Re: [R] Rcmdr X11 protocol error message
Michael Bibo [EMAIL PROTECTED] writes: What's happening? Both Java and R are reading events from the X11 event loop ``at the same time''. As a result, R is reading events meant for Java and vice-versa and bad things happen! There are two solutions: * use the Java graphics device which uses the Java graphics facilities to do the rendering; * don't run the GUI and the X11 device at the same time. The second solution is clearly not ideal. What we can do is to compile the X11 device code to be thread-safe. However, the same effect will be seen if we load another X11 event handler (e.g. the Tcl/Tk library). I don't think that is the issue. The X11 device and Tk sit on different connections to the X device, so they shouldn't be seeing eachother's events (the event loops can starve eachother though since they are multiplexed). I'd rather suspect a race condition with the Tk code itself - sending events to a window that is in the process of being destroyed and things like that. Which Tk version are we talking about, by the way? -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] D'agostino test
Alexandre, I found once a code to calculate Shapiro_Francia and D'Agostino tests of normality written long time ago by Peter B. Mandeville. Below his code and references (It is in spanish, I assume): Gregor #-- I coded the Shapiro-Francia test which handles from 5 to 5000 observations from Patrick Royston 1993 A Pocket-Calculator Algorithm for the Shapiro-Francia Test for Non-Normality: An Application to Medicine. Statistics in Medicine vol 12, 181-184 1991 Estimating Departure from Normality. Statistics in Medicine. vol 10, 1283-1293 1983 A Simple Method for Evaluating the Shapiro-Francia W' Test of Non-Normality. The Statistician 32 (1983) 297-300 and the D'Agostino tests from Ralph B. D'Agostino, Albert Belanger, and Ralph B. D'Agostino, Jr. 1990 A Suggestion for Using Powerful and Informative Tests of Normality. The American Statistician, November 1990, Vol. 44, No. 4 for testing normality in R. # Lee 1996:33-34 MOMENTS - function(data,r) sum((data-mean(data))^r)/length(data) # Peter SKEW - function(data) MOMENTS(data,3)/(MOMENTS(data,2)*sqrt(MOMENTS(data,2))) # Peter KURTOSIS - function(data) MOMENTS(data,4)/(MOMENTS(data,2)*MOMENTS(data,2)) # D'Agostino, Belanger and D'Agostino 1990:316-321 DAGOSTINO - function(data){ cat( PRUEBAS DE D'AGOSTINO\n) n - length(data) cat(SIMETRIA\n) cat( COEFICIENTE DE SIMETRIA:,sqrtb1 - SKEW(data),\n) if(n8){ y - sqrtb1*sqrt((n+1)*(n+3)/(6*(n-2))) beta2 - 3*(n*n+27*n-70)*(n+1)*(n+3)/((n-2)*(n+5)*(n+7)*(n+9)) w - sqrt(-1+sqrt(2*(beta2-1))) delta - 1/sqrt(log(w)) ALPHA - sqrt(2/(w*w-1)) cat( ZETA CALCULADA:,zb1 - delta*log(y/ALPHA+sqrt((y/ALPHA)^2+1)),\n) cat( PROBABILIDAD:,2*(1-pnorm(abs(zb1))),\n) }else cat(LA PRUEBA DE SESGO REQUIERE POR LO MENOS 9 REPETICIONES\n) cat(KURTOSIS\n) cat( COEFICIENTE DE KURTOSIS:,b2 - KURTOSIS(data),\n) if(n19){ meanb2 - 3*(n-1)/(n+1) varb2 - 24*n*(n-2)*(n-3)/((n+1)*(n+1)*(n+3)*(n+5)) x - (b2-meanb2)/sqrt(varb2) moment - 6*(n*n-5*n+2)/((n+7)*(n+9))*sqrt(6*(n+3)*(n+5)/(n*(n-2)*(n-3))) a - 5+8/moment*(2/moment+sqrt(1+4/(moment*moment))) cat( ZETA CALCULADA:,zb2 - (1-2/(9*a)-((1-2/a)/(1+x*sqrt(2/(a-4^(1/3))/sqrt(2/(9*a)),\n) cat( PROBABILIDAD:,2*(1-pnorm(abs(zb2))),\n) cat(OMNIBUS\n) cat( JI-CUADRADA CALCULADA:,k2 - zb1*zb1+zb2*zb2,\n) cat( GRADOS DE LIBERTAD: 2\n) cat( PROBABILIDAD:,probji2 - 1-pchisq(k2,2),\n) }else cat(LAS PRUEBAS DE KURTOSIS Y OMNIBUS REQUIEREN POR LO MENOS 20 REPETICIONES\n) } # Royston 1993:183-184 SHAPIRO.FRANCIA - function(data){ cat( PRUEBA DE NORMALIDAD DE SHAPIRO-FRANCIA\n) n - length(data) if(n5 || n5000) cat(REQUIERE ENTRE 5 Y 5000 REPETICIONES\n) else{ xbar - mean(data) sdata - sort(data) resid - sdata-xbar uniform - seq(1,n) np - qnorm((uniform-0.375)/(n+0.25)) cat(W':,w - (sum(np*sdata))^2/(sum(np*np)*sum(resid*resid)),\n) u - log(n) v - log(u) muy - -1.2725+1.0521*(v-u) sigmay - 1.0308-0.26758*(v+2/u) cat(ZETA CALCULADA:,zeta - (log(1-w)-muy)/sigmay,\n) cat(PROBABILIDAD:,probz - 1-pnorm(zeta),\n) } } Peter B. Peter B. Mandeville [EMAIL PROTECTED] Jefe del Depto. de Informática y Bioestadística [EMAIL PROTECTED] Facultad de MedicineTel: 48 26-23-45 ext. 232 Universidad Autónoma de San Luis Potosí Fax: 48 28-23-52 Av. V. Carranza 2405 Col. Los Filtros Apartado Postal 145 San Luis Potosí, S.L.P. 78210 México #-- -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Alexandre Bournery Sent: Montag, 30. August 2004 18:08 To: [EMAIL PROTECTED] Subject: [R] D'agostino test Hi, Does anyone know if the D'agostino test is available with R ? Alex __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Any information in this communication is confidential and ma...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] (no subject)
Hi this is a very simple question ! if a have a set of date like x=c(1,2,3,4,4,5,6,8,7,8,8) how can I find out the percentile of 7 ? thanks Paolo __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] (no subject)
sum(x=7)/length(x)*100 *** REPLY SEPARATOR *** On 8/31/2004 at 8:16 AM Paolo Tommasini wrote: Hi this is a very simple question ! if a have a set of date like x=c(1,2,3,4,4,5,6,8,7,8,8) how can I find out the percentile of 7 ? thanks Paolo __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Dipl. bio-chem. Witold Eryk Wolski @ MPI-Moleculare Genetic Ihnestrasse 63-73 14195 Berlin'v' tel: 0049-30-83875219/ \ mail: [EMAIL PROTECTED]---W-Whttp://www.molgen.mpg.de/~wolski [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] Rcmdr X11 protocol error message
Dear Michael, A question: Do you observe these problems with tcltk in general or just with the Rcmdr package? Is it possible for you to test a Tcl/Tk program outside of R? Regards, John -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Peter Dalgaard Sent: Tuesday, August 31, 2004 2:32 AM To: Michael Bibo Cc: [EMAIL PROTECTED] Subject: Re: [R] Rcmdr X11 protocol error message Michael Bibo [EMAIL PROTECTED] writes: What's happening? Both Java and R are reading events from the X11 event loop ``at the same time''. As a result, R is reading events meant for Java and vice-versa and bad things happen! There are two solutions: * use the Java graphics device which uses the Java graphics facilities to do the rendering; * don't run the GUI and the X11 device at the same time. The second solution is clearly not ideal. What we can do is to compile the X11 device code to be thread-safe. However, the same effect will be seen if we load another X11 event handler (e.g. the Tcl/Tk library). I don't think that is the issue. The X11 device and Tk sit on different connections to the X device, so they shouldn't be seeing eachother's events (the event loops can starve eachother though since they are multiplexed). I'd rather suspect a race condition with the Tk code itself - sending events to a window that is in the process of being destroyed and things like that. Which Tk version are we talking about, by the way? -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] appending data to a dataframe
Dear R users, I am sorry to ask you such a pathetic newbie question, but how does one append data at the end of a data frame? I am working with GRASS/R library, but the question is about R. I have a data.frame containing the following variables basinID, distoutlet, drainage_area, slope These variables are stored for all pixels of Grass Raster objects. For each drainage basin (basinID), I'd like to find the maximum of distoutlet. How can I store the pair of info (bv$ID, max(bv$distoutlet)) at each step of the for loop ? I presume something like this could do for (i 1:max(basinID)){ bid - i length - max(distoutlet[which(basinID == I(i))]) } But how do I handle the output and record bid and length in parallel in an object??? Any hint is welcome Thomas *** Le contenu de cet e-mail et de ses pièces jointes est destin...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] appending data to a dataframe
On Tue, 31 Aug 2004 14:05:59 +0200, Dewez Thomas [EMAIL PROTECTED] wrote : Dear R users, I am sorry to ask you such a pathetic newbie question, but how does one append data at the end of a data frame? The rbind() function should work: it adds rows to matrices and dataframes. However, I'd do something different for the problem you give: I am working with GRASS/R library, but the question is about R. I have a data.frame containing the following variables basinID, distoutlet, drainage_area, slope These variables are stored for all pixels of Grass Raster objects. For each drainage basin (basinID), I'd like to find the maximum of distoutlet. How can I store the pair of info (bv$ID, max(bv$distoutlet)) at each step of the for loop ? I presume something like this could do for (i 1:max(basinID)){ bid - i length - max(distoutlet[which(basinID == I(i))]) } But how do I handle the output and record bid and length in parallel in an object??? Appending a row at a time is very slow. You'll find it much faster to set up a matrix to hold the results in advance: results - matrix(NA, max(basinID), 2) for (i in 1:max(basinID)) { bid - i length - max(distoutlet[which(basinID == i)]) results[i,] - c(i, length) } Duncan Murdoch __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] blockwise sums
I am looking for a function like my.blockwisesum(vector, n) that computes sums of disjoint subsequences of length n from vector and can work with vector lengths that are not a multiple of n. It should give me for instance my.blockwisesum(1:10, 3) == c(6, 15, 24, 10) Is there a builtin function that can do this? One could do it by coercing the vector into a matrix of width n, and then use apply, but that is cumbersome if the length is not divisible by n, is it not? Any other ideas? Lutz Prof. Dr. Lutz Prechelt; [EMAIL PROTECTED] Institut fuer Informatik; Freie Universitaet Berlin Takustr. 9; 14195 Berlin; Germany +49 30 838 75115; http://www.inf.fu-berlin.de/inst/ag-se/ __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] blockwise sums
If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } Andy From: Lutz Prechelt I am looking for a function like my.blockwisesum(vector, n) that computes sums of disjoint subsequences of length n from vector and can work with vector lengths that are not a multiple of n. It should give me for instance my.blockwisesum(1:10, 3) == c(6, 15, 24, 10) Is there a builtin function that can do this? One could do it by coercing the vector into a matrix of width n, and then use apply, but that is cumbersome if the length is not divisible by n, is it not? Any other ideas? Lutz Prof. Dr. Lutz Prechelt; [EMAIL PROTECTED] Institut fuer Informatik; Freie Universitaet Berlin Takustr. 9; 14195 Berlin; Germany +49 30 838 75115; http://www.inf.fu-berlin.de/inst/ag-se/ __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] blockwise sums
Liaw, Andy wrote: If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } Did you test that? I get: my.blockwisesum(1:10, 3) Error in tapply(x, seq(1, length(x), by = n), sum, ...) : arguments must have same length Here's my solution with tapply and rep() to generate a vector like c(1,1,1,2,2,2,3,3,3,4): baz.blockwisesum= function(v,n){tapply(v,rep(1:(1+length(v)/n),each=n)[1:length(v)],sum)} baz.blockwisesum(1:10,3) 1 2 3 4 6 15 24 10 - just ignore the 1 to 4 names, they cant hurt you. Baz __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] blockwise sums
Hi! ind-c(sort(rep(1:floor(length(x)/ 3 ) , 3 )) , floor(length(x)/ 3 )+1) by(x,ind,sum) my.blockwisesum-function(x,n,...) { ind-c(sort(rep(1:floor(length(x)/ n ) , n )) , floor(length(x)/ n )+1) return(tapply(x,ind,sum)) } /Eryk *** REPLY SEPARATOR *** On 8/31/2004 at 2:19 PM Lutz Prechelt wrote: I am looking for a function like my.blockwisesum(vector, n) that computes sums of disjoint subsequences of length n from vector and can work with vector lengths that are not a multiple of n. It should give me for instance my.blockwisesum(1:10, 3) == c(6, 15, 24, 10) Is there a builtin function that can do this? One could do it by coercing the vector into a matrix of width n, and then use apply, but that is cumbersome if the length is not divisible by n, is it not? Any other ideas? Lutz Prof. Dr. Lutz Prechelt; [EMAIL PROTECTED] Institut fuer Informatik; Freie Universitaet Berlin Takustr. 9; 14195 Berlin; Germany +49 30 838 75115; http://www.inf.fu-berlin.de/inst/ag-se/ __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Dipl. bio-chem. Witold Eryk Wolski @ MPI-Moleculare Genetic Ihnestrasse 63-73 14195 Berlin'v' tel: 0049-30-83875219/ \ mail: [EMAIL PROTECTED]---W-Whttp://www.molgen.mpg.de/~wolski [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] blockwise sums
From: Barry Rowlingson Liaw, Andy wrote: If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } Did you test that? I get: Of course not (slap on wrist)!! My apologies... Andy my.blockwisesum(1:10, 3) Error in tapply(x, seq(1, length(x), by = n), sum, ...) : arguments must have same length Here's my solution with tapply and rep() to generate a vector like c(1,1,1,2,2,2,3,3,3,4): baz.blockwisesum= function(v,n){tapply(v,rep(1:(1+length(v)/n),each=n)[1:length( v)],sum)} baz.blockwisesum(1:10,3) 1 2 3 4 6 15 24 10 - just ignore the 1 to 4 names, they cant hurt you. Baz __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] blockwise sums
Hi Lutz, you could try the following: blockwisesum - function(x, n){ nx - length(x) if(nx%%n) x. - c(x, rep(0., n*ceiling(nx/n)-nx)) else x. - x x. - matrix(x., ncol=n, byrow=TRUE) rowSums(x.) } blockwisesum(1:10, 3) I hope this helps. Best, Dimitris Dimitris Rizopoulos Doctoral Student Biostatistical Centre School of Public Health Catholic University of Leuven Address: Kapucijnenvoer 35, Leuven, Belgium Tel: +32/16/396887 Fax: +32/16/337015 Web: http://www.med.kuleuven.ac.be/biostat/ http://www.student.kuleuven.ac.be/~m0390867/dimitris.htm - Original Message - From: Lutz Prechelt [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, August 31, 2004 2:19 PM Subject: [R] blockwise sums I am looking for a function like my.blockwisesum(vector, n) that computes sums of disjoint subsequences of length n from vector and can work with vector lengths that are not a multiple of n. It should give me for instance my.blockwisesum(1:10, 3) == c(6, 15, 24, 10) Is there a builtin function that can do this? One could do it by coercing the vector into a matrix of width n, and then use apply, but that is cumbersome if the length is not divisible by n, is it not? Any other ideas? Lutz Prof. Dr. Lutz Prechelt; [EMAIL PROTECTED] Institut fuer Informatik; Freie Universitaet Berlin Takustr. 9; 14195 Berlin; Germany +49 30 838 75115; http://www.inf.fu-berlin.de/inst/ag-se/ __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] blockwise sums
Liaw, Andy wrote: If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } Did you test that? I get: my.blockwisesum(1:10, 3) Error in tapply(x, seq(1, length(x), by = n), sum, ...) : arguments must have same length Here's my solution with tapply and rep() to generate a vector like c(1,1,1,2,2,2,3,3,3,4): baz.blockwisesum= function(v,n){tapply(v,rep(1:(1+length(v)/n),each=n)[1:length( v)],sum)} baz.blockwisesum(1:10,3) 1 2 3 4 6 15 24 10 - just ignore the 1 to 4 names, they cant hurt you. Baz To complete the picture: here is another one: my.blockwisesum - function(vec, n){ vec - as.vector(vec) n - as.integer(n) total - length(vec) if(total = n){ stop(\nn should be smaller than length of vector.\n) } start - seq(1, total, n) end - start + n - 1 end[end total] - max(start) index - 1 : length(start) return(sapply(index, function(x)sum(test[start[x]:end[x]]))) } test - 1:150 ptn - proc.time() baz.blockwisesum(test,3) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 6 15 24 33 42 51 60 69 78 87 96 105 114 123 132 141 150 159 168 177 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 186 195 204 213 222 231 240 249 258 267 276 285 294 303 312 321 330 339 348 357 41 42 43 44 45 46 47 48 49 50 366 375 384 393 402 411 420 429 438 447 proc.time()-ptn [1] 0.00 0.00 0.22 NA NA ptn - proc.time() my.blockwisesum(test,3) [1] 6 15 24 33 42 51 60 69 78 87 96 105 114 123 132 141 150 159 168 [20] 177 186 195 204 213 222 231 240 249 258 267 276 285 294 303 312 321 330 339 [39] 348 357 366 375 384 393 402 411 420 429 438 447 proc.time()-ptn [1] 0.00 0.00 0.19 NA NA HTH, Bernhard The information contained herein is confidential and is inte...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] blockwise sums
Barry Rowlingson [EMAIL PROTECTED] writes: Liaw, Andy wrote: If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } Did you test that? I get: my.blockwisesum(1:10, 3) Error in tapply(x, seq(1, length(x), by = n), sum, ...) : arguments must have same length Here's my solution with tapply and rep() to generate a vector like c(1,1,1,2,2,2,3,3,3,4): baz.blockwisesum= function(v,n){tapply(v,rep(1:(1+length(v)/n),each=n)[1:length(v)],sum)} baz.blockwisesum(1:10,3) 1 2 3 4 6 15 24 10 Slight variant pd.blockwisesum - function(v,n){N - length(v); tapply(x,gl(ceiling(N/n), n, N), sum)} pd.blockwisesum(1:10,3) 1 2 3 4 6 15 24 10 -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] blockwise sums
Lutz Prechelt prechelt at pcpool.mi.fu-berlin.de writes: : : I am looking for a function like : my.blockwisesum(vector, n) : that computes sums of disjoint subsequences of length n from vector : and can work with vector lengths that are not a multiple of n. : : It should give me for instance : my.blockwisesum(1:10, 3) == c(6, 15, 24, 10) tapply(v, (seq(v)-1)%/%n, sum) __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] I've forgotten, why is box() the default?
I've searched on CRAN for axes, axis, and other terms I've already forgotten, without (re)discovering the reason for S using non-joining axes by default, instead of box(l). MASS points me towards Cleveland (1993) but I don't have ready access to this any more. Could someone give me a one-liner to justify this choice to a sceptic? It's something to do with not mis-interpreting the axes intersection as (0,0), isn't it? TIA, Simon Simon Fear Senior Statistician Syne qua non Ltd Tel: +44 (0) 1379 69 Fax: +44 (0) 1379 65 email: [EMAIL PROTECTED] web: http://www.synequanon.com Number of attachments included with this message: 0 This message (and any associated files) is confidential and\...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] D'agostino test
-Original Message- From: Alexandre Bournery [mailto:[EMAIL PROTECTED] Sent: Monday, August 30, 2004 11:08 AM To: [EMAIL PROTECTED] Subject: [R] D'agostino test Hi, Does anyone know if the D'agostino test is available with R ? Alex Hi, See below, taken pretty much verbatim from Zar (1999) Douglas G. Scofield, PhDDepartment of Biology [EMAIL PROTECTED]Indiana University off: (812) 856-0115 1001 E. 3rd St. fax: (812) 855-6705 Bloomington, IN 47405-3700 cell: (786) 514-9141 --- dagostino.pearson.test - function(x) { # from Zar (1999), implemented by Doug Scofield, [EMAIL PROTECTED] DNAME - deparse(substitute(x)) n - length(x) x2 - x * x x3 - x * x2 x4 - x * x3 # compute Z_g1 k3 - ((n*sum(x3)) - (3*sum(x)*sum(x2)) + (2*(sum(x)^3)/n)) / ((n-1)*(n-2)) g1 - k3 / sqrt(var(x)^3) sqrtb1 - ((n - 2)*g1) / sqrt(n*(n - 1)) A - sqrtb1 * sqrt(((n + 1)*(n + 3)) / (6*(n - 2))) B - (3*(n*n + 27*n - 70)*(n+1)*(n+3)) / ((n-2)*(n+5)*(n+7)*(n+9)) C - sqrt(2*(B - 1)) - 1 D - sqrt(C) E - 1 / sqrt(log(D)) F - A / sqrt(2/(C - 1)) Zg1 - E * log(F + sqrt(F*F + 1)) # compute Z_g2 G - (24*n*(n-2)*(n-3)) / (((n+1)^2)*(n+3)*(n+5)) k4 - (((n*n*n + n*n)*sum(x4)) - (4*(n*n + n)*sum(x3)*sum(x)) - (3*(n*n - n)*sum(x2)^2) + (12*n*sum(x2)*sum(x)^2) - (6*sum(x)^4)) / (n*(n-1)*(n-2)*(n-3)) g2 - k4 / var(x)^2 H - ((n-2)*(n-3)*abs(g2)) / ((n+1)*(n-1)*sqrt(G)) J - ((6*(n*n - 5*n + 2)) / ((n+7)*(n+9))) * sqrt((6*(n+3)*(n+5)) / (n*(n-2)*(n-3))) K - 6 + (8/J)*(2/J + sqrt(1 + 4/(J*J))) L - (1 - 2/K) / (1 + H*sqrt(2/(K-4))) Zg2 - (1 - 2/(9*K) - (L^(1/3))) / (sqrt(2/(9*K))) K2 - Zg1*Zg1 + Zg2*Zg2 pk2 - pchisq(K2, 2, lower.tail=FALSE) RVAL - list(statistic = c(K2 = K2), p.value = pk2, method = D'Agostino-Pearson normality test\n\nK2 is distributed as Chi-squared with df=2, alternative = distribution is not normal, data.name = DNAME) class(RVAL) - htest return(RVAL) } __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] (no subject)
On Mon, 30 Aug 2004, Austin, Matt wrote: A correction. You either need to open the plotting device prior to the simulation that includes your plotting commands and close it after the simulation or have the name change dynamically in your simulation so that the runs go in separate files. An example of the first method would be postscript(file=/where/to/put/file/filename.ps) It's also worth pointing out that many people will find pdf() more convenient than postscript() -thomas ## your simulation commands dev.off() for the second method more info would be needed to know how you are running your simulation to help with dynamic naming. --Matt -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Austin, Matt Sent: Monday, August 30, 2004 18:56 PM To: 'Loke Chok Kang'; [EMAIL PROTECTED] Subject: RE: [R] (no subject) You need to use a device to print to such as postscript(file=/where/to/put/file/filename.ps) ##your plotting code here dev.off() Writing a plot at each iteration of your simulation can impact the runtime greatly. Please read the posting guide at the bottom of the e-mail, it can help you get more helpful . . . help --Matt -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Loke Chok Kang Sent: Monday, August 30, 2004 18:42 PM To: [EMAIL PROTECTED] Subject: [R] (no subject) Hi, Sorry, I have a problem that require some help. As I am doing a project with R and this project requires me to do a lot of plotting as I run my simulation, I need R to help me store my plots automatically as the simulation is run. Could anyone advise me on how this can be done? If possible I need all the plots to be written to a file. PS:In all I have over 500 over plots, thus it will not be possible to manually plot all the plots out. From yours faithfully, Chok Kang [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Thomas Lumley Assoc. Professor, Biostatistics [EMAIL PROTECTED] University of Washington, Seattle __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] blockwise sums
On Tue, 31 Aug 2004, Liaw, Andy wrote: If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } If x is large, rowsum() should be faster than tapply(). -thomas Andy From: Lutz Prechelt I am looking for a function like my.blockwisesum(vector, n) that computes sums of disjoint subsequences of length n from vector and can work with vector lengths that are not a multiple of n. It should give me for instance my.blockwisesum(1:10, 3) == c(6, 15, 24, 10) Is there a builtin function that can do this? One could do it by coercing the vector into a matrix of width n, and then use apply, but that is cumbersome if the length is not divisible by n, is it not? Any other ideas? Lutz Prof. Dr. Lutz Prechelt; [EMAIL PROTECTED] Institut fuer Informatik; Freie Universitaet Berlin Takustr. 9; 14195 Berlin; Germany +49 30 838 75115; http://www.inf.fu-berlin.de/inst/ag-se/ __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Thomas Lumley Assoc. Professor, Biostatistics [EMAIL PROTECTED] University of Washington, Seattle __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] I've forgotten, why is box() the default?
On Tue, 31 Aug 2004 14:13:49 +0100, Simon Fear [EMAIL PROTECTED] wrote : I've searched on CRAN for axes, axis, and other terms I've already forgotten, without (re)discovering the reason for S using non-joining axes by default, instead of box(l). MASS points me towards Cleveland (1993) but I don't have ready access to this any more. Could someone give me a one-liner to justify this choice to a sceptic? It's something to do with not mis-interpreting the axes intersection as (0,0), isn't it? I'm not sure I understand your question: the default in S is a full box unless you ask for no box (bty=n), isn't it? The justification for this is The four scale lines also provide a clearly defined region where our eyes can search for data. With just two, data can be camouflaged by virtue of where they lie (p. 35). There's also advice to keep the data away from the axes and to put the ticks outside the box to avoid hiding extreme points. I can't spot a discussion in Cleveland of the reason the axes don't join when bty=n is specified. Duncan Murdoch __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] I've forgotten, why is box() the default?
Not helped by my typo and being in a rush! It's true that plot.default and surely many others do a box, since par(bty) [1] o But this isn't what happens if you do things via axis() ... (see first example in help(axis)). Anyway, point taken, I couldn't find it because it is NOT what is recommended! -Original Message- From: Duncan Murdoch [mailto:[EMAIL PROTECTED] Sent: 31 August 2004 14:45 To: Simon Fear Cc: R-help Subject: Re: [R] I've forgotten, why is box() the default? Security Warning: If you are not sure an attachment is safe to open contact Andy on x234. There are 0 attachments with this message. On Tue, 31 Aug 2004 14:13:49 +0100, Simon Fear [EMAIL PROTECTED] wrote : I've searched on CRAN for axes, axis, and other terms I've already forgotten, without (re)discovering the reason for S using non-joining axes by default, instead of box(l). MASS points me towards Cleveland (1993) but I don't have ready access to this any more. Could someone give me a one-liner to justify this choice to a sceptic? It's something to do with not mis-interpreting the axes intersection as (0,0), isn't it? I'm not sure I understand your question: the default in S is a full box unless you ask for no box (bty=n), isn't it? The justification for this is The four scale lines also provide a clearly defined region where our eyes can search for data. With just two, data can be camouflaged by virtue of where they lie (p. 35). There's also advice to keep the data away from the axes and to put the ticks outside the box to avoid hiding extreme points. I can't spot a discussion in Cleveland of the reason the axes don't join when bty=n is specified. Duncan Murdoch Simon Fear Senior Statistician Syne qua non Ltd Tel: +44 (0) 1379 69 Fax: +44 (0) 1379 65 email: [EMAIL PROTECTED] web: http://www.synequanon.com Number of attachments included with this message: 0 This message (and any associated files) is confidential and\...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Restore function in R
R-help, Any function in R that restore an object to its original state so that if not satisfied with the modifications made to it there is no need to redefine again ( 'undo' action) I believe there is something like that in S-plus. Thank you __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] N-dimensional delaunay tesselation voronoi diagrams
Hi, I've been looking for functions that can do delaunay tesselation and generate voronoi cells. I came across deldir and tripack but both seem to be restricted to 2D points. Are there any packages that can do a tesselation in N dimensions? I know that Matlab and Mathematica use the qhull package to provide functions for this. Does anybody know of any R packages that do this (maybe by calling on qhull)? Thanks, --- Rajarshi Guha [EMAIL PROTECTED] http://jijo.cjb.net GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04 06F7 1BB9 E634 9B87 56EE --- Q: What do you get when you cross a mosquito with a mountain climber? A: Nothing. You can't cross a vector with a scaler. __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] More efficient matrix computation
I have a 20x3 matrix as follows: m - replicate(3, matrix(rnorm(20),20,1)) I need to compute, say, 95th and 99th percentiles of each column such that the resulting matrix becomes 2x3 with each row representing the respective percentile. My best effort is to compute one column at a time as follows: quantile(m[,1], c(0.95, 0.99)) To do the same for columns 2 and 3, I would simply change the column number accordingly. Clearly, this is not very efficient as I may have a large matrix (e.g., 100,000x500) to work with. Any help with the code is appreciated. Jack Wang __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] More efficient matrix computation
JTW wrote: I have a 20x3 matrix as follows: m - replicate(3, matrix(rnorm(20),20,1)) Why not m - matrix(rnorm(60), 20, 3) I need to compute, say, 95th and 99th percentiles of each column such that the resulting matrix becomes 2x3 with each row representing the respective percentile. My best effort is to compute one column at a time as follows: quantile(m[,1], c(0.95, 0.99)) That's what apply() is made for: apply(m, 2, quantile, c(0.95, 0.99)) Uwe Ligges To do the same for columns 2 and 3, I would simply change the column number accordingly. Clearly, this is not very efficient as I may have a large matrix (e.g., 100,000x500) to work with. Any help with the code is appreciated. Jack Wang __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] More efficient matrix computation
Hi! apply(m,2,quantile,c(0.95,0.99)) /Eryk *** REPLY SEPARATOR *** On 8/31/2004 at 9:16 AM JTW wrote: I have a 20x3 matrix as follows: m - replicate(3, matrix(rnorm(20),20,1)) I need to compute, say, 95th and 99th percentiles of each column such that the resulting matrix becomes 2x3 with each row representing the respective percentile. My best effort is to compute one column at a time as follows: quantile(m[,1], c(0.95, 0.99)) To do the same for columns 2 and 3, I would simply change the column number accordingly. Clearly, this is not very efficient as I may have a large matrix (e.g., 100,000x500) to work with. Any help with the code is appreciated. Jack Wang __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Dipl. bio-chem. Witold Eryk Wolski @ MPI-Moleculare Genetic Ihnestrasse 63-73 14195 Berlin'v' tel: 0049-30-83875219/ \ mail: [EMAIL PROTECTED]---W-Whttp://www.molgen.mpg.de/~wolski [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] blockwise sums
This is a cute problem, so how about a cute algorithm? Since all that is wanted is the sum, matrix multiplication can be used. If n=vector length and b = blocksize, then the issue comes down to constructing a block diagonal matrix with nblk=n%/%b columns with b 1's in each block to multiply the first b*nblk elements of the vector by (the sum of any remaining elements can be appended to the result). This muliplier matrix is easily constructed using diag(): matrix(rep(diag(nblk),each=b),ncol=nblk) There may even be slicker ways to do it. Anyway, as no implicit looping (tapply) is used, I suspect this would run considerably faster,too, although this may be unimportant, and I haven't tested it. Of course, this approach doesn't generalize. But it is a nice example of some of R's little tricks. Cheers, -- Bert Gunter Genentech Non-Clinical Statistics South San Francisco, CA The business of the statistician is to catalyze the scientific learning process. - George E. P. Box -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Liaw, Andy Sent: Tuesday, August 31, 2004 5:42 AM To: 'Barry Rowlingson' Cc: [EMAIL PROTECTED] Subject: RE: [R] blockwise sums From: Barry Rowlingson Liaw, Andy wrote: If you insist, here's one way: my.blockwisesum - function(x, n, ...) { tapply(x, seq(1, length(x), by=n), sum, ...) } Did you test that? I get: Of course not (slap on wrist)!! My apologies... Andy my.blockwisesum(1:10, 3) Error in tapply(x, seq(1, length(x), by = n), sum, ...) : arguments must have same length Here's my solution with tapply and rep() to generate a vector like c(1,1,1,2,2,2,3,3,3,4): baz.blockwisesum= function(v,n){tapply(v,rep(1:(1+length(v)/n),each=n)[1:length( v)],sum)} baz.blockwisesum(1:10,3) 1 2 3 4 6 15 24 10 - just ignore the 1 to 4 names, they cant hurt you. Baz __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] More efficient matrix computation
A million thanks to all for the prompt reply: apply() works! Jack Wang __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Rcmdr X11 protocol error message
John Fox jfox at mcmaster.ca writes: Dear Michael, A question: Do you observe these problems with tcltk in general or just with the Rcmdr package? Is it possible for you to test a Tcl/Tk program outside of R? And Peter asked which version of Tcl/Tk. Apparently my system has version 8.4 installed, specifically: Tcl-8.4.5-3-mdk (tclsh8.4) and Tk-8.4.5-3-mdk (libtk8.4.so). As I understand it, these are installed from RPMs on the installation DVD, and I note that they are mandrake specific. John - I wasn't sure if I had any other Tcl/Tk applications installed (it's not always obvious when installing from RPM's). I have certainly not encountered these error messages with any other application. I quickly downloaded WISH Supernotepad 1.2.3. This application requires Tcl/Tk 8.4 or greater. There are no graphics window in this application, but plenty of dialogue boxes. It gave no errors. Is this an appropriate test? If this is a mandrake-configuration-specific problem, it may not be worth spending too much time investigating, as R Commandr still works. I can always try re-installing Tcl/Tk from source when/if I have time. Regards, Michael __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] after lm-fit: equality of two regression coefficients
Dear Christoph, You can use the package gregmisc require(gregmisc) C-c(0,1,-1) glh.test(my.lm,C) I hope that helps Alexandre Galvão Hi Let's assume, we have a multiple linear regression, such as the one using the Scottish hills data (MASS, data(hills)): one dependent variable: time two independent var (metric): dist, climb if I am interested, after (!) fitting a lm: my. lm - lm(time ~ dist + climb, data = hills) in the equivalence (or non-equivalence) of the two predictors dist and climb: H0: dist = climb I think you intend to ask if the *coefficients* in the fit should be equal, which is nonsense in this example of course. Is there any function in R, which lets me calculate this, in just giving the lm-object my.lm and e.g. a vector such as c(1, -1), operationalizing the hypothesis H0: t(c(1, -1)) %*% c(dist, climb) = 0 ? library(car) ?linear.hypothesis Christoph, If you are interested in testing for b1=b2 in a regression model, say y=b0+b1*x1+b2*x2+e you can compare the two models o1 and o2 o1-lm(y~x1+x2) o2-lm(y~I(x1+x2)) best, vito __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] enter browser on error
Is there a way I can get R to automatically enter the browser inside a user-defined function on the generation of an error? Specifically, I'm trying to debug this: Error in as.double.default(sapply(lis, FUN)) : (list) object cannot be coerced to double In addition: There were 38 warnings (use warnings() to see them) traceback() 8: as.double.default(sapply(lis, FUN)) 7: as.numeric(sapply(lis, FUN)) 6: numeric.sapply(function(x) { [EMAIL PROTECTED] }) On detection of the error, I would like browser() to be called at the level of numeric.sapply(), so that I can examine x. I'm wondering if this can be done by modifying the default error handling. Using try() with browser() didn't work. Thanks, David _ David Bickel http://davidbickel.com Research Scientist Pioneer Hi-Bred International Bioinformatics Exploratory Research 7250 NW 62nd Ave., PO Box 552 Johnston, Iowa 50131-0552 515-334-4739 Tel 515-334-6634 Fax [EMAIL PROTECTED], [EMAIL PROTECTED] This communication is for use by the intended recipient and ...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] enter browser on error
Have you tried options(error=recover)? On Tue, 31 Aug 2004, Bickel, David wrote: Is there a way I can get R to automatically enter the browser inside a user-defined function on the generation of an error? Specifically, I'm trying to debug this: Error in as.double.default(sapply(lis, FUN)) : (list) object cannot be coerced to double In addition: There were 38 warnings (use warnings() to see them) traceback() 8: as.double.default(sapply(lis, FUN)) 7: as.numeric(sapply(lis, FUN)) 6: numeric.sapply(function(x) { [EMAIL PROTECTED] }) On detection of the error, I would like browser() to be called at the level of numeric.sapply(), so that I can examine x. I'm wondering if this can be done by modifying the default error handling. Using try() with browser() didn't work. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] enter browser on error
use options(error=recover), e.g.: remove(x) NULL Warning message: remove: variable x was not found (function() {x})() Error in (function() { : Object x not found options(error=recover) (function(y=1) {x})(2) Error in (function(y = 1) { : Object x not found Enter a frame number, or 0 to exit 1:(function(y = 1) { Selection: 1 Called from: eval(expr, envir, enclos) Browse[1] y [1] 2 Browse[1] Enter a frame number, or 0 to exit 1:(function(y = 1) { Selection: 0 At Tuesday 11:26 AM 8/31/2004, Bickel, David wrote: Is there a way I can get R to automatically enter the browser inside a user-defined function on the generation of an error? Specifically, I'm trying to debug this: Error in as.double.default(sapply(lis, FUN)) : (list) object cannot be coerced to double In addition: There were 38 warnings (use warnings() to see them) traceback() 8: as.double.default(sapply(lis, FUN)) 7: as.numeric(sapply(lis, FUN)) 6: numeric.sapply(function(x) { [EMAIL PROTECTED] }) On detection of the error, I would like browser() to be called at the level of numeric.sapply(), so that I can examine x. I'm wondering if this can be done by modifying the default error handling. Using try() with browser() didn't work. Thanks, David _ David Bickel http://davidbickel.com Research Scientist Pioneer Hi-Bred International Bioinformatics Exploratory Research 7250 NW 62nd Ave., PO Box 552 Johnston, Iowa 50131-0552 515-334-4739 Tel 515-334-6634 Fax [EMAIL PROTECTED], [EMAIL PROTECTED] This communication is for use by the intended recipient and ...{{dropped}} __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] sample size for t-tests
Dear all, Could any one please tell me the exact formula R uses to calculate the sample size for one-sample and two-sample t-tests? Thanks, Caimiao [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] More efficient matrix computation
JTW wrote: I have a 20x3 matrix as follows: m - replicate(3, matrix(rnorm(20),20,1)) I need to compute, say, 95th and 99th percentiles of each column such that the resulting matrix becomes 2x3 with each row representing the respective percentile. My best effort is to compute one column at a time as follows: quantile(m[,1], c(0.95, 0.99)) To do the same for columns 2 and 3, I would simply change the column number accordingly. Clearly, this is not very efficient as I may have a large matrix (e.g., 100,000x500) to work with. Any help with the code is appreciated. Jack Wang How about: apply(m, 2, quantile, c(0.95, 0.99)) --sundar __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] sample size for t-tests
On Tue, 31 Aug 2004, Caimiao Wei wrote: Dear all, Could any one please tell me the exact formula R uses to calculate the sample size for one-sample and two-sample t-tests? Thanks, There isn't a formula in closed form. The exact procedure is in the code. In words: the critical value for the t-test comes from a t-distribution. The distribution of the t-statistic under the alternative is a non-central t-distribution, and the sample size is adjusted to get the requested power. -thomas __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Obtaining Least Square Means with lme (mixed models)
Hello, I was wondering how one could obtain LSM for fixed factors in a mixed-effect model using lme. In other words, if I have model such as: model-lme(yield~period+treatment+period*treatment, data=data,random=~1|cow) how can I get LSM for period, treatment and their interaction? Thanks a lot! Alex __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] spatial autcorrelation in glmmPQL
I am unable to specify error spatial autocorrelation structure in glmmPQL: DATA x and y coordinates for sample points at which presence/absence of seedlings and canopy openness were recorded in different forest stands. QUESTION Does seedling density increase with canopy openness? ANALYSIS Initial analysis using glmmPQL with binomial errors and stand as random effect result - glmmPQL(seedlings ~ canopy, random = ~1|stand, family=binomial, data=regen) Semivariogram of residuals from this fit using geoR showed spatial autocorrelation with range 34.9 m and relative nugget of 75%. Therefore I tried to create a corStruct object: corel - corSpher(value = c(34.9,0.75), form = ~x+y, nugget=TRUE) corel - Initialize(corel, data = regen) And specify the structure in a new glmmPQL result2 - glmmPQL(seedlings ~ canopy, random = ~|stand, family = binomial, correlation = corel, data=regen) PROBLEM I get the error message: iteration 1 Error in eval(expr, envir, enclos) : Object x not found (Running R 1.9.1 on Windows ME) I appear to be implementing corSpher and Initialize incorrectly- any help greatly appreciated. Dr. Daniel P. Bebber Department of Plant Sciences University of Oxford South Parks Road Oxford OX1 3RB Tel. 01865 275060 --- __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] add single contour line to levelplot
Hello, I want to add a single contour line to a levelplot but can't figure out how to do it 'on-the-fly'. When I include the last line in the code below, I get the following error: Error in NextMethod([) : Argument subscripts is missing, with no default Any tips on how to fix this are greatly appreciated! Ian Jonsen levelplot(dens~nu*sigma,data=tsurflikB,at=c(-20,-10,-5,-4,-3,-2,-1), col.regions=gray(seq(0.5,0.9,length=14)),region=T,colorkey=F,aspect=1, panel=function(x,y,z,...){ panel.levelplot(x=x,y=y,z=z,...) lpoints(x[z==0],y[z==0],pch=16,cex=0.5) panel.levelplot(z~x*y,at=-3,contour=T,aspect=1) } ) ___ Ian Jonsen, Postdoctoral Fellow Dept Biology, Dalhousie University, Halifax, NS, CAN Phone: 902 494 3910; Fax: 902 494 3736 [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Rcmdr X11 protocol error message
Michael Bibo [EMAIL PROTECTED] writes: John Fox jfox at mcmaster.ca writes: Dear Michael, A question: Do you observe these problems with tcltk in general or just with the Rcmdr package? Is it possible for you to test a Tcl/Tk program outside of R? And Peter asked which version of Tcl/Tk. Apparently my system has version 8.4 installed, specifically: Tcl-8.4.5-3-mdk (tclsh8.4) and Tk-8.4.5-3-mdk (libtk8.4.so). As I understand it, these are installed from RPMs on the installation DVD, and I note that they are mandrake specific. John - I wasn't sure if I had any other Tcl/Tk applications installed (it's not always obvious when installing from RPM's). I have certainly not encountered these error messages with any other application. I quickly downloaded WISH Supernotepad 1.2.3. This application requires Tcl/Tk 8.4 or greater. There are no graphics window in this application, but plenty of dialogue boxes. It gave no errors. Is this an appropriate test? If this is a mandrake-configuration-specific problem, it may not be worth spending too much time investigating, as R Commandr still works. I can always try re-installing Tcl/Tk from source when/if I have time. I don't think we have evidence that it's the Tcl installation, although it could be (Google suggests that there have been problems with at least some versions, although most references seem rather old). I can't seem to reproduce the effect with SUSE's tk-8.4.6-28 either. If it is a bug in Rcmdr, then we'd want to find it and you have the only handle on it BTW, sometimes Tk errors allow you to see a trace of the execution. Would this happen to be one of those situations? -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] I've forgotten, why is box() the default?
On Tue, 31 Aug 2004 15:02:13 +0100, Simon Fear [EMAIL PROTECTED] wrote : Not helped by my typo and being in a rush! It's true that plot.default and surely many others do a box, since par(bty) [1] o But this isn't what happens if you do things via axis() ... (see first example in help(axis)). I suspect it's just done that way for programming convenience. Since it's hard or impossible to delete a line once drawn, axis draws the minimal sensible one; if you want more, you use box() to draw it. Duncan Murdoch __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Problem (bug?) with vector indices
Hi all, I found a problem on R that looks like a bug to me. I am working with R for windows verson 1.9.0, but the problem also happens on the linux version 1.9.0. I hope that the problem is on my reasoning, and I would be grateful if someone could explain me why this happens. There goes the example: x-seq(-2,2,0.001) y-rep(0,length(x)) for (i in x){y[1000*i+2001]-i^2} plot(x,y)# Here I found something wrong which(y==0) [1] 244 353 357 361 365 369 482 486 490 494 579 583 587 591 595 [16] 599 603 607 611 615 619 708 712 716 720 724 728 732 736 740 [31] 744 837 841 845 849 853 857 861 865 869 962 966 970 974 978 [46] 982 986 990 994 998 2001 x[244] [1] -1.757 i-x[244] 1000*i+2001 [1] 244 y[1000*i+2001] [1] 3.087049 y[244] [1] 0 y[243] [1] 3.087049 What means that the expression 1000*i+2001, which is equal to 244, inside the brackets is evaluated as 243! The same occurs with every value of x that results in a zero value of y, the right value is stored in y one position before it should be, overwriting the previous value at this position. The bug is on R or on my mind? Thanks a lot. Rodrigo Drummond Rodrigo D. Drummond Laboratorio Genoma Funcional Centro de Biologia Molecular e Eng. Genetica Universidade Estadual de Campinas Caixa Postal 6010 13083-875 - Campinas - SP - Brasil Tel: xx-19-3788-1119 Fax: xx-19-3788-1089 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Problem (bug?) with vector indices
Rodrigo Drummond [EMAIL PROTECTED] writes: Hi all, I found a problem on R that looks like a bug to me. I am working with R for windows verson 1.9.0, but the problem also happens on the linux version 1.9.0. I hope that the problem is on my reasoning, and I would be grateful if someone could explain me why this happens. There goes the example: x-seq(-2,2,0.001) y-rep(0,length(x)) for (i in x){y[1000*i+2001]-i^2} ... x[244] [1] -1.757 i-x[244] 1000*i+2001 [1] 244 y[1000*i+2001] [1] 3.087049 y[244] [1] 0 y[243] [1] 3.087049 What means that the expression 1000*i+2001, which is equal to 244, inside the brackets is evaluated as 243! The same occurs with every value of x that results in a zero value of y, the right value is stored in y one position before it should be, overwriting the previous value at this position. The bug is on R or on my mind? The latter: Have a look at (1000*x[244]+2001)-244 Beware floating point arithmetic! -- O__ Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] add single contour line to levelplot
Ian Jonsen wrote: Hello, I want to add a single contour line to a levelplot but can't figure out how to do it 'on-the-fly'. When I include the last line in the code below, I get the following error: Error in NextMethod([) : Argument subscripts is missing, with no default Any tips on how to fix this are greatly appreciated! Ian Jonsen levelplot(dens~nu*sigma,data=tsurflikB,at=c(-20,-10,-5,-4,-3,-2,-1), col.regions=gray(seq(0.5,0.9,length=14)),region=T,colorkey=F,aspect=1, panel=function(x,y,z,...){ panel.levelplot(x=x,y=y,z=z,...) lpoints(x[z==0],y[z==0],pch=16,cex=0.5) panel.levelplot(z~x*y,at=-3,contour=T,aspect=1) } ) You have misued panel.levelplot and also forgot (neglected?) to include ... in the second call. I think you want something like: # from ?levelplot library(lattice) x - seq(pi/4, 5 * pi, length = 100) y - seq(pi/4, 5 * pi, length = 100) r - as.vector(sqrt(outer(x^2, y^2, +))) grid - expand.grid(x = x, y = y) grid$z - cos(r^2) * exp(-r/(pi^3)) levelplot(z ~ x * y, grid, cuts = 50, scales=list(log=e), xlab = , ylab = , main = Weird Function, sub = with log scales, panel = function(x, y, z, at, contour, ...) { panel.levelplot(x, y, z, at = 0, contour = TRUE, ...) lpoints(x[z = -0.5],y[z = -0.5], pch = 16, cex = 0.5) }, colorkey = FALSE, region = TRUE) Note the usage of I also think you do not need two calls to panel.levelplot, though I may be misunderstanding your intention. One will suffice, as it does here. --sundar __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Sparse Matrices in R
I have data in i,j,r format, where r is the value in location A[i,j] for some imaginary matrix A. I need to build this matrix A, but given the sizes of i and j, I believe that using a sparse format would be most adequate. Hopefully this will allow me to perform some basic matrix manipulation such as multiplication, addition, rowsums, transpositions, subsetting etc etc. Is there any way to achieve this goal in R? Thanks, Danny [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Sparse Matrices in R
Take a look at the package SparseM url:www.econ.uiuc.edu/~rogerRoger Koenker email [EMAIL PROTECTED] Department of Economics vox:217-333-4558University of Illinois fax:217-244-6678Champaign, IL 61820 On Aug 31, 2004, at 4:52 PM, Danny Heuman wrote: I have data in i,j,r format, where r is the value in location A[i,j] for some imaginary matrix A. I need to build this matrix A, but given the sizes of i and j, I believe that using a sparse format would be most adequate. Hopefully this will allow me to perform some basic matrix manipulation such as multiplication, addition, rowsums, transpositions, subsetting etc etc. Is there any way to achieve this goal in R? Thanks, Danny [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Sparse Matrices in R
Hi! help.search(sparse matrix) graph2SparseM(graph)Coercion methods between graphs and sparse matrices tripletMatrix-class(Matrix) Class tripletMatrix sparse matrices in triplet form SparseM.hb(SparseM) Harwell-Boeing Format Sparse Matrices image,matrix.csr-method(SparseM) Image Plot for Sparse Matrices etc . /E *** REPLY SEPARATOR *** On 8/31/2004 at 5:52 PM Danny Heuman wrote: I have data in i,j,r format, where r is the value in location A[i,j] for some imaginary matrix A. I need to build this matrix A, but given the sizes of i and j, I believe that using a sparse format would be most adequate. Hopefully this will allow me to perform some basic matrix manipulation such as multiplication, addition, rowsums, transpositions, subsetting etc etc. Is there any way to achieve this goal in R? Thanks, Danny [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Dipl. bio-chem. Witold Eryk Wolski @ MPI-Moleculare Genetic Ihnestrasse 63-73 14195 Berlin'v' tel: 0049-30-83875219/ \ mail: [EMAIL PROTECTED]---W-Whttp://www.molgen.mpg.de/~wolski [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Problem with seq.dates in chron
I get faulty output from seq.dates() if I specify a length that is too long. For example, I ask for 129 months in the following call to the function, but it returns 131: R.version.string [1] R version 1.9.1, 2004-06-21 startdatetime - chron(dates=01/01/1995, times=00:00:00) beg.month.datetimes - seq.dates(from=startdatetime, by=months, length=129) beg.month.datetimes [1] 01/01/95 02/01/95 03/01/95 04/01/95 05/01/95 06/01/95 07/01/95 08/01/95 [9] 09/01/95 10/01/95 11/01/95 12/01/95 01/01/96 02/01/96 03/01/96 04/01/96 [17] 05/01/96 06/01/96 07/01/96 08/01/96 09/01/96 10/01/96 11/01/96 12/01/96 [25] 01/01/97 02/01/97 03/01/97 04/01/97 05/01/97 06/01/97 07/01/97 08/01/97 [33] 09/01/97 10/01/97 11/01/97 12/01/97 01/01/98 02/01/98 03/01/98 04/01/98 [41] 05/01/98 06/01/98 07/01/98 08/01/98 09/01/98 10/01/98 11/01/98 12/01/98 [49] 01/01/99 02/01/99 03/01/99 04/01/99 05/01/99 06/01/99 07/01/99 08/01/99 [57] 09/01/99 10/01/99 11/01/99 12/01/99 01/01/00 02/01/00 03/01/00 04/01/00 [65] 05/01/00 06/01/00 07/01/00 08/01/00 09/01/00 10/01/00 11/01/00 12/01/00 [73] 01/01/01 02/01/01 03/01/01 04/01/01 05/01/01 06/01/01 07/01/01 08/01/01 [81] 09/01/01 10/01/01 11/01/01 12/01/01 01/01/02 02/01/02 03/01/02 04/01/02 [89] 05/01/02 06/01/02 07/01/02 08/01/02 09/01/02 10/01/02 11/01/02 12/01/02 [97] 01/01/03 02/01/03 03/01/03 04/01/03 05/01/03 06/01/03 07/01/03 08/01/03 [105] 09/01/03 10/01/03 11/01/03 12/01/03 01/01/04 02/01/04 03/01/04 04/01/04 [113] 05/01/04 06/01/04 07/01/04 08/01/04 09/01/04 10/01/04 11/01/04 12/01/04 [121] 01/01/05 02/01/05 03/01/05 04/01/05 05/01/05 06/01/05 07/01/05 08/01/05 [129] 09/01/05 10/01/05 11/01/05 Is this a bug, and can it be fixed? Scott Waichler Pacific Northwest National Laboratory scott.waichleratpnl.gov __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] subselect install problem
Trying to install subselect v0.8 on Redhat 7.3 and R 1.8.1 fails (below). Any help is greatly appreciated. Xiao-Jun * Installing *source* package 'subselect' ... ** libs f2c anneal.f anneal.c anneal: Error on line 263: Declaration error for fica: adjustable dimension on non-argument Error on line 263: Declaration error for valp: adjustable dimension on non-argument Error on line 263: Declaration error for auxw: adjustable dimension on non-argument Error on line 263: wr_ardecls: nonconstant array size Error on line 263: wr_ardecls: nonconstant array size Error on line 263: wr_ardecls: nonconstant array size make: *** [anneal.o] Error 1 ERROR: compilation failed for package 'subselect' __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
Re: [R] Problem with seq.dates in chron
Waichler, Scott R Scott.Waichler at pnl.gov writes: I get faulty output from seq.dates() if I specify a length that is too long. For example, I ask for 129 months in the following call to the function, but it returns 131: R.version.string [1] R version 1.9.1, 2004-06-21 startdatetime - chron(dates=01/01/1995, times=00:00:00) beg.month.datetimes - seq.dates(from=startdatetime, by=months, length=129) If you look at the seq.dates R source, it sets `to' like this: to - from + (length. - 1) * c(1, 7, 31, 366)[i] where in your case where i is 3 for months. It then picks out those days between from and to having the same day of the month as the from date (or uses a more complex algorithm if the day of the month exceeds 28). The multiplication by 31 gives an overestimate which is OK for a small lengths but overflows to give too many months if length is too large. Since we know the error will always be on the high side, until it is fixed as a workaround you could just subset it: seq.dates(from=startdatetime, by=months, length=129)[1:129] (Perhaps you could forward this to the maintainer of chron.) __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] simpleboot, pairs.boot
Hi. I am a new R user trying to obtain bootstrap confidence intervals around vector correlations of Principal components. My data is a data frame of eigenvector loadings from PC1 in the columns the rows are different species. I want to calculated the vector correlation and bootstrap confidence intervals for all species pairs. I have tried using simpleboot as pairs.boot seems to be an easy way of sampling from two vectors of data. The function I wrote veccor below works outside of the pairs.boot function. However in this function it seems to calculate all 1000 vector correlations as -1. veccor function(x,y) ((sum((x)*(y)))/((sqrt(sum(y^2)))*(sqrt(sum(x^2) x-as.vector(faVC[1,]) y-as.vector(faVC[2,]) boot-pairs.boot(x,y,veccor,1000) # I have also tried using the boot function but only end up sampling from on of my vectors of data. Any suggestions would be greatly appriciated! Thanks! Best, Becca __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html