Subject: Re: [R] Introductory Resources From: Spencer Graves Date: Fri, 06 Jun 2003 09:44:56 -0700 To: RBaskin@ahrq.gov CC: R-help@stat.math.ethz.ch I've been using S-Plus almost daily for roughly 7 years and testing R daily for the past 4 months. In my opinion, roughly 90% of the differences I've seen are in R's favor, e.g., the "log" options in the probability functions. I think the future lies with R, because I guessing that much of the most advanced algorithm development must be in R: I can modify an R function and give it to the world; if others find my modification useful, I get recognized for having made a contribution to science and humanity. If I do that with an S-Plus function, I'm a thief, because I've stolen Insightful's intellectual property. In brief, I believe that R is already superior to S-Plus, and the difference is increasing. I'm telling my collaborators who use S-Plus that our new software development should be tested in both R and S-Plus so we can switch whenever we need to. Hope this helps. Spencer Graves p.s. You'll find much more discussion of this by searching the r-news archives, "http://www.r-project.org/" -> Search -> "R Search Site". RBaskin@ahrq.gov wrote: I am interested in R as an alternative for a statistical tool at our firm. Ditto... I have recently moved to this agency from a company where I had access to Splus. There is also a coworker here who had used Splus at a previous employer. We both would like some access to the S language. We are considering either begging loud and long to try to get the agency to purchase two copies of Splus or trying to convert to R. This is not an easy decision under the circumstances and purchasing Splus will mean giving up something else. On the other hand I have never used R before and I fear the learning curve for using R plus possibly ESS. There are a few things I am trying to determine before I really decide what to do. I have been trying to convert some of my old Splus script files at home to run under R 1.7.0. Small (less than 50 lines or so) script files that I have tested run exactly as before. I tried to run one large simulation and it took about a week of screaming hell to get the error messages out (well all but one error message anyway). 1) I want a test suite for R. I noted in the messages (Date: Mon Feb 24 2003 - 22:18:03 EST) that Prof Ripley wrote "Well, R itself has lots of tests in its test suite (see directory tests in the sources) packages..." but I was too stupid to find them. Q1: Can someone provide directions to this test suite that even an idiot can follow? 2) Most of the problems I ran into had to do with missing values (in effect I have ragged arrays). One silly example is that I had made use of which.na 'which' apparently is not defined in R 1.7.0. There are multiple workarounds such as simply defining a function which.na but of course it would be untested and you can loop back to 1). The problem in my script files is the same I had in Splus. I want the defaults on all of the functions (such as mean, median, etc.) I am using to be reset GLOBALLY so that the default is to ignore missing values or not. Q2: Can the missing defaults be set globally for all functions. In other words, I want the default for how to treat NAs in all functions to be set at startup. 3) What I really want to do is pass a function name and extra arguments to another function. For example, in Splus, you can pass a function such as median to the bootstrap function. The bootstrap function says that you can pass arguments to the median function through the bootstrap function but unfortunately I could never make this work. This functionality would probably solve most of my NA problems if I could make it work. (I don't seem to be able to properly use the ellipses:) Pseudo-Example: The Splus bootstrap can be called as Bootstrap(variable-name, median, sampler=sample-function, na.rm=T) But I never figured out how to pass the na.rm=T as an argument to median so that the function being bootstrapped is median(variable-name, na.rm=T). Q3: Is there some way in R to pass alternative arguments through a function to another? 4) Any general thoughts on Splus versus R that you are willing to share? This is way too much blithering for one day but thanks in advance for any thoughts. Bob Baskin All the usual disclaimers that my statements don't represent the agency etc. etc. -----Original Message----- From: Marc Schwartz [mailto:mschwartz@medanalytics.com] Sent: Thursday, June 05, 2003 1:19 PM To: 'Fohr, Marc [AM]'; 'R-help@lists.R-project.org' Subject: RE: [R] Introductory Resources -----Original Message----- From: r-help-bounces@stat.math.ethz.ch [mailto:r-help-bounces@stat.math.ethz.ch] On Behalf Of Fohr, Marc [AM] Sent: Thursday, June 05, 2003 11:46 AM To: 'R-help@lists.R-project.org' Subject: [R] (no subject) Hello, I am interested in R as an alternative for a statistical tool at our firm. I do know RATS an SPSS but not S+. As I read that R is close to S+, I would like to know if you could recommend me any books as an introduction to S+ or R. Best regards Marc Marc, Reviewing R FAQs 2.7 and 3.x on the main R site would be a good place to start. The former lists books and other documents (some online) that serve as excellent introductions, while the latter helps to differentiate R and S/S+. Spending some time with those references and the R FAQs will serve as a good foundation, with keyword searches of the R-help list archive serving as an additional strong resource. HTH, Another Marc ______________________________________________ R-help@stat.math.ethz.ch mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help [[alternate HTML version deleted]] ______________________________________________ R-help@stat.math.ethz.ch mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help ______________________________________________ R-help@stat.math.ethz.ch mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help