Re: How good is R?


casioculture@xxxxxxxxx wrote:
> I'm investing my time in learning R, and also gnumeric. I guess
> gnumeric is somewhat trivial, and so far I found R easy to learn. I
> must say I'm spending most of my time relearning statistics itself. I
> made the decision to do so as they're open source. I decided not to
> learn SPSS and Excel, as I don't have SPSS and Excel is equally
> proprietary, though I'm inclined to think that by understanding R and
> being familiar with gnumeric, and the relevant statistics, then getting
> to know SPSS and Excel if I encounter them somewhere should be no major
> issue. Am I correct?

I have not used SPSS in years and like you I am trying to teach myself
R but as I remember SPSS, it is nothing like R.

I have used Excel and played a bit with Gnumeric but not for
statistics. I don't think you should ever consider using a spread***
for statistics unless you have absolutely no choice and even then you
might not want to use Excel. It has a fairly large number of known
problems or inadequacies

Here are a few links you might find useful
http://www.cs.uiowa.edu/~jcryer/JSMTalk2001.pdf
http://pages.stern.nyu.edu/~jsimonof/classes/1305/pdf/excelreg.pdf
http://www.csdassn.org/reportdetail.cfm?ID=508

>
> Also, I know that gnumeric is more powerful than excel. How does R
> compare to SPSS?

.

Loading