convore.json/groups/rstats/python-interfaces-to-r/messages.json

[{"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1299786984.5450251, "message": "there seems to be many... e.g. RSPython, rpy, and even a redesigned version rpy2 (see http://rpy.sourceforge.net) -- what works well for you, and why?", "group_id": 6150, "id": 315830}, {"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1299787351.741003, "message": "My *nix alternative would be to use <<HereDocument scripting with Python and R within a bash script to also get access to the shell as a bonus feature.", "group_id": 6150, "id": 315874}, {"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1299915843.9908781, "message": "Relatively new interface called \"PypeR\" -- see http://www.jstatsoft.org/v35/c02/paper for PDF or for the source code http://rinpy.sourceforge.net/", "group_id": 6150, "id": 330741}, {"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1299916425.7911971, "message": "Ryan Rosario gives a walkthrough of his data mining code using RPy2 as interface to R, http://www.bytemining.com/2010/10/accessing-r-from-python-using-rpy2/ -- the video gets going after the first ten minutes (35 minutes total running time). The caveats he points out are worth noting.", "group_id": 6150, "id": 330771}, {"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1300042895.904784, "message": "R vs. Python/Numpy/SciPy : worth reading, http://stackoverflow.com/questions/1177019/what-can-be-done-in-r-that-cant-be-done-with-python-numpy-scipy", "group_id": 6150, "id": 342550}, {"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1300134322.1402881, "message": "about statistical computing in Python proper: http://stats.stackexchange.com/questions/1595/python-as-a-statistics-workbench", "group_id": 6150, "id": 350651}, {"user_id": 10890, "stars": [], "topic_id": 12097, "date_created": 1300134673.2061219, "message": "generally Google does not use R in production, however, it is used extensively by  software engineers for exploratory data analysis and to prototype models. These models are then coded in Python (or C++) since that's a better fit for distributed computing with massive data.", "group_id": 6150, "id": 350725}]