Re: [R] Exhaustive CHAID package
Many thanks for your response, sir. Here are two of the references to which I referred. I've also personally explored several data sets in which the outcomes are 'known' and have seen high variability in the topology of the trees being produced but, typically Exhaustive CHAID predictions match the 'known' results better than any of the others, using default settings. http://www.hindawi.com/journals/jam/2014/929768/ http://interstat.statjournals.net/YEAR/2010/articles/1007001.pdf By inference, many research papers are choosing Exhaustive CHAID. My concern is not that these procedures produce mildly variant trees but dramatically variant, with not even the same set of variables included. Is CHAID available for use as an R package? I thought R-FORGE was solely for developers? Again, many thanks. MCG -Original Message- From: Achim Zeileis [mailto:achim.zeil...@uibk.ac.at] Sent: Wednesday, April 22, 2015 3:30 AM To: Michael Grant Cc: r-help@R-project.org Subject: Re: [R] Exhaustive CHAID package On Tue, 21 Apr 2015, Michael Grant wrote: Dear R-Help: From multiple sources comparing methods of tree classification and tree regressions on various data sets, it seems that Exhaustive CHAID (distinct from CHAID), most commonly generates the most useful tree results and, in particular, is more effective than ctree or rpart which are implemented in R. I searched a bit on the web for exhaustive CHAID and didn't find any convincing evidence that this method is most commonly the most useful. I doubt that such evidence exists because the methods are applicable to so many different situations that uniformly better results are essentially never obtained. Nevertheless, if you have references of comparison studies, I would still be interested. Possibly these provide insight in which situations exhaustive CHAID performs particularly well. I see that CHAID, but not Exhaustive CHAID, is in the R-forge, and I write to ask if there are plans to create a package which employs the Exhaustive CHAID strategy. I wouldn't know of any such plans. But if you want to adapt/extend the code from the CHAID package, this is freely available. Right now the best source I can find is in SPSS-IBM and I feel a bit disloyal to R using it. I wouldn't be concerned about disloyalty. If you feel that exhaustive CHAID is the most appropriate tool for your problem and you have access to it in SPSS, why not use it? Possibly you can also export it from SPSS and import it into R using PMML. The partykit package has an example with an imported QUEST tree from SPSS. Michael Grant Professor University of Colorado Boulder [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Exhaustive CHAID package
Dear R-Help: From multiple sources comparing methods of tree classification and tree regressions on various data sets, it seems that Exhaustive CHAID (distinct from CHAID), most commonly generates the most useful tree results and, in particular, is more effective than ctree or rpart which are implemented in R. I see that CHAID, but not Exhaustive CHAID, is in the R-forge, and I write to ask if there are plans to create a package which employs the Exhaustive CHAID strategy. Right now the best source I can find is in SPSS-IBM and I feel a bit disloyal to R using it. Michael Grant Professor University of Colorado Boulder [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] lmer, p-values and all that
Dear Help: I am trying to follow Professor Bates' recommendation, quoted by Professor Crawley in The R Book, p629, to determine whether I should model data using the 'plain old' lm function or the mixed model function lmer by using the syntax anova(lmModel,lmerModel). Apparently I've not understood the recommendation or the proper likelihood ratio test in question (or both) for I get this error message: Error: $ operator not defined for this S4 class. Would someone be kind enough to point out my blunder? Thank you, Michael Grant [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Newly installed version; can't run lm function
New installation seems to have behavior I cannot figure out. Here is illustrative sequence where I load a small data set (test) from Crawley's files and try to run a simple linear model and get an error message. Oddly, R reports that the variable 'test$ozone' is numeric while, after attaching test, the variable ozone is not numeric. Can someone please help? This behavior is occurring with multiple data sets loaded from outside R. Thank you in advance. Michael Grant Example: test ozone garden 1 3 A 2 5 B 3 4 A 4 5 B 5 4 A 6 6 B 7 3 A 8 7 B 9 2 A 10 4 B 11 3 A 12 4 B 13 1 A 14 3 B 15 3 A 16 5 B 17 5 A 18 6 B 19 2 A 20 5 B is.data.frame(test) [1] TRUE is.numeric(test$ozone) [1] TRUE is.factor(test$garden) [1] TRUE lm(ozone~garden) Error in model.frame.default(formula = ozone ~ garden, drop.unused.levels = TRUE) : invalid type (list) for variable 'ozone' attach(test) is.numeric(ozone) [1] FALSE is.numeric(test$ozone) [1] TRUE [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Sub-Directories
I've searched hard in texts, email threads, FAQs etc. and cannot find out how to successfully utilize sub-directories below my WorkingDirectory. I can create them, and create R objects within the sub-directories but (a) the objects() command lists ONLY the objects in the WorkingDirectory and none of those in the sub-directory although I can use them (if I remember their names!). I'm an old (and I do mean old!) UNIX user now working in a Windows environment. I'm sure there is a simple way to do this but I seem unable to discover it. I'd like to have several different 'projects' going at the same time, with each being unrecognized by the other within R. Many thanks, Michael Grant [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Testing non-exhaustive Null and Alternative Hypothesis
I wish to test the hypothesis of mu equal to or less than 5 against the specific alternative mu equal to or greater than 7. I am unable to find how to persuade R to do this with any function (e.g. t.test). Suggestions? Michael Grant [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] choose.dir() gone?
Maybe 'tk_choose.dir' in the tcltk package will do what you want. --- On Wed, 9/8/10, Johannes Graumann johannes_graum...@web.de wrote: From: Johannes Graumann johannes_graum...@web.de Subject: Re: [R] choose.dir() gone? To: r-h...@stat.math.ethz.ch Date: Wednesday, September 8, 2010, 3:18 AM OK. Just checked and choose.file/choose.dir exists in the windows version - apparently not in the linux one ... does anybody have a nice platform-agnostic solution for this? Thanks, Joh Johannes Graumann wrote: Hi, I fail to find choose.dir() in my current R install (see below)? Didn't that exist at some point? How to achieve file.choose() equivalent functionality for directories? Thanks for any hints, Joh sessionInfo() R version 2.11.1 (2010-05-31) x86_64-pc-linux-gnu locale: [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C [3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8 [7] LC_PAPER=en_US.UTF-8 LC_NAME=en_US.UTF-8 [9] LC_ADDRESS=en_US.UTF-8 LC_TELEPHONE=en_US.UTF-8 [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=en_US.UTF-8 attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] rkward_0.5.3 loaded via a namespace (and not attached): [1] tools_2.11.1 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] annotating a filled contours plot with a grid of points
Dario, Checkout the 'plot.axes' argument to 'filled.contour'. Michael Grant --- On Fri, 2/6/09, Dario Martin-Benito dmar...@inia.es wrote: From: Dario Martin-Benito dmar...@inia.es Subject: [R] annotating a filled contours plot with a grid of points To: r-help@r-project.org Date: Friday, February 6, 2009, 8:23 AM Dear R-help members, I am trying to plot annotate a filled contours plot (with filled.contour) with a grid of points. I have read ways of annotating it with individual points but not with grids in another matrix. Any ideas? Thank you very much. Dario ___ Dario Martin-Benito CIFOR-INIA Dpto. Sistemas y Recursos Forestales Ctra. La Coruña, Km. 7.5 E-28040 Madrid (Spain ) Tel.: +34 91 347 1461 e-mail: dmar...@inia.es /// dmartinben...@gmail.com http://dmartinbenito.googlepages.com [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] Incorrect p value for binom.test?
I believe the binom.test procedure is producing one tailed p values rather than the two tailed value implied by the alternative hypothesis language. A textbook and SAS both show 2*9.94e-07 = 1.988e-06 as the two tailed value. As does the R summation syntax from R below. It looks to me like the alternative hypothesis language should be revised to something like ... greater than or equal to ... Am I mistaken? M.C.Grant 2*sum(dbinom(c(10:25),25,0.061)) [1] 1.987976e-06 binom.test(10,25,0.061) Exact binomial test data: 10 and 25 number of successes = 10, number of trials = 25, p-value = 9.94e-07 alternative hypothesis: true probability of success is not equal to 0.061 95 percent confidence interval: 0.2112548 0.6133465 sample estimates: probability of success 0.4 [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] 3D trajectory plot?
Are you talking about a 'streamline' like might be made with 'vtk'? or in a number of fluid dynamics, groundwater packages, particle physics codes, etc.? There was an exploratory hook up of R and vtk by M. Kondrin a year or two back (linux only). Search the archives. Parsing a vtk dataset in R should relatively straight-forward. HTH Michael Grant --- On Tue, 11/11/08, Chris Jarman [EMAIL PROTECTED] wrote: From: Chris Jarman [EMAIL PROTECTED] Subject: [R] 3D trajectory plot? To: r-help@r-project.org Date: Tuesday, November 11, 2008, 9:55 PM Hello, I'm attempting to create a smooth, 3D plot of a trajectory (rather than the cloud or wireframe functions). I would rather the individual data points not be visible. I've had no luck finding this on the graphics or help pages. Thank you in advance. Chris Some example data, just in case: 2532 40 1225 32 2 12 25 2 2 12 202 2 1 20 2 6 1 20 5 6 1 5 5 6 __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] Bug in is ?
Tie a rock to the 7 and tie another rock of equal mass to the 7.0. Throw them both into opposite ends of a large pond on the first Tuesday after after a new moon. If the former number floats it is not the integer 10. If the latter floats it likely is not the larger integer 13. This is because of the greater buoyancy of the '0'. Divide the difference (13-10) into 1 and you have 1/3, but you do not have 7 or even 7.0. Since this result is neither '7' nor '7.0', it absolutely must be a '7.'. This is why pi is irrational. Now, I hope that settles everything. It is easily demonstrated in four and a half lines of R code or one line of APL. HTH :O) Michael Grant --- On Mon, 9/29/08, Lucke, Joseph F [EMAIL PROTECTED] wrote: From: Lucke, Joseph F [EMAIL PROTECTED] Subject: Re: [R] Bug in is ? To: Stefan Evert [EMAIL PROTECTED], R-help Mailing List r-help@r-project.org Date: Monday, September 29, 2008, 10:24 AM Stefan You are right. Briefly put, the existence of 7 requires only Peano's axiom for successive integers. Strictly speaking, 7 is not an integer but a natural number. But natural numbers can be embedded in the integers which can be embedded in the rationals which can be embedded the reals which can be embedded in the complex. Little of this is relevant to a programming language's two basic storage modes for numbers. Confusing a variable type with a mathematical set is an elementary, if entertaining, logical error. Joe -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Stefan Evert Sent: Saturday, September 27, 2008 12:27 PM To: R-help Mailing List Subject: Re: [R] Bug in is ? Hi everyone! Sorry for taking an already overlong discussion thread slightly off- topic, but ... quote: No doubt, 7.0 is integer in math. But if people can write 7 why people need to write 7.0 (I do not see any reason to do this). endquote What is true in mathematics is not necessarily true in R. ... am I the only one who thinks that the integer 7 is something entirely different from the real number 7.0? (The latter most likely being an equivalence class of sequences of rational numbers, but that depends on your axiomatisation of real numbers.) Integers can be embedded in the set of real numbers, but that doesn't make them the same mathematically. So the original complaint should really have said that is.integer() doesn't do what a naive user (i.e. someone without a background in computer science or maths) might expect it to do. :-) That said, I've fallen into the same trap more than once, and painfully, myself (often in connection with MySQL). Best wishes, and with a tiny grain of salt, Stefan Evert [ [EMAIL PROTECTED] | http://purl.org/stefan.evert ] [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] model II regression - how do I do it?
Dear Danilo: Here is one approach with the formal reference being: Computational Statistics Data Analysis 23 ( 1997 ) 355-372 COMPUTATIONAL STATISTICS DATA ANALYSIS Generalization of the geometric mean functional relationship Norman R. Draper, Yonghong (Fred) Yang Department of Statistics, 1210 West Dayton Street, Madison, WI 53706-1685, USA Received February 1995; revised February 1996 Here is the S version of their program (for two predictor variables): Appendix The Splus code has been used to specify the weight functions and fit the model: Specify the weight function: weight - function(y,x 1,x2,b0,b 1,b2) { pred -b0+b l*x 1 ÷b2*x2 parms -abs(b 1b2)^(1/3) (y-pred)/parms } Fit the model gmfit -nls(~weight(y,x 1,x2,b0,b 1,b2), observe,list(starting value)) Hope this helps. MCG -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Danilo Muniz Sent: Friday, August 29, 2008 11:37 AM To: r-help@r-project.org Subject: [R] model II regression - how do I do it? I need to do a model II linear regression, but I could not find out how!! I tryed to use the lm function, but I did not discovered how to specify the model (type I or type II) to the function... could you help me? -- Danilo Muniz [Gruingas Abdiel] [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Re: [R] piper diagram
Sorry no previous message text or addresses, but I just cleaned my mailbox and then found something relevant. Regarding the Piper diagram. I just noticed the 'hydrogeo' package on CRAN, courtesy of one Myles English. That should be what you need or close to it. Best regards, Michael Grant [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.