Re: [R] g2 test...

Fernando Marmolejo Ramos adelaide.edu.au> writes: > [snip] > is it appropriate to use a Log likelihood ratio (G-test) test of independence when dealing with repeated > categorical responses (e.g. 2 by 2 table) instead of the McNemar test? [snip] Not an R question. Try http://stats.stack

Re: [R] Substituting Greek symbols in some tick labels

David, thanks for the reply. No, I don't want to do it by hand because the actual data that I need to label is much larger and variable. The deltas occur in the label character vector for some labels and not others and are in different positions in the character strings. In the example I gave, th

Re: [R] Substituting Greek symbols in some tick labels

On Jul 4, 2013, at 8:14 PM, Eric Archer - NOAA Federal wrote: I have a character vector that I'm using to label ticks in a dotchart. Some of the elements in the vector have an asterisk (*) where a Greek Delta needs to be placed when the plot is generated. Here's a simple example: x <- 1:4

[R] Substituting Greek symbols in some tick labels

I have a character vector that I'm using to label ticks in a dotchart. Some of the elements in the vector have an asterisk (*) where a Greek Delta needs to be placed when the plot is generated. Here's a simple example: x <- 1:4 x.lab <- c("a*a", "bbb", "c*c", "ddd") dotchart(x, labels = x.lab) Th

[R] g2 test...

dear R users is it appropriate to use a Log likelihood ratio (G-test) test of independence when dealing with repeated categorical responses (e.g. 2 by 2 table) instead of the McNemar test? thanks fer [[alternative HTML version deleted]] __

Re: [R] Non-linear modelling with several variables including a categorical variable

Dear Prof. Nash, I tried to run nls with the nlxb function and as you mention it is fairly slower in terms of running the code. However, if I would have used this function earlier I would have saved a lot of time trying to find the start values. The output looks a little bit sloppy but I think it

Re: [R] bootstrapping respecting subject level information

Josh's comment prompted me to check mty go-to reference. Davison and Hinckley (1997 Section 3.8) recommend sampling the Subjects, but not within the Subjects. Cheers Andrew On Thu, Jul 04, 2013 at 05:53:58PM -0700, Joshua Wiley wrote: > Hi, > > It is not the easiest to follow code, but when I w

Re: [R] bootstrapping respecting subject level information

Hi, It is not the easiest to follow code, but when I was working at UCLA, I wrote a page demonstrating a multilevel bootstrap, where I use a two stage sampler, (re)sampling at each level. In your case, could be first draw subjects, then draw observations within subjects. A strata only option doe

Re: [R] bootstrapping respecting subject level information

I think that in the case of a 2*2 balanced, replicated design such as this one, interpreting the interaction should be safe. Cheers Andrew On Fri, Jul 5, 2013 at 9:38 AM, David Winsemius wrote: > > On Jul 3, 2013, at 7:19 PM, Sol Lago wrote: > >> Hi there, >> >> This is the first time I use thi

Re: [R] bootstrapping respecting subject level information

I'd like to preface this answer by suggesting that if you have multiple measurements within subjects then you should possibly be thinking about using mixed-effects models. Here you have a balanced design and seem to be thinking about a constrained bootstrap, but I don't know whether the resulting

Re: [R] Subsetting multiple rows of a data frame at once

> xt<- c(1.05, 2.85, 3.40, 4.25, 0.25, 3.05, 3.70, 0.20, 0.30, 0.70, 1.05, > 1.20, 1.40, 1.90, > 2.70, 3.25, 3.55, 4.60, 2.05, 2.15, 3.70, 4.85, 4.90, 1.60, 2.45, 3.20, 3.90, > 4.45) > > yt<- c(0.25, 0.10, 0.90, 0.25, 1.05, 1.70, 2.05, 2.90, 2.35, 2.60, 2.55, > 2.15, 2.75, 2.05, > 2.70, 2.25, 2

Re: [R] R and MatLab implementations of the same model differs

On Jul 5, 2013, at 00:41 , Jannetta Steyn wrote: > Hi All > > Thanks for all the comments. I have looked at everything you have pointed > out and this is the situation at the moment: > >> I don't think so. >> In the R script you have >> >> init = c(v_axon_AB=-55,mNa_axon_AB=1,hNa_axon_AB=0,mK_

Re: [R] bootstrapping respecting subject level information

On Jul 3, 2013, at 7:19 PM, Sol Lago wrote: > Hi there, > > This is the first time I use this forum, and I want to say from the start I > am not a skilled programmer. So please let me know if the question or code > were unclear! > > I am trying to bootstrap an interaction (that is my test sta

Re: [R] setClass confusion

On Jul 3, 2013, at 4:54 PM, Erin Hodgess wrote: > Dear R People: > > > I am experimenting with S4 classes and methods but am having trouble with > setting up a class. > > Here is an example: > >> buzz <- setClass("buzz",slots=c(x="matrix"), > + validity <- function(object) { > + if(is.matrix(

Re: [R] R and MatLab implementations of the same model differs

Hi All Thanks for all the comments. I have looked at everything you have pointed out and this is the situation at the moment: > I don't think so. > In the R script you have > > init = c(v_axon_AB=-55,mNa_axon_AB=1,hNa_axon_AB=0,mK_axon_AB=1) > > That is not the same as in your Matlab script. To m

Re: [R] Lattice barchart with error bars

Shaun: I understand that this type of plot is standard in many disciplines, but it really is awful (google on 'Dynamite plots' for some more erudite perspectives). Have you considered bwplot() for your unaggregated data instead? (No need to reply. It's July 4, and I'm just waving a little flag fo

Re: [R] Help

Well, it implies either 2help or help^2, right? In this case, it seems oddly appropriate, as really needing help getting packages is sort of greater in a "need hierarchy" than coding help, as the latter is largely useless without the former. At one point, I seem to recall discussion of whether th

Re: [R] Help

Please (for crying out loud!) use a meaningful subject line!!! This list is called "r-help"!!! Using a subject line of "Help" is completely vacuous. cheers, Rolf Turner On 04/07/13 20:33, Aboagye-sarfo, Patrick wrote: I am trying to download packages and the message I get is as f

[R] Lattice barchart with error bars

Hi, I'd like to draw a lattice barchart of means with error bars to show the standard deviation. I have the barchart, how do I add the error bars? require(datasets) require(lattice) x <- aggregate(weight ~ Diet, ChickWeight, function(x) c(mean=mean(x), sd=sd(x))) barchart(weight[,'mean'] ~ Diet,

[R] coxph doesn't converge when including factor variables

Hello, I have a dataset of 371 observations. When I run coxph with numeric variables it works fine. However, when I try to add factor variables it returns "Ran out of iterations and the model did not converge" There is something very strange with the factors - some of them should actually be

Re: [R] R and MatLab implementations of the same model differs

On Jul 4, 2013, at 20:14 , Berend Hasselman wrote: > > On 04-07-2013, at 19:56, peter dalgaard wrote: > >> >> On Jul 4, 2013, at 19:11 , Berend Hasselman wrote: >> >>> >>> On 04-07-2013, at 18:42, Jannetta Steyn wrote: >>> Hi Ben and others I don't quite know how to explai

Re: [R] Meta-analysis on a repeated measures design with multiple trials per subject using metafor

Dear Marc, Let me see if I understand the type of data you have. You say that you have 5 experiments. And within each experiment, you have n subjects and for each subject, you have data in the form described in your post. Now for each subject, you want to calculate some kind of measure that qua

Re: [R] help on selecting values of an object

Hi, You could use: d1<- data.frame(a,b) k1<-data.frame(a=k) library(plyr) join(k1,d1,by="a")[,2] #[1] 4 4 6 6 7 7 6 A.K. - Original Message - From: Andras Farkas To: r-help@r-project.org Cc: Sent: Thursday, July 4, 2013 2:09 PM Subject: [R] help on selecting values of an object Dear

Re: [R] help on selecting values of an object

I think this is what you are looking for > f <- d[match(k, d[,1]), 2] > f [1] 4 4 6 6 7 7 6 - David L Carlson Associate Professor of Anthropology Texas A&M University College Station, TX 77840-4352 -Original Message- From: r-help-boun...@r-project.org

Re: [R] R and MatLab implementations of the same model differs

On 04-07-2013, at 19:56, peter dalgaard wrote: > > On Jul 4, 2013, at 19:11 , Berend Hasselman wrote: > >> >> On 04-07-2013, at 18:42, Jannetta Steyn wrote: >> >>> Hi Ben and others >>> >>> I don't quite know how to explain the "doesn't work" in more detail without >>> any visual aid. >>

[R] help on selecting values of an object

Dear List, please provide some input on the following: we have a <-c(0,1,2,3) b <-c(4,5,6,7) d <-cbind(a,b) k <-c(0,0,2,2,3,3,2) "k" in this case consists of some values of "d[,1]" in a random sequence. What I am trying to do is to create an object "f" that would have the values of "d[,2]" in

Re: [R] R and MatLab implementations of the same model differs

On Jul 4, 2013, at 19:11 , Berend Hasselman wrote: > > On 04-07-2013, at 18:42, Jannetta Steyn wrote: > >> Hi Ben and others >> >> I don't quite know how to explain the "doesn't work" in more detail without >> any visual aid. > > You said that R got into an indefinite loop, whatever that may

Re: [R] R and MatLab implementations of the same model differs

On 04-07-2013, at 18:42, Jannetta Steyn wrote: > Hi Ben and others > > I don't quite know how to explain the "doesn't work" in more detail without > any visual aid. You said that R got into an indefinite loop, whatever that maybe. > When you run the two scripts it is easy to see the > differe

[R] iterative methods

Dear R users, Please help me with some documentation for newbie about R programming, algorithms, create iterative C++ function (like for, while, if , etc). Thank you very much! -- --- Catalin-Constantin ROIBU Lecturer PhD, Forestry engineer Forestry Faculty of Suceava Str. Universitatii no. 13

Re: [R] R and MatLab implementations of the same model differs

Hi Ben and others I don't quite know how to explain the "doesn't work" in more detail without any visual aid. When you run the two scripts it is easy to see the difference. MatLab produces a line on x= -55. This is what I expect - a more or less straight line. R on the other hand the result drops

Re: [R] R and MatLab implementations of the same model differs

Berend Hasselman xs4all.nl> writes: > > > On 04-07-2013, at 17:15, Jannetta Steyn henning.org> wrote: > > > Hi folks > > > > I have implemented a model of a neuron using Hodgkin Huxley equations in > > both R and MatLab. My preference is to work with R but R is not giving me > > the correct

Re: [R] R and MatLab implementations of the same model differs

On 04-07-2013, at 17:15, Jannetta Steyn wrote: > Hi folks > > I have implemented a model of a neuron using Hodgkin Huxley equations in > both R and MatLab. My preference is to work with R but R is not giving me > the correct results. I also can't use ode45 as it just seems to go into an > indef

[R] R and MatLab implementations of the same model differs

Hi folks I have implemented a model of a neuron using Hodgkin Huxley equations in both R and MatLab. My preference is to work with R but R is not giving me the correct results. I also can't use ode45 as it just seems to go into an indefinite loop. However, the MatLab implementation work fine with

Re: [R] how to choose dates data?

Also > yrs <- as.Date(c("2007-01-01", "2009-12-31")) > day[day>=yrs[1] & day<=yrs[2]] [1] "2008-04-12" "2008-04-12" "2008-04-12" Should be fast if there are many days to process since it converts the search criteria, not the data. - David L Carlson Associate P

Re: [R] ggplot2

Thanks Alex but the idea is to use dput() for the data so that readers can simply copy and paste it into R and have a working dataset. I have a very small data.frame called dd. Var1 Var2 1 A1 2 B1 3 C1 4 A2 5 B2 6 C2 7 A3 8 B3 9

Re: [R] ggplot2

Here last function (x, ...) { UseMethod("last") } I can share 100 rows if you want, but I don`t you will find it useful, since I want to perform my time-series on a single row. Here`s one: dezdiff.dt[nrow(dezdiff.dt),] datetime X1X2X3 X4X5

Re: [R] ggplot2

I think we need to see some of the original data. See ?dput for how to supply data in a email. We probably don't need all the data, I'd suggest perhaps 100 rows or so. Try dput(head(yourdata, 100)) BTW what is "last"? John Kane Kingston ON Canada > -Original Message- > From: alexa

Re: [R] how to choose dates data?

Hi, You could try: day<-as.Date(c("2008-04-12","2011-07-02","2011-09-02","2008-04-12","2008-04-12"))  indx<-gsub("-.*","",day)  day[indx>="2007" & indx<="2009"] #[1] "2008-04-12" "2008-04-12" "2008-04-12" #or library(xts) xt1<- xts(seq_along(day),day) index(xt1["2007/2009"]) #[1] "2008-04-12" "200

[R] ggplot2

Hello Folks, I have a database of 2000+ days with 35 observations each. I am trying to modeling a time series by day, but it seems a problem that I don str(last(dezdiff)) An ‘xts’ object on 2013-07-04/2013-07-04 containing: Data: num [1, 1:35] 1.9016 0.7545 0.225 0.0727 0.1357 ... - attr(*, "di

Re: [R] polr?

On Jul 4, 2013, at 14:38 , Dániel Kehl wrote: > Dear Prof Ripley, > > could you be just a little more specific? He'll likely find that difficult. It's sort of like if you had data like this 25 75 75 25 25 75 and did a trend test. The trend test _assumes_ that the effect is increasing, and

Re: [R] list construction with automatic names

>> I often find myself (wanting t)o constructing lists or >> data.frames > > Apologies; previous post should have said "Read R inforno on 'Growing > Objects'" and should have added the URL: > > http://www.burns-stat.com/pages/Tutor/R_inferno.pdf > > S Ellison > > > > ***

Re: [R] polr?

Dear Prof Ripley, could you be just a little more specific? Thanks a lot daniel Feladó: Prof Brian Ripley [rip...@stats.ox.ac.uk] Küldve: 2013. július 4. 14:14 To: Dániel Kehl Cc: r-help Tárgy: Re: [R] polr? On 04/07/2013 12:59, Dániel Kehl wrote: > Dear

Re: [R] list construction with automatic names

> I often find myself (wanting t)o constructing lists or > data.frames: Try something like fl<-list() for (ix in id) { fl[[basename(ix)]] <- list.files(ix) } But read R Inferno on inrementing objects first. It isn't efficient in memory terms (unless that's improved in recent versions of

Re: [R] list construction with automatic names

> I often find myself (wanting t)o constructing lists or > data.frames Apologies; previous post should have said "Read R inforno on 'Growing Objects'" and should have added the URL: http://www.burns-stat.com/pages/Tutor/R_inferno.pdf S Ellison **

Re: [R] polr?

On 04/07/2013 12:59, Dániel Kehl wrote: Dear R users, I have a dataset with two ordered variables, tr_x1 and tr_y1. A crosstable of them can bee seen below. tr_x1 tr_y1 -101 -1 629 100 629 0 1396 4353 1443 1 668 126 655 It is clear that if tr_x1 is 0, it ha

[R] list construction with automatic names

I often find myself (wanting t)o constructing lists or data.frames like so: #Find files in subdirs ii <- 0 for (ix in id) { ii <- ii + 1 if (ii == 1) { fl <- list(basename(ix) = list.files(ix)) } else { fl <- c(fl, list(basename(ix) = list.fi

[R] polr?

Dear R users, I have a dataset with two ordered variables, tr_x1 and tr_y1. A crosstable of them can bee seen below. tr_x1 tr_y1 -101 -1 629 100 629 0 1396 4353 1443 1 668 126 655 It is clear that if tr_x1 is 0, it has an effect on tr_y1. A chi-square statistic

Re: [R] Help

> I am trying to download packages and the message I get is as follows > > Warning: unable to access index for repository > http://cran.ms.unimelb.edu.au/bin/windows/contrib/3.0 Your internet connection is probably either not working or (more likely) blocked by firewall or site access policie

[R] Help

I am trying to download packages and the message I get is as follows > utils:::menuInstallPkgs() --- Please select a CRAN mirror for use in this session --- Warning: unable to access index for repository http://cran.ms.unimelb.edu.au/bin/windows/contrib/3.0 Warning: unable to access index for rep

Re: [R] how to choose dates data?

On 04/07/2013 11:36, Dante.py wrote: > They are really good methods. Thank you very much. > Could I ask one more question. Is there obvious difference between these > two methods? Yes. It is obvious that the type of x differs. > > > 2013/7/4 Prof Brian Ripley >

Re: [R] Non-linear modelling with several variables including a categorical variable

Thanks for the advise Prof Nash, but I'm not sure if I understood it right. I managed to make a new model based on what I think you meant. What I did is; I created 3 variables (cat1, cat2, cat3) one for each category with either the value 1 or 0 and added these to the model so they work as differen

Re: [R] how to choose dates data?

They are really good methods. Thank you very much. Could I ask one more question. Is there obvious difference between these two methods? 2013/7/4 Prof Brian Ripley > On 04/07/2013 08:04, Dante.py wrote: > >> I have a method which is not so smart --use grep to match the pattern. >> for example:

[R] Binomial Regression and nnet

Dear All, I am playing with different models/packages (random forest, logistic regression, gbm etc...) for a problem of binomial regression (i.e. the outcome is 0/1, dead or alive etc...). I have used in the past the multinom function from the nnet library which uses the neural networks for multino

Re: [R] how to choose dates data?

On 04/07/2013 08:04, Dante.py wrote: I have a method which is not so smart --use grep to match the pattern. for example: dates <- c("02/27/92", "02/27/92", "01/14/92", "02/28/92", "02/01/91") day <- as.Date(dates, "%m/%d/%y") day: [1] "1992-02-27" "1992-02-27" "1992-01-14" "1992-02-28" "1991-02-0

Re: [R] modify timestemp

On Wed, 03 Jul 2013, Ye Lin writes: > Hey All, > > I want to standardize my timestamp which is formatted as hh:mm:ss > > My data looks like this: > > Date Time > 01/01/2013 00:09:01 > 01/02/2013 00:10:14 > 01/03/2013 00:11:27 > 01/04/2013 00:12:40 > 01/05/2013 00:13:53 > 01/06/2013 00:1

Re: [R] how to choose dates data?

I have a method which is not so smart --use grep to match the pattern. for example: dates <- c("02/27/92", "02/27/92", "01/14/92", "02/28/92", "02/01/91") day <- as.Date(dates, "%m/%d/%y") day: [1] "1992-02-27" "1992-02-27" "1992-01-14" "1992-02-28" "1991-02-01" If I want to search for 1991, I can

[R] cluster analysis

I want to do Agglomerative Hierarchical clustering using complete linkage method in R using the function agnes or hclust. 1. Can i do a cluster analysis of h=(n+p+1)/2 out of n observation?  note that p=nomber of variables(dependent and independent) 2. Can i plot the dendrogram and get the clust

Re: [R] Subsetting multiple rows of a data frame at once

Hi, carbon.fit = expand.grid(list(x=seq(0, 5, 0.01), y=seq(0, 5, 0.01)))  dim(carbon.fit) #[1] 251001  2  xtNew<-sprintf("%.2f",xt)  ytNew<- sprintf("%.2f",yt)  carbon.fit[]<- lapply(carbon.fit,function(x) sprintf("%.2f",x)) res<-do.call(rbind,lapply(seq_along(xtNew),function(i) subset(carb

Re: [R] Subsetting multiple rows of a data frame at once

Hi, Possibly, FAQ 7.31 Using the same example: set.seed(24) df<- data.frame(x=sample(seq(0.25,4.25,by=.05),1e5,replace=TRUE),y= sample(seq(0.10,1.05,by=.05),1e5,replace=TRUE),z=rnorm(1e5)) dfOld<- df  df[,1:2]<- lapply(df[,1:2],function(x) sprintf("%.2f",x)) x1<- c(1.05,2.85,3.40,4.25,0.25) y1<-

[R] bootstrapping respecting subject level information

Hi there, This is the first time I use this forum, and I want to say from the start I am not a skilled programmer. So please let me know if the question or code were unclear! I am trying to bootstrap an interaction (that is my test statistic) using the package "boot". My problem is that for e