I wonder whether there is/are reliable Graphical User Interface distributions of R statistical software comparable to commercial software packages such as SPSS. Existence of detailed manuals, being free of charge are the most important desirable qualities for me.
In general, I would recommend against using a menu-driven GUI for R. Instead I would work through some websites, books, or MOOCs to get familiar with how to use R. For example, the following two websites focus on simple, reproducible examples of common statistical tests without asking you to learn much about the R language.
http://rcompanion.org/rcompanion/
http://www.statmethods.net/
I recommend using RStudio as an interface to make working with R easier (https://www.rstudio.com/).
If you do want to use a menu-driven GUI, the package Rcmdr is pretty good (http://www.rcommander.com/). I like that it shows you the R code, so you can learn how to code the tests. I believe there are also extensions for this package to expand the range of analyses.
Another possibility is BlueSky statistics, which is a commercial product but has a free version (http://www.blueskystatistics.com/). It also shows you the code it is running.
I think this style of GUI is okay to get someone started in learning R, but pretty quickly they will find it quicker and more flexible to just use the code directly.
In general, I would recommend against using a menu-driven GUI for R. Instead I would work through some websites, books, or MOOCs to get familiar with how to use R. For example, the following two websites focus on simple, reproducible examples of common statistical tests without asking you to learn much about the R language.
http://rcompanion.org/rcompanion/
http://www.statmethods.net/
I recommend using RStudio as an interface to make working with R easier (https://www.rstudio.com/).
If you do want to use a menu-driven GUI, the package Rcmdr is pretty good (http://www.rcommander.com/). I like that it shows you the R code, so you can learn how to code the tests. I believe there are also extensions for this package to expand the range of analyses.
Another possibility is BlueSky statistics, which is a commercial product but has a free version (http://www.blueskystatistics.com/). It also shows you the code it is running.
I think this style of GUI is okay to get someone started in learning R, but pretty quickly they will find it quicker and more flexible to just use the code directly.
I think most people I know use R Studio as suggested by Salvatore......
R for Dummies, a useful introduction to R, is available at the link below. I agree with the previous two suggestions. Best wishes.
http://www.ievbras.ru/ecostat/Kiril/R/Biblio/R_eng/R%20dummies.pdf
65
I am using RStudio, which is a powerful and intuitive interface for R. I second the recommendation above. RStudio will help you by not only displaying the syntax of commands, but by showing lists of things that you can select from.
https://www.rstudio.com/products/rstudio/download/
I second the previous advice against GUIs for R : programming is a bitch, but gives you benefits unreachable by point-and-click interfaces. The current "best" solution is to work with tools allowing you to mix text (explaining your work) with R code. Look at the R packages knitr for the (currently) best realization of this idea ; that's what RStudio uses.
A nice alternative to RStudio is the emacs text editor and the ESS emacs package (interface between emacs and R). It is exceedingly retro (text-mode interface with key chords, plus a few menus as a (reluctant) concession to modernity), has a steep learning curve, but gives you a large productivity boost over all GUI editors I'm aware of.
If you just want a nice workbook, use markdown (the polymode emacs package is handy, and works well with the Rmarkdown R package, which "wraps" knitr) ; if you plan to publish, use LaTeX (I highly recommend the auctex emacs package for that) : again, steep learning curve, but with enormous benefits...
Thank you all for your contributions. I have installed R studio and now am able to conduct many basic statistics on R also thanks to your replies. It is amazing to see what enormous variety of things people come out with using R.
There is nothing better than RStudio running on a 8-core AMD/SSD Linux Mint (optimized for speed) workstation and a large (very large) screen. Free as in Freedom! A true weapons system, I would say! Cheers!
RStudio is a good start. But as told above, use some time to understand basic R, Rmarkdown (there are lots of very good reproductible stuff online in Rpubs and other sites). A good work for latin, spanish and portuguese readers are: http://saree.com.mx/econometriaR/sites/default/files/Ebook_econometriaR.pdf; and the course at this github: https://github.com/pedrocostaferreira/SeriesTemporais
Here comes a new challenger: R++, the Next Step
Well, it is not ready yet. The beta test will start september 5th. But here is a 1 minute preview: http://bit.do/erdLx
Jamovi and JASP are also possibilities.
We'll see what R++ has to offer.
But, honestly, any point-and-click GUI I've seen falls short of what I want to present to undergraduate students.
I will not start a troll "point-and-click" vs "code" because I myself am a big fan of coding. Bug there is coding and coding:
- Coding a nested analysis, longitudinal kmeans in 3D or random forest, yes! I love that, it is what I am good at.
- Coding to find & correct abnormal value, or to merge "Male", "male", "M" and "malle" into "Male"? Boring...
That's the spirit of R++: click for the boring part, code for the expert part. But as you say: we'll see!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics