R and WinBugs has good statistical capabilities but has a learning curve. Tools like Bayesialab are too costly. Are there any good "in-between" alternatives which can be used for building an applied statistics course around it?
For a full list of statistical packages, please take a look at the Wikipedia's list of statistical softwares and packages at:
http://en.wikipedia.org/wiki/List_of_statistical_packages
Moreover, I think Salstat and Bayesian Filtering Library may be good suggestions.
I'd strongly recommend mlWIN, which requires NO familiarity with Bayesian modeling. One just chooses "MCMC" as the estimation method. Incredibly simple. It runs many of the most popular models, including hierarchical ones. It is free to academics in the UK, and has reduced prices for students, I believe.
I've used WinBUGS a lot, and it's also terrific, although there is a bit of a learning curve.
I am using RStudio with JAGS. Installation instructions are available on John Kruschke's blog.
http://doingbayesiandataanalysis.blogspot.com/search/label/installation
For Bayesian analysis of inequality constrained hypotheses, there is BIEMS:
http://www.jstatsoft.org/v46/i02
Also, this book has a section that describes a range of software packages for running Bayesian analyses.
http://www.amazon.com/Informative-Hypotheses-Behavioral-Scientists-Statistics/dp/1439880514/ref=sr_1_1?s=books&ie=UTF8&qid=1365644596&sr=1-1
Hi All, As an update, please find below an important link for most popular 50 free statistical tools.
www.predictiveanalyticstoday.com/top-free-statistical-software/
Good luck!
I know this thread is old, but the question seems to reappear frequently.
In my view, JAGS has surpassed WinBUGS/OpenBUGS, both in terms of available features and speed. Calling JAGS from R (or another software) using an interface function is definitely the way to go.
First Bayes (http://www.tonyohagan.co.uk/1b/) is designed specifically for the purpose you are describing, but it seems that the project is no longer active.
Finally, if you want something menu driven, take a look at BayES (http://bayeconsoft.com/). It is primarily menu driven, but supports very few models. However, it provides graphical interfaces to JAGS and OpenBUGS, which may take some of the rough edges of using the R interface functions off. DISCLAIMER: I am the developer of BayES.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software