Comments on: R-[{Perl,Python}] Interface http://hea-www.harvard.edu/AstroStat/slog/2008/r-interface/ Weaving together Astronomy+Statistics+Computer Science+Engineering+Intrumentation, far beyond the growing borders Fri, 01 Jun 2012 18:47:52 +0000 hourly 1 http://wordpress.org/?v=3.4 By: TomLoredo http://hea-www.harvard.edu/AstroStat/slog/2008/r-interface/comment-page-1/#comment-230 TomLoredo Thu, 22 May 2008 04:14:29 +0000 http://hea-www.harvard.edu/AstroStat/slog/?p=307#comment-230 Astronomers familiar with Python who would like to access R's capabilities from within Python should also explore: <a href="http://rpy.sourceforge.net/" rel="nofollow">RPy</a> It's a one-way interface (you can call R from Python, but not Python from R), as opposed to RSPython (Hyunsook's "R/SPlus-Python" link). I briefly played with both, with Bill Jefferys (a big R fan), back at SAMSI in Spring 2006. From the Python side, we felt RPy was the more natural (and more stable) tool, at least back then. RPy is directly inspired by RSPython, but focuses on the R-from-Python part of the problem, and so may be a superior solution if that's the only direction you need. While on the topic of Python, I'd like to point NumPy/SciPy users to the <a href="http://www.scipy.org/Developer_Zone/DocMarathon2008" rel="nofollow">NumPy documentation marathon</a>, a project spearheaded by astronomer & Python fan Joe Harrington, seeking to make Python's scientific computing documentation competitive with that of Matlab and IDL. The Marathon team hopes to accomplish this by combining funded support of some dedicated writers and editors, with extensive community input (documentation text submitted via a Wiki web page). Right now they are focusing on NumPy (hoping to have good docs in place by Fall); SciPy will follow. Readers who use Python should check out the project and see if they can contribute a docstring or two, if they have some expertise with NumPy. The statistical expertise of Slog readers could be especially valuable when they move to SciPy and need documentation for the stats module. Astronomers familiar with Python who would like to access R’s capabilities from within Python should also explore:

RPy

It’s a one-way interface (you can call R from Python, but not Python from R), as opposed to RSPython (Hyunsook’s “R/SPlus-Python” link). I briefly played with both, with Bill Jefferys (a big R fan), back at SAMSI in Spring 2006. From the Python side, we felt RPy was the more natural (and more stable) tool, at least back then. RPy is directly inspired by RSPython, but focuses on the R-from-Python part of the problem, and so may be a superior solution if that’s the only direction you need.

While on the topic of Python, I’d like to point NumPy/SciPy users to the NumPy documentation marathon, a project spearheaded by astronomer & Python fan Joe Harrington, seeking to make Python’s scientific computing documentation competitive with that of Matlab and IDL. The Marathon team hopes to accomplish this by combining funded support of some dedicated writers and editors, with extensive community input (documentation text submitted via a Wiki web page). Right now they are focusing on NumPy (hoping to have good docs in place by Fall); SciPy will follow. Readers who use Python should check out the project and see if they can contribute a docstring or two, if they have some expertise with NumPy. The statistical expertise of Slog readers could be especially valuable when they move to SciPy and need documentation for the stats module.

]]>