Re: [R] dependent p.values in R
dear michael thanks for your input i do agree in visual tests (indeed a recent paper in TAS=http://amstat.tandfonline.com/doi/abs/10.1080/00031305.2015.1077728 ) as a matter of fact, the test i'm after is simply for comparative purposes with some visualisation techniques best f Fernando Marmolejo-Ramos Postdoctoral Fellow Gösta Ekman Laboratory Department of Psychology Stockholm University Frescati Hagväg 9A, Stockholm 114 19 Sweden ph = +46 08-16 46 07 website = http://sites.google.com/site/fernandomarmolejoramos/ From: Michael Friendly <frien...@yorku.ca> Sent: Sunday, 10 July 2016 7:51 PM To: Fernando Marmolejo Ramos; Marc Girondot; r-help@r-project.org Subject: Re: dependent p.values in R Hello Fernando First, ask yourself what Gosta Ekman would have said if you asked him this question. He would have asked "does it make any difference to your conclusion?" He might also have asked you "Did you do a visual test?" Plot your data as a QQ plot or density plot? If the test doesn't make a difference in conclusions, it is a waste of your time (and ours) to worry about how to cite a 'combined p.value' (if such an animal exists), presumably to more decimal places than is worth worrying about. If the test *does* make a difference about normality, then ask yourself does the degree of non-normality impede my substantive conclusions. HTH, -ichael On 7/10/16 3:39 AM, Fernando Marmolejo Ramos wrote: > hi marc > > say i have a vector with some x number of observations > > x = c(23, 56, 123, . ) > > and i want to know how normal it is > > as there are many normality tests, i want to combine their p.values > > so, suppose i use shapiro.wilk, anderson darling and jarque bera and each > will give a pvalue > > i could simply average those p,values but to my knowledge that approach is > biased > > so i thought, in the same way there is a method to combine independent > pvalues (e.g. stouffer method); is there a way to combine dependent pvalues? > > best > > f > > > Fernando Marmolejo-Ramos > Postdoctoral Fellow > Gösta Ekman Laboratory > Department of Psychology > Stockholm University > Frescati Hagväg 9A, Stockholm 114 19 > Sweden > > ph = +46 08-16 46 07 > website = http://sites.google.com/site/fernandomarmolejoramos/ > > > > ________ > From: Marc Girondot <marc.giron...@u-psud.fr> > Sent: Sunday, 10 July 2016 8:25 AM > To: r-help@r-project.org; Fernando Marmolejo Ramos > Subject: Re: [R] dependent p.values > > Le 09/07/2016 à 17:17, Fernando Marmolejo Ramos a écrit : >> hi all >> >> >> does any one know a method to combine dependent p.values? >> >> > First, it is a stats question and not a R question. So you could have > better chance to ask this in stackexchange forum. > Second, your question is difficult to answer without context: why > p.values are dependent ? Do they come from the same dataset ? Or are > they linked by an external source ? For both these situations, combining > dependent p.values seems strange for me. > When you will ask question in stackexchange, be more precise. > Sincerely, > Marc Girondot > > -- > __ > Marc Girondot, Pr > > Laboratoire Ecologie, Systématique et Evolution > Equipe de Conservation des Populations et des Communautés > CNRS, AgroParisTech et Université Paris-Sud 11 , UMR 8079 > Bâtiment 362 > 91405 Orsay Cedex, France > > Tel: 33 1 (0)1.69.15.72.30 Fax: 33 1 (0)1.69.15.73.53 > e-mail: marc.giron...@u-psud.fr > Web: http://www.ese.u-psud.fr/epc/conservation/Marc.html > Skype: girondot > > __ 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.
Re: [R] dependent p.values in R
hi marc say i have a vector with some x number of observations x = c(23, 56, 123, . ) and i want to know how normal it is as there are many normality tests, i want to combine their p.values so, suppose i use shapiro.wilk, anderson darling and jarque bera and each will give a pvalue i could simply average those p,values but to my knowledge that approach is biased so i thought, in the same way there is a method to combine independent pvalues (e.g. stouffer method); is there a way to combine dependent pvalues? best f Fernando Marmolejo-Ramos Postdoctoral Fellow Gösta Ekman Laboratory Department of Psychology Stockholm University Frescati Hagväg 9A, Stockholm 114 19 Sweden ph = +46 08-16 46 07 website = http://sites.google.com/site/fernandomarmolejoramos/ From: Marc Girondot <marc.giron...@u-psud.fr> Sent: Sunday, 10 July 2016 8:25 AM To: r-help@r-project.org; Fernando Marmolejo Ramos Subject: Re: [R] dependent p.values Le 09/07/2016 à 17:17, Fernando Marmolejo Ramos a écrit : > hi all > > > does any one know a method to combine dependent p.values? > > First, it is a stats question and not a R question. So you could have better chance to ask this in stackexchange forum. Second, your question is difficult to answer without context: why p.values are dependent ? Do they come from the same dataset ? Or are they linked by an external source ? For both these situations, combining dependent p.values seems strange for me. When you will ask question in stackexchange, be more precise. Sincerely, Marc Girondot -- __ Marc Girondot, Pr Laboratoire Ecologie, Systématique et Evolution Equipe de Conservation des Populations et des Communautés CNRS, AgroParisTech et Université Paris-Sud 11 , UMR 8079 Bâtiment 362 91405 Orsay Cedex, France Tel: 33 1 (0)1.69.15.72.30 Fax: 33 1 (0)1.69.15.73.53 e-mail: marc.giron...@u-psud.fr Web: http://www.ese.u-psud.fr/epc/conservation/Marc.html Skype: girondot __ 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] dependent p.values
hi all does any one know a method to combine dependent p.values? best Fernando Marmolejo-Ramos Postdoctoral Fellow G�sta Ekman Laboratory Department of Psychology Stockholm University Frescati Hagv�g 9A, Stockholm 114 19 Sweden ph = +46 08-16 46 07 website = http://sites.google.com/site/fernandomarmolejoramos/ [[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] 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]] __ 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] calculating the mode in R...
Dear R users Im aware that the package modest is useful to find the mode in an array. However, Id like to know if someone has translated the mode function built-in in MATLAB into R language. This function finds the most frequent value in an array (http://www.mathworks.com/help/techdoc/ref/mode.html). Best Fer __ 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] confidence intervals around p-values
Dear all I wonder if anyone has heard of confidence intervals around p-values... Any pointer would be highly appreciated. Best Fer __ 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] R-help Digest, Vol 91, Issue 9
dear all I wonder if anyone has heard of confidence intervals around p-values... Any pointer would be highly appreciated. Best Fer __ 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] simulating Likert-type data
Dear R members Could someone indicate me how to simulate Likert-type data using the GAMLSS library and how to fit that data? Lets say, 50 random numbers in a variable that goes from -4 to +4 in steps of 1 (i.e., -4, -3, -2, -1, 0, 1, 2, 3, 4) with a mean of 2 and SD of 0.15 and belonging to a normal distribution? Cheers Fer __ 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] transformation and outliers
Dear R people I ask again 1. Is there a published reference presenting the normal score transformation? Is there a published paper (in any field) using that transformation in the analysis of data? And this is a new question 2. The outliers library has a function called rm.outlier. It offers the option of either removing one outlier (the largest or the smallest) or replacing the identified outlier by the mean or median of the data. Does anyone know of a published paper introducing that particular approach (replacing the outlier by the groups mean/median)? Cheers Fer __ 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] normal score transformation...
Dear R members Im interested in reviewing some data transformations. Ive found literature regarding Box-Cox, logarithmic and square-root transformations. I know there is a library called coin which contains a transformation called normal score transformation (see normal_trafo function). However, Ive not been able to find literature on that regard. Does anyone know about bibliography reporting the use of the normal score transformation? And published papers in any field implementing such transformation? Cheers Fer __ 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] Fitting weibull and exponential distributions to left censoring data
Dear Barja Have you looked at gamlss y gamlss.dist libaries? There some functions called WEI, WEI2, and WEI3 Cheers, Fer __ 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] multiple visualisation of a bootstrap
Dear R team Jim Lemon recently said regarding new improvements in PrettyR and plotrix: Remember, it is your whingeing and moaning that have helped to make these packages what they are today. So, this is my whinge: out there is a freeware called Vista, which runs, among other things, bootstrap analyses. Most of the numerical reports given by Vista can be done using packages such as boot, bootstrap, and simpleboot. Ive not explored all the functions in these packages thoroughly (very casually), but the first thing I noticed is that the graphical output is majorly histograms. Vista offers a so-called Multiple Visualisation (MV) of the output of a bootstrap (it shows on the same window scatter plots for a particular variable showing the bootstrapped CI, a scatter plot of the evolution of the bootstrapped mean over many sample sizes, a box plot, a QQ plot, and a histogram). I wonder if there is any manner in which this sort of MV could me mimicked by R in the case of bootstrap. My best guess is that Id have to create a 3 by 2 frame an insert each graph separately or maybe resort to Trellis graphics am I right? Does anyone have a more sophisticated solution? Cheers, Fer __ 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] Interpretation of t.test results
Dear Omar Your t.test is telling that the 2 groups are not different. Now it is not clear if you want to compute the CI for the difference or for each group. If the former, then the outputs provides you with that information. Itd be clearer if you provide the full syntax of your t.test. For example, lets suppose I have 2 groups as vectors a and b, they are independent, they have similar homogeneity, I want to use a two-tailed test, and I want a 95%CI for the difference, so I type t.test(a, b, alternative = c(two.sided), paired = FALSE, var.equal = TRUE, conf.level = 0.95) Cheers, Fer __ 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] A line for SD, colouring simplevioplots, limiting Y axis in barplots, axis in vioplot
Dear R people I have 4 questions about graphics. QUESTION 1 2 --- I have the following data, which is also set as a dataframe. Then I plot it using a violin plot 1. How can a draw a line representing 2 SD for each group? 2. How can I colour the simple violin plots in different colours (e.g., Group A white, Group B grey)? set.seed(1) a = rnorm(25, 100, 50) b = rnorm(25, 300, 50) # --- dataframing the vectors... adf - data.frame(Group = Group A , Measure = a) bdf - data.frame(Group = Group B , Measure = b) ## Combine into a dataframe using rbind abtruncdata - rbind(adf, bdf) attach(abtruncdata) require(UsingR) with(abtruncdata, simple.violinplot(Measure~Group)) with(abtruncdata, plotmeans(Measure ~ Group, pch=16, mean.labels=TRUE, col= c(blue, red), digits=2, add=T, p=0.95, barcol=c(blue, red), use.t=T, n.label=F, connect=F, ci.label=F)) QUESTION 3 --- I have the following bar plot for the same data 3. how can I set the ylim, lets say from 50 to 350 in steps of 25? tmp = split(Measure, Group) means = sapply(tmp, mean) stdev = sapply(tmp, sd) require(gplots) plotCI(barcol=blue, pch=16, col=blue, barplot(means, col=c(white, gray90), ylim=c(0,max(means + stdev)), xlab = Error bars: ±1SD, ylab=Measure), means, stdev, add=TRUE) QUESTION 4 --- The same data frame now as violin plot 4. How can I keep a violinplot frame a la simple.violinplot? ## set up frame, without axes plot(1,1,xlim=c(0.5,2.5),ylim=range(c(a,b)),type=n, xlab= ,ylab=Median measure,axes=FALSE) ## bottom axis, with user-specified labels axis(side=1,at=1:2,labels=c(Group A,Group B)) axis(side=2) # but as soon as I call the first violinplot, the upper and right axes appear again! In advance, thank you for your answers! Cheers, Fer __ 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] about granova library
Dear all Recently the granova package was launched. I installed but after when I invoked it in R it requested for other libraries. They were downloaded and install automatically. I tried to run the example syntax of granova.1w and granova.2w but two things happened: i) either a file called granova.rdb wasnt existent or ii) the GUI clashed and R shut down. Has anyone else experience this? Do the developers have an answer for this troubleshot? Im using a Windows Vista system and I have the R version 2.7.2. Cheers, Fer __ 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] dixon test
hi giov about the dixon test... i just run a simple test with a sample of 40 and I got: Error in dixon.test(x) : Sample size must be in range 3-30 So it seems that most of the test in the outliers package are designed for small samples. See also the Rnews article published in May 2006 (vol 6/2) called processing data for outliers by Lukasz Komsta (the developer of the package). However there is in that package a function called scores which works for big samples. You can also see the p-values and z scores for the observations you have and determine which values are considered outliers. Try this simple syntax: library(outliers) library(gamlss.dist) # this produces a exponential+Gaussian distribution (which usually has heaps of outliers!) x - rexGAUS(100,2000,3000,5000) # this confirms that Dixon works for samples between 3 and 30!!! dixon.test(x) # just to see what the data set looks like and visually confirm the outliers boxplot(x, notch=T) # sort the scores in ascending order sort(x) # returns probability of each score (using z scores) to be an outlier in order sort(scores(x, type=z, prob=1)) # determines which scores are considered outliers with a 95% confidence sort(scores(x, prob=0.95)) The author points regarding the prob part... prob If set, the corresponding p-values instead of scores are given. If value is set to 1, p-value are returned. Otherwise, a logical vector is formed, indicating which values are exceeding specified probability. In z and mad types, there is also possibility to set this value to zero, and then scores are confirmed to (n-1)/sqrt(n) value, according to Shiffler (1998). The iqr type does not support probabilities, but lim value can be specified. The reference of Shiffler is not as the one that appears in the help. It is this one: Schiffler, R.E (1988). Maximum Z scores and outliers. Am. Stat. 42, 1, 79-80. I hope this helps, Fernando -- View this message in context: http://www.nabble.com/dixon-test-tp18940260p18953571.html Sent from the R help mailing list archive at Nabble.com. __ 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] about the 95%CI around the median...
Dear people I've learnt that by using the boxplot.stats command in the grDevices library I can get the 5-number summaries of a boxplot, plus other important information, like the confidence interval around the median. I'm interested in knowing the actual formula to used in that package to calculate that confidence interval. Can someone help me with this? Cheers, Fernando __ 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] adding the mean and standard deviation to boxplots
Dear users This is a message I was directing to Harold Baize but because I pressed the wrong button the message got lost g!!! So I’m doing it all over again: Lets suppose I have three batches of data: a - rnorm(50,2500,300) b - rnorm(50,3500,250) c - rnorm(50,4000,200) # Now I want to plot them as boxplots and violin plots require(vioplot) vioplot (a,b,c, horizontal=T, col=“white”) boxplot (a,b,c, horizontal=T, col=“white”) As we know boxplot show the least-greates values, lower-upper quartiles, the mean, and outliers (when present). However, for some data is not important the MEDIAN but the MEAN. Also, it is more relevant to show ERROR BARS instead of quartiles. So, how could I see (for the batches of data I introduced above)… 1. a boxplot showing the MEAN and the SD instead of the lower/upper quartile? 2. a boxplot showing the MEAN and the STANDARD ERROR OF THE MEAN instead of the lower/upper quartile? 3. a boxplot showing the MEAN and the 95% CI instead of the lower/upper quartile? (I think in all these cases is preferable to have visual access, or to have the line that shows, the LEAST and the GREATEST VALUES.) In other words, that the ERROR BARS (95% CI, SD, SE) proposed here take the place of the boxes usually used to represent the lower/upper quartile. Now, the big question, is all this jazz possible to be implemented in violin plots as well? How could that be done? Cheers, Fernando -- View this message in context: http://www.nabble.com/adding-the-mean-and-standard-deviation-to-boxplots-tp15271398p18604571.html Sent from the R help mailing list archive at Nabble.com. __ 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.
