Open Source Stats Programs?
Posted 3 months ago at 1:22 pm. 6 comments
My wife is looking for an Open Source Statistics program for Linux that does multinomial logistic regression and can import from spss, stata, or other popular statistic programs. Please e-mail me (mike.basinger(at)gmail.com) if you know of any programs that could work.
There is PSPP, which is a free software implementation of SPSS. Of course, some have said that it stands for, “People Should Prefer PSPP.”
I think that it is command-line only, though, and I don’t think that it is a complete implementation of the SPSS language, but I don’t know all about it.
I’ll be following this thread to see what others say here, too.
Apparently our stats professors now teach students R, you might look to see if that meets your needs.
Hi there. I was exploring different stat programs a while back myself and just realized that many if not most of them use different programming languages at the very core. SPSS uses a proprietary language and by extension unique commands and syntax afaik, and so does SAS. I don’t know how this affects interoperability between them.
‘R’ is known to be a powerful opensource stats package and I suggest you take a look at it. There are many GUIs available that work with R, most noteworthy of which is R-Commander. Cheereo.
Yes: I am a statistician by education and I can seriously say that R (http://www.r-project.org/) is the most prominent statistical software suite around, regardless whether the competitors are commercial or open source ones. CRAN (the CPAN-comparable) repository) contains almost anything you can imagine; all packages are based on research. The community is active and mostly academic; many involved people are known names within the scientific statistical community.
Learning curve can be a little deep, but it is all worth it. Also: unlike others, I would recommend not to use GUIs for R.
Finally, also Gretl (GNU regression, Econometrics and Time-Series Library; http://gretl.sourceforge.net/) can handle all basic regressions, including logistic ones, easily.
My statistics course at university was also taught using R.