the code, modify it and use in R software for longer computations.
Some more R tips are here: