Comments on: Considering R as a Python Supplement http://micahelliott.com/2009/03/considering-r-as-python-supplement/ Stories from my Startup Journey Wed, 08 Sep 2010 00:19:02 +0000 hourly 1 http://wordpress.org/?v=3.1.2 By: Micah Elliott http://micahelliott.com/2009/03/considering-r-as-python-supplement/comment-page-/#comment-55 Micah Elliott Thu, 19 Mar 2009 16:35:00 +0000 http://micahelliott.com/?p=311#comment-55 @Michael: Thanks for sharing those insights. Yes, I'm finding the "after" features to be really impressive. And having a lot of fun with some of the free literature and tutorials now. I'm looking forward to more of your illuminating posts (on your blog). I'd be really interested in making it to one of your Bay Area R group meetings if I can time a trip right. We were getting our own R study group into the works here in PDX, but recently switched it to the more general PDX-data-viz. We'll be covering R quite a bit there, first meeting Monday. @Michael: Thanks for sharing those insights. Yes, I'm finding the "after" features to be really impressive. And having a lot of fun with some of the free literature and tutorials now. I'm looking forward to more of your illuminating posts (on your blog). I'd be really interested in making it to one of your Bay Area R group meetings if I can time a trip right. We were getting our own R study group into the works here in PDX, but recently switched it to the more general PDX-data-viz. We'll be covering R quite a bit there, first meeting Monday.

]]> By: Michael E. Driscoll http://micahelliott.com/2009/03/considering-r-as-python-supplement/comment-page-1/#comment-51 Michael E. Driscoll Sun, 15 Mar 2009 10:52:37 +0000 http://micahelliott.com/?p=311#comment-51 Micah - Great overview. As an avid R user for a number of years, I can tell you the hard truth that R is far from perfect. Its syntax can be frustrating. On occasion, matrices become vectors without notice. But, like you, having munged large data sets in a variety of languages -- Perl, Python, Matlab, Mathematica -- I find R is the most capable at slicing and dicing data. For me, no other language allows so many ways to index into a data matrix. With R, unique character names, numeric indices, boolean values all work. Finally, it's what R can do after the data is sorted out. Few tools offer such high-level functionality for statistical analysis and graphics visualization. Micah – Great overview. As an avid R user for a number of years, I can tell you the hard truth that R is far from perfect. Its syntax can be frustrating. On occasion, matrices become vectors without notice. But, like you, having munged large data sets in a variety of languages — Perl, Python, Matlab, Mathematica — I find R is the most capable at slicing and dicing data. For me, no other language allows so many ways to index into a data matrix. With R, unique character names, numeric indices, boolean values all work.

Finally, it's what R can do after the data is sorted out. Few tools offer such high-level functionality for statistical analysis and graphics visualization.

]]> By: Micah Elliott http://micahelliott.com/2009/03/considering-r-as-python-supplement/comment-page-/#comment-52 Micah Elliott Tue, 10 Mar 2009 15:53:09 +0000 http://micahelliott.com/?p=311#comment-52 @Parand: Wonder how many of us are in that boat. I still think it will be worth broadening the horizon to see what's out there. I don't expect to hear from a lot of people who have done both. I'll plan to report on how compelling it turns out to be over the various Python facilities (which are obviously many). @Parand: Wonder how many of us are in that boat. I still think it will be worth broadening the horizon to see what's out there. I don't expect to hear from a lot of people who have done both. I'll plan to report on how compelling it turns out to be over the various Python facilities (which are obviously many).

]]> By: Micah Elliott http://micahelliott.com/2009/03/considering-r-as-python-supplement/comment-page-1/#comment-53 Micah Elliott Tue, 10 Mar 2009 15:49:38 +0000 http://micahelliott.com/?p=311#comment-53 @Charlie: Thanks for the rpy2 link. I was reading about rpy but nice to know there's a second version (which I hadn't noticed). I will certainly have a need to be integrating the languages. This could make bridging the two feasible for a lot of folks and go a long ways in helping R adoption, if Pythonistas are convinced that R's features are worth the combination effort (which doesn't look to bad). @Charlie: Thanks for the rpy2 link. I was reading about rpy but nice to know there's a second version (which I hadn't noticed). I will certainly have a need to be integrating the languages. This could make bridging the two feasible for a lot of folks and go a long ways in helping R adoption, if Pythonistas are convinced that R's features are worth the combination effort (which doesn't look to bad).

]]> By: Charlie Roosen http://micahelliott.com/2009/03/considering-r-as-python-supplement/comment-page-1/#comment-54 Charlie Roosen Tue, 10 Mar 2009 09:34:17 +0000 http://micahelliott.com/?p=311#comment-54 Take a look at rpy2 (<a href="http://rpy.sourceforge.net/rpy2.html)." target="_blank">http://rpy.sourceforge.net/rpy2.html).</a> You can use it to stick with Python as your main language and start using R for statistics/graphics as needed. Take a look at rpy2 (http://rpy.sourceforge.net/rpy2.html). You can use it to stick with Python as your main language and start using R for statistics/graphics as needed.

]]> By: Parand http://micahelliott.com/2009/03/considering-r-as-python-supplement/comment-page-1/#comment-56 Parand Mon, 09 Mar 2009 19:00:11 +0000 http://micahelliott.com/?p=311#comment-56 I also intended to take a brief look at R, but I started using SciPy and matplotlib and it turned out to have more functionality than I'll need, and was in Python, so I never made it to R. I'm happily using SciPy. I also intended to take a brief look at R, but I started using SciPy and matplotlib and it turned out to have more functionality than I'll need, and was in Python, so I never made it to R. I'm happily using SciPy.

]]>