Re:Starting data analysis with R: Things I wish I'd been told
by sytelus on Monday November 24, 2014 @04:15AM
(#8651587)
R is very acquired taste - or rather you struggle through until you can memorize all the gotchas. For people who are not doing doing data analysis as a full time and perhaps only job, remembering intricacies of R is going to be deal breaker. Personally I would shift towards Python unless there are exclusive packages in R that you have to have it - and that's becoming rarer by the dat. Combined with iPython Notebook and recent wave of migrations of packages to Python, I see little need to deal with R on regular basis.
[ Parent
| on HN ]
Re:Starting data analysis with R: Things I wish I'd been told
by micheljansen on Monday November 24, 2014 @04:59AM
(#8651679)
Thanks for introducing me to iPython Notebook. I had never heard of this, but it looks very promising. The browser-based interface does not look as polished as R's studio application and I'm a bit confused why this functionality is not in the Qt console, but I'll check it out nonetheless.
Re:Starting data analysis with R: Things I wish I'd been told
by afarrell on Monday November 24, 2014 @09:26AM
(#8652463)
Much of python's power for data science comes from the scipy/numpy libraries and the tools built on top of them. Unfortunately, installing these requires first installing a bunch of fortran libraries. Fortunately, There is a free distribution of python that comes with them and comes with a package manager you can use to install python libraries that have C extensions. You can also install R packages and use them from iPython notebook. https://store.continuum.io/cshop/anaconda/