3 ms·
Using lisp as language for statistics has it's precursors like xlisp-stat. In a historical perspective, S (or similar functional languages) have taken the crown
by xtho 17y ago
Using lisp as language for statistics has it's precursors like xlisp-stat. In a historical perspective, S (or similar functional languages) have taken the crown though.
- whyenot 17y agoSomewhere in the archives of comp.lang.lisp there is a post by Ross Ihaka where he mentions that they wanted to use scheme or another lisp-like language for R, but decided to go with something more familiar to most potential users. If I remember right, he said if they could do it all over again, they would use lisp. There's also this, which I believe partly motivated Incanter's creation: http://books.google.com/books?id=8Cf16JkKz30C&pg=PA21#v=onepage&q=&f=false http://books.google.com/books?id=8Cf16JkKz30C&pg=PA21#v=...
- xtho 17y agoIMHO the popularity of R is grounded on its compatibility with S/S-Plus. if they hadn't used S as language, they would have never reached the position where they are now and R wouldn't be increasingly adapted as standard statistics package---competing with and/or replacing SPSS/PASW, SAS, stata etc. On the incanter homepage there is a link to http://www.jstatsoft.org/v13 http://www.jstatsoft.org/v13 which includes a few interesting articles. Anyway, I find the integration with java interesting. That's a strong argument in favour of a solution like incanter.