(GUIs, was R GUI ( was Re: [R] R Documentation(s)))

Kjetil Kjernsmo kjetil.kjernsmo at astro.uio.no
Fri May 12 17:52:37 CEST 2000


I'm following up, just a bit late... :-)

On Tue, 9 May 2000, Prof Brian D Ripley wrote:

>> Peter Dalgaard BSA wrote:
>> > 
>> > Actually, I think there are valid reasons for people to want GUI
>> > interfaces to statistics. 

There most certainly are.

>> >One argument that I hear frequently is that
>> > nonstatisticians tend to do other things than analyse data for
>> > extended periods of time, and it becomes difficult to remember the
>> > formalities from one time to the next. So they get scared of the
>> > prospect of staring empty-minded at the R prompt. 

Well, I'm not a statistician by any standards, the most advanced thing I
ever did in a course was 
glm(formula = kyphosis ~ al + an + sa, family = binomial(link = "logit"), 
    data = data),
but of course, since I am an astrophysicist, I'm pretty used to
programming, command-line interfaces etc. But, as Prof Ripley says:

>Um.  In all the statistical GUIs I have used (including those two and
>S_PLUS) the menu/dialogs interface shows you only a part of what can be
>done. (Actually, I think in SPSS for Windows it is all that can be done,
>but SPSS itself can do more.)  So my experience is the other way round: I
>know it can be done, but I can't find out how from the GUI and as GUIs are
>supposed to me self-explanatory, no other documentation is to hand.

it is far from easy to find the things you want in menu systems. If you
haven't understood the concepts, the path to doing what you need may in
fact be shorter at a command line: Take for example my glm above. I may
think that I should be looking under "hypothesis testing". I mean, there
are levels of significanse and everything, you now. :-) I will not find
anything there in a menu system under "hypothesis testing", I might find
it under "regression". On the other hand, with a piece of well written
documentation, I might not only find the glm()-function more swiftly, I
might also realize that I wasn't really doing hypothesis testing. 

I'm not against a GUI for R, but I think that well written documentation
intended for beginners of different fields of the command line interface
is far more urgent, and is something I think would attract beginners to R.

There are of course certain operations that could be performed more easily
with a GUI. For example, if I have a figure on the X-device, doing a
right-click on the figure to get a pop-up menu that lets me send it to a
postscript device or a different file (blurring the distinction) would be
great. But not urgent.

In general, I think that a GUI may be nice if the options you have are
few, but if the options you have are many, then it is just a barrier
between you and the really powerful stuff. The GUI may be designed to help
you do tasks that are common, but if you are doing things that are
somewhat out of the mainstream, even simple operations may be difficult to
perform.

Best,

Kjetil
-- 
Kjetil Kjernsmo
Graduate astronomy-student                    Problems worthy of attack
University of Oslo, Norway            Prove their worth by hitting back
E-mail: kjetikj at astro.uio.no                                - Piet Hein
Homepage <URL:http://www.astro.uio.no/~kjetikj/>
Webmaster at skepsis.no 


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._



More information about the R-help mailing list