I was going to chip in with Bugs and winBugs, mainly for Bayesians, but then saw that Bugs etc are also mentioned in the article.
The two or three newly graduated statisticians who we recruited in the 90s to help me do stats and maths for Food Scientists came with expertise in Genstat and S. I see Nelder’s Genstat is still around (https://www.vsni.co.uk/software/genstat/). Cliff was very keen on that, while Robin preferred S, the commercial precursor to its freeware alter ego, R. Ifirc, the very expensive SAS was the preferred supporting software in papers offered to Academic Journals; SPSS, also expensive, was considered a bit suspect for some models, I forget where. S and Genstat were highly regarded. Bugs etc are very impressive. I expect Cambridge, UK, is still offering 2-3 day induction Courses on Markov Chain Monte Carlo modelling, Gibbs Sampling and Bugs. I certainly enjoyed my few days on the course. Prof Spiegelhalter was one of the high-powered presenters. We used to try to justify our modelling efforts by looking at likelihood. In particular, we would attempt to apply GLIM, the generalised linear model, whether within Genstat or in the APL statistical library which was developed in the mid-90s. Sorry, a bit OTT, but might be of passing interest, Mike Sent from my iPad > On 26 Jul 2019, at 06:40, Skip Cave <[email protected]> wrote: > > Interesting evaluation of programming languages for statistical uses. > > https://www.r-bloggers.com/whats-the-best-statistical-software-a-comparison-of-r-python-sas-spss-and-stata/ > > Skip > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
