Re: [R] time series processing - count of datestamp delta's, per group

Hi Martin, it sounds like you want the difference between the first and the last observation per user, not, e.g., all the date differences between successive observations of each separate user. Correct me if I'm wrong. That said, let's build some toy data: set.seed(1) dataset -

Re: [R] Citing Package Contributing Authors

Hi, it usually is a good idea to look at the output of citation() (which, however, also often is auto-generated) or at the authors listed in package vignettes. And thanks for citing R package authors. When I review papers, I often have to remind authors of this... Best Stephan On

Re: [R] Saving model and other objects from caret

Have you looked at ?save and ?load? As I already wrote here: http://stackoverflow.com/questions/14761496/saving-and-loading-a-model-in-r Best, Stephan On 07.02.2013 22:33, James Jong wrote: Say I train a model in caret, e.g.: RFmodel - train(X,Y,method='rf',trControl=myCtrl,tuneLength=1)

Re: [R] Problems with stepAIC

Hi Katja, try fitting the original model using ML (not REML) with the parameter method = ML: PModell1 -lme(sqrt(Earthwormsm.2)~ Treatment+Pflanzenfrischmasse+aBodenfeuchte+bBodenfeuchte+Gfrischmasse+Ltrockenmasseanteil+KCN+I+Eindringtiefe, random=~1|Block/Treatment/Cluster/Patch,

Re: [R] percentile of a given value: is there a reverse quantile function?

?ecdf Best, Stephan On 03.03.2012 13:37, drflxms wrote: Dear all, I am familiar with obtaining the value corresponding to a chosen probability via the quantile function. Now I am facing the opposite problem I have a value an want to know it's corresponding percentile in the distribution. So

Re: [R] Plot many x and y

Hi Alex, could you be a little more specific as to what exactly you mean by plotting many x's and y's with one legend per plot? Please note what appears at the bottom of every R-help mail: PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented,

Re: [R] Force the for loop to stop

Hi, you could set a dummy variable to FALSE outside the outermost loop. If the break condition is met in the inner loop, set the dummy variable to TRUE before breaking and test its truth status in the outer loop. HTH Stephan Am 01.06.2011 21:25, schrieb Salih Tuna: Hi, I am looking for a

Re: [R] Force the for loop to stop

Hi Salih, here you go: dummy - FALSE for ( ii in 1:5 ) { for ( jj in 3:6 ) { cat(ii=,ii,; jj=,jj,\n,sep=) if ( ii == jj ) { dummy - TRUE break } } if ( dummy ) break } ### Note

Re: [R] Problems Dating....

Hi Nat, I guess something like as.Date(as.character(3/4/2007),format=%d/%m/%Y) should work - as.character() coerces the factors to characters, which the as.Date() function can work with, given the right format argument. HTH Stephan Am 01.06.2011 22:59, schrieb Struckmeier, Nathanael: I'm

Re: [R] Applying a function to a subset

Or just include is.na=TRUE in the definition of kurtosis(): kurtosis-function(x) { m4-sum((x-mean(x,na.rm=TRUE))^4,na.rm=TRUE)/length(x) s4-var(x,na.rm=TRUE)^2 m4/s4 - 3 } HTH Stephan Am 29.05.2011 11:34, schrieb Jim Holtman: kurtosis(fem[!is.na(fem)]) Sent from my iPad On May 29,

Re: [R] What are the common Standard Statistical methods used for the analysis of a dataset

Dear all, may I suggest the acronym IOTT for the inter-ocular trauma test? Now we just need someone to implement iot.test(). I assume it will appear on CRAN within the next 24 hours. Looking forward to yet another base package, Stephan Am 25.05.2011 23:36, schrieb Greg Snow: How can

Re: [R] minimum distance between line segments

Hi, this sounds like a standard problem in Computational Geometry - I guess game developers have to deal with something like this all the time. You may want to look at a textbook or two. An article with the promising title On fast computation of distance between line segments can be found

Re: [R] Anyone know a forum for stats advice?

Hi Clare, you want to go here: http://stats.stackexchange.com/questions HTH Stephan Am 04.03.2011 12:08, schrieb Clare Embling: Hi, I know this forum is for R-related issues, but the question I have is a statistical question I was wondering if anyone could recommend a good statistics

Re: [R] Comparing decimal numbers

Hi, this is R FAQ 7.31. http://cran.r-project.org/doc/FAQ/R-FAQ.html HTH, Stephan Am 01.02.2011 14:49, schrieb mlancee: Hi, I have a seemingly easy question that has been keeping be busy for quite a while. The problem is the following: 0.1 + 0.1 + 0.1 == 0.3 [1] FALSE Why is this false?

Re: [R] Uniform Distribution

Hi Alex, help.search(uniform) HTH, Stephan Am 08.09.2010 15:36, schrieb Alaios: Hello, I would like to uniformly distribute values from 0 to 200. Can someone help me find the appropriate uniform distribution generator? I would like to thank you in advance for your help. Best Regards Alex

Re: [R] repeated measurements ANOVA

Hi Alex, I'm slightly unclear as to why you would want to restructure your nice six-column data.frame (why six? One column for the data and four for the factors should make five, shouldn't it? I guess you have a subject ID in one column?) into some monstrosity which I assume you would fill

Re: [R] R time series analysis

Hi, basically, you know 5 periods later. If you use a good error measure, that is. I am a big believer in AIC for model selection. I believe that arima() also gives you the AIC of a fitted model, or try AIC(arima1). Other ideas include keeping a holdout sample or some such. I'd recommend

[R] predict.lm, matrix in formula and newdata

Dear all, I am stumped at what should be a painfully easy task: predicting from an lm object. A toy example would be this: XX - matrix(runif(8),ncol=2) yy - runif(4) model - lm(yy~XX) XX.pred - data.frame(matrix(runif(6),ncol=2)) colnames(XX.pred) - c(XX1,XX2) predict(model,newdata=XX.pred) I

Re: [R] Problems with normality req. for ANOVA

Hi, simulating would still require you to operationalize the lack of normality. Are the tails too heavy? Is the distribution skewed? Does it have multiple peaks? I suspect that the specific choices you would make here would *strongly* influence the result. My condolences on the client you

Re: [R] Help with a problem

for the help! Michael Stephan Kolassa 07/17/10 4:50 PM Mike, I am slightly unclear on what you want to do. Do you want to check rows 1 and 7 or 1 *to* 7? Should c1 be at least 100 for *any one* or *all* rows you are looking at, and same for c2? You can sort your data like this: data - data

Re: [R] Help with a problem

Mike, I am slightly unclear on what you want to do. Do you want to check rows 1 and 7 or 1 *to* 7? Should c1 be at least 100 for *any one* or *all* rows you are looking at, and same for c2? You can sort your data like this: data - data[order(data$ds),] Type ?order for help. But also do this

Re: [R] Can anybody help me understand AIC and BIC and devise a new metric?

Hi, one comment: Claeskens and Hjort define AIC as 2*log L - 2*p for a model with likelihood L and p parameters; consequently, they look for models with *maximum* AIC in model selection and averaging. This differs from the vast majority of authors (and R), who define AIC as -2*log L + 2*p and

Re: [R] how to define a function in R

Hi, I recommend that you look at the following help pages and experiment a little (maybe create a toy directory with only three or four files with a few lines each): ?files ?dir ?grep ?strsplit Good luck! Stephan jd6688 schrieb: Here are what i am going to accomplish: I have 400 files

Re: [R] process of stepwise selection

Hi Elaine, in general, stepwise selection is a very bad idea: Whittingham, M. J.; Stephens, P. A.; Bradbury, R. B. Freckleton, R. P. Why do we still use stepwise modelling in ecology and behaviour? Journal of Animal Ecology, 2006, 75, 1182-1189 HTH Stephan elaine kuo schrieb: Dear list,

Re: [R] Finding Lower Envelope of Points on a Plot?

Hi, one possibility would be to calculate the convex hull using chull(). I believe that the hull points are returned by chull() in a clockwise order (?), so the points between the rightmost and the leftmost point in the chull() result are the lower half of the convex hull. Remove these

Re: [R] Finding Lower Envelope of Points on a Plot?

the percentage of points outside it? Thanks! Asha Stephan Kolassa wrote: Hi, one possibility would be to calculate the convex hull using chull(). I believe that the hull points are returned by chull() in a clockwise order (?), so the points between the rightmost and the leftmost point

Re: [R] What is wrong with this code?

Hi, your problem is called string matching. Search for that term on rseek.org, there are a couple of functions and packages. And Wikipedia can tell you everything you ever wanted to know about string matching (and more). HTH, Stephan Cable, Samuel B Civ USAF AFMC AFRL/RVBXI schrieb:

Re: [R] Exponential Smoothing: Forecast package

this. And I did find a way around could this allow me to set MAPE as a criteria? Phani On Tue, Jun 29, 2010 at 12:47 AM, Stephan Kolassa stephan.kola...@gmx.dewrote: Hi Phani, to get the best Holt's model, I would simply wrap a suitable function calling ets() within optim() and optimize for alpha

Re: [R] Exponential Smoothing: Forecast package

Hi Phani, to get the best Holt's model, I would simply wrap a suitable function calling ets() within optim() and optimize for alpha and beta - the values given by ets() without constraints would probably be good starting values, but you had better start the optimization with a variety of

Re: [R] Reading file

Have you set the correct working directory? ?setwd ?getwd HTH Stephan Robert Tsutakawa schrieb: I am trying to read a source program into a mac pro laptop, which uses Snow Leopard. R is unable to find the file containing my source program. I'm using the function source( file name). I

Re: [R] create new variable: percentile value of variable in data frame

Hi Jon, does the empirical cumulative distribution function do what you want? dat$q.score - ecdf(dat$score)(dat$score) ?ecdf HTH Stephan Jonathan Beard schrieb: Hello all, Thanks in advance for you attention. I would like to generate a third value that represents the quantile value of a

Re: [R] shapiro.test

Hi, David Winsemius schrieb: snip This would imply that ozon is a list or dataframe. snip And you tried to give the whole list to a function that only wants a vector. And whenever you suspect that your data types clash, try str() to find out just what kind of thing your data is. Here:

Re: [R] adding column to data frame conditionally

Hi Kristina, Thierry's solution is certainly the correct one in terms of keeping within R's philosophy... but I personally find a series of conditional assignments easier to understand - see below for an example. HTH, Stephan # # Example

Re: [R] glm.diag

Hi Helga, did you load the boot library, which contains glm.diag(), by calling library(boot)? HTH Stephan Helga margrete holmestad schrieb: I have made a poisson regression by using glm. Then I want to check if I have over-dispersion in my model. I tried to use glm.diag(fit.1), but then R

Re: [R] From date to week day

Or: weekdays(as.Date(2010-05-24)) HTH, Stephan Wu Gong schrieb: ?strptime will helps. d - as.Date(01/05/2007,%m/%d/%Y) format(d, %A, %b %d, %Y) [1] Friday, Jan 05, 2007 - A R learner. __ R-help@r-project.org mailing list

Re: [R] function rep

Hi, this is FAQ 7.31: pb and pr are floating-point numbers that are coerced to integer for rep(), and this does not always work the way you want. HTH Stephan Covelli Paolo schrieb: Hi, I've got the following code: p - 0.34 pb - p*100 pr - (1-p)*100 A - rep(0,pb) # a vector with 34

Re: [R] help in histogram

Hi, first get the densities without plotting the histogram: foo - hist(x, plot=FALSE) then plot the histogram and feed the rounded densities, converted to character, to the labels argument (instead of just labels=TRUE): hist(x, freq=F, xlab='',ylab=Percent of Total, col=skyblue,

Re: [R] interpretation of p values for highly correlated logistic analysis

Hi Claus, welcome to the wonderful world of collinearity (or multicollinearity, as some call it)! You have a near linear relationship between some of your predictors, which can (and in your case does) lead to extreme parameter estimates, which in some cases almost cancel out (a coefficient of

Re: [R] reshaping data

Hi Kim, look at the reshape() command with direction=wide. Or at the reshape package. HTH, Stephan Kim Jung Hwa schrieb: Hi All, Can someone help me reshape following data: Var1 Var2 Val A X 1 A Y 2 A Z 3 B X 4 B Y 5 B Z 6 to some kind of matrix/tabular format (preferably as a matrix),

Re: [R] ODD and EVEN numbers

(foo %% 2) == 0 See ?%% HTH Stephan tj schrieb: Hi, anyone here who knows how to determine if an integer is odd or even in R? Thanks. tj __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the

Re: [R] Summing data based on certain conditions

?by may also be helpful. Stephan Steve Murray schrieb: Dear all, I have a dataset of 1073 rows, the first 15 which look as follows: data[1:15,] date year month day rammday thmmday 1 3/8/1988 1988 3 81.430.94 2 3/15/1988 1988 3 152.860.66 3 3/22/1988

Re: [R] Finding common an unique elements in character vectors

Hi Thomas, %in% does the trick: vector_1 - c(Belgium, Spain, Greece, Ireland, Luxembourg, Netherlands,Portugal) vector_2 - c(Denmark, Luxembourg) vector_1[!(vector_1 %in% vector_2)] HTH, Stephan Thomas Jensen schrieb: Dear R-list, I have a problem which I think is quite basic, but so far

Re: [R] Creating dataframe of all possible variable combinations

Hi Mike, the following works for me: SITE - ordered(c(101,102,103,104)) WDAY - ordered(c(MON,TUE,WED,THR,FRI),levels=c(MON,TUE,WED,THR,FRI)) TOD - ordered(c(MORN,AFTN),levels=c(MORN,AFTN)) foo - expand.grid(SITE=SITE,WDAY=WDAY,TOD=TOD) foo[order(foo$SITE),] If this doesn't solve your

Re: [R] Arima forecasting

Hi Matteo, just use forecast.Arima() with h=2 to get forecasts up to 2 steps ahead. R will automatically use forecast.Arima() if you call forecast() with an Arima object. library(forecast) model - auto.arima(AirPassengers) forecast(model,h=2) HTH, Stephan Matteo Bertini schrieb: Hello

Re: [R] for() loop

Hi Martin, it is slightly unclear to me what you are trying to achieve... are you trying to tabulate how often each value appears in datjan[,4]? Then table(datjan[,4]) may be what you want. HTH Stephan Schmidt Martin schrieb: Hello I'm working with R since a few month and have still many

Re: [R] Is this a bug (or a feature) in hist(x)$density ??

Hi Tal, basically, by summing over the (pointwise) density, you are approximating the integral over the density (which should be around 1) - but to really do a rectangular approximation, you will of course need to multiply each function value by the width of the corresponding rectangle. I'd

Re: [R] Making multiple columns to a single column

Hi Hyo, how about as.vector(ttx1) or as.vector(t(ttx1))? HTH Stephan Hyo Lee schrieb: Hi guys, I have a very simple question. I'm trying to make multiple columns to a single column. For example, *ttx1* is a 46*72 matrix. so, I tried this. *d1=ttx1[,1] d2=ttx1[,2] ... d72=ttx1[,72]*

Re: [R] Randomly sampling subsets of dataframe variable

Hi Mike, take an index vector that selects Monday and Tuesday out of each week, and then run a restricted random permutation on this vector which only permutes indices within each week. rperm() is in the sna package. library(sna) foo - rep(c(TRUE,TRUE,FALSE,FALSE,FALSE),26)

Re: [R] Shade area under curve

Hi, use dnorm() for the density and polygon() to shade the area underneath, with suitably many x values so your density looks smooth. HTH, Stephan claytonmccandless schrieb: I want to shade the area under the curve of the standard normal density. Specifically color to the left of -2 and

Re: [R] data.frame question

Hi, data.frame(x=x,y=as.numeric(x%in%y)) HTH, Stephan joseph schrieb: hello can you show me how to create a data.frame from two factors x and y. column 1 should be equal to x and column 2 is 1 if it is common to y and 0 if it is not. x=factor(c(A,B,C,D,E,F,G)) y=factor(c(B,C,G)) the

Re: [R] Equation for model generated by auto.arima

Hi, In that case, I'd recommend reading a good book on time series analysis. Forecasting: Methods and Applications by Makridakis, Wheelwright and Hyndman is very accessible. Alternatively, there are probably tons of webpages on ARIMA, so google around. Best, Stephan testuser schrieb:

Re: [R] Equation for model generated by auto.arima

Hi, the help page for arima() suggests looking at predict.Arima(), so take a look at ?predict.Arima(). You will probably not use the coefficients, but just feed it the output from arima(). And take a look at auto.arima() in the forecast package. HTH Stephan testuser schrieb: I would like

Re: [R] Exporting Graphs

Hi Karthik, I think you will need to do something like jpeg(histograms.jpg) hist(rnorm(100)) dev.off() HTH Stephan Karthik schrieb: Hello Tal, This is the code. hist(rnorm(100)) jpeg(histogram.jpeg) --- Even when I decrease the quality, I

Re: [R] GARCH AIC

Hi David, str(g) gives you a ton of output, and the @fit slot has a $ics component, part of which has the promising name of AIC... (g...@fit)$ics[1] HTH, Stephan David Rubins schrieb: Hi, Is there anyway to extract the AIC and BIC from the summary statistics in fGarch package? g -

Re: [R] Comparing dates in dataframes

Hi, it looks like when you read in your data.frames, you didn't tell R to expect dates, so it treats the Date columns as factors. Judicious use of something along these lines before doing your comparisons may help: arr$Date - as.Date(as.character(arr$Date),format=something) Then again, it

Re: [R] bootstrapping

Hi Aaron, try the argument statistic=mean. Then boot() will give you the mean turn angle in your actual data (which appears to be 6 degrees, judging from what you write), as well as the means of the bootstrapped data. Then you can get (nonparametric) bootstrap CIs by

Re: [R] = returns wrong result? Why

Hi Trafim, take a look at FAQ 7.31. HTH Stephan Trafim Vanishek schrieb: Dear all, Does anybody know the probable reason why = gives false when it should give true? These two variables are of the same type, and everything works in the cycle but then it stops when they are equal. this is

Re: [R] Operating on each row of data frame

Hi, does this do what you want? d - cbind(d,apply(d[,c(2,3,4)],1,mean),apply(d[,c(2,3,4)],1,sd)) HTH, Stephan Abhishek Pratap schrieb: Hi All I have a data frame in which there are 4 columns . Column 1 : name Column 2-4 : values I would like to calculate mean/Standard error of values

[R] Reading a file with mixed cyrillic/latin characters

Dear useRs, I am trying to read a tab-delimited Unicode text file containing both latin and cyrillic characters and failing miserably. The file looks like this (I hope it comes across right): A B C 3 foo ФОО 5 bar БАР read.table(foo.txt,sep=\t,header=TRUE) I

Re: [R] Beginer data.frame

Hi Jean-Baptiste, two points: 1) Your variable df is a *local* variable which you define in your function myfunc(), so it is not known outside myfunc(). When you ask is.data.frame(df), R looks at the global definition of df - which is the density function of the F distribution. To make your

Re: [R] coerce vector into array - change filling sequence

Hi, you can permute array dimensions using aperm(): x - 1 : 24 z - array(x, dim=c(6,2,2)) y - aperm(z,perm=c(3,2,1)) y[1,1,] HTH, Stephan Kohleth Chia schrieb: Dear all, When I coerce a vector into a multi dimensional array, I would like R to start filling the array along the last

Re: [R] Question regarding if statement in while loop

Hi Stephanie, it sounds like R's exception handling may help, something like this: foo - try(eblest(i, dir5, sterr5, weight5, aux5)) if ( class(foo) == try-error ) next Take a look at ?try. HTH, Stephan Stephanie Coffey schrieb: Hi all, I'm running R version 2.9.2 on a PC. I'm having a

[R] boot() with an array-valued statistic?

Dear guRus, is there a version of boot() that deals with array data and array-valued statistics? For example: foo - array(rnorm(120),dim=c(3,5,8)) means - apply(foo, MARGIN=c(2,3), FUN=mean) means contains the means over the first dimension of foo: a 5x8 array. Now I would like to bootstrap

Re: [R] plotting

Hi teo, try lines() instead of points(). HTH Stephan teo schrieb: Hi: Could you please guide me how to plot weekly.t2. Below are my values and I am only able to plot weekly t1 but failing to add a second line representng weekly t2. Thanks. weekly.t1 [1] 228.5204 326.1224 387.2449 415.4082

Re: [R] Generating permutations that always include one specific element

Hi Ray, First possibility: just select those combinations that contain AL: combos.with.AL - possible.combos[rowSums(possible.combos==AL)0,] Second possibility: create all 3-combos *without* AL: bands.without.AL - c(B, DB, DG, G, K, LB, LG, MG, O, P, PI, PK, PU, R, V, W, Y)

Re: [R] integer(0) and NA do not equal FALSE

Hi Jonathan, grep() returns a vector giving either the indices of the elements of 'x' that yielded a match or, if 'value' is 'TRUE', the matched elements of 'x' (quoting from the help page, see ?grep). So you probably want to test whether this vector is empty or not - in other words,

Re: [R] forecasting

?predict HTH Stephan DispersionMap schrieb: I have some data that i ran a regression on and got the usual r output, with the a intercept and b coefficient for my independent variable. i want to forecast the number of future events using these parameters. What function / packages in R would

Re: [R] Literature analysis

Hi, from what I understand, you may be interested in text mining, so perhaps you want to look at the tm package. Then again, depending on what you are really trying to do, you may be better served with perl, awk and similar tools than with R... HTH, Stephan Schwan schrieb: Dear all, i

Re: [R] R Packages Crack the 3,000 Mark!

Hi Bob, Muenchen, Robert A (Bob) wrote: Does anyone have a program that graphs the growth of R packages? I don't know if that historical data is around. John Fox had a slide on this in his useR 2008 talk The Social Organization of the R Project (page 7), with package counts up to March

Re: [R] modeling and forecasting commodity time series?

Hi Luna, you may want to look at the IIF website, http://www.forecasters.org They have a mailing list for forecasters - you may get more of a response there than on a dedicated R list. HTH, Stephan Luna Moon schrieb: Hi all, Could anybody please shed some lights on me about good

Re: [R] exclude data for boxplot stats using mathematical operator

Try this: boxplot.stats(x[x[,2]0,2],do.conf=FALSE) HTH, Stephan e-letter schrieb: Readers, I have a data set as follows: 1,1 2,2 3,3 4,4 5,3 6,2 7,-10 8,-9 9,-3 10,2 11,3 12,4 13,5 14,4 15,3 16,2 17,1 I entered this data set using the command 'read.csv'. I want to exclude values fewer than

Re: [R] State Space models in R

Hi, Rob Hyndman's forecast package does exponential smoothing forecasting based on state space models (and lots of other stuff, ARIMA et al.). It's not exactly the companion package to his book, but it comes close. The book's (Forecasting with Exponential Smoothing - The State Space

[R] ggplot2: deterministic position_jitter geom_line with position_jitter

Dear guRus, I am starting to work with the ggplot2 package and have two very dumb questions: 1) deterministic position_jitter - the jittering is stochastic; is there any way to get a deterministic jittering? For instance: example.data -

Re: [R] Normalize data set

Are you looking for reshape()? HTH, Stephan Edward Chen schrieb: Hi all, I have a mxn matrix that consists of 28077 rows of features and 30 columns of samples. I want to normalize each row for the samples for each feature. I have tried normalize and scale functions but they don't seem to

Re: [R] plotting the graph of density with an unknown distribution

Hi, Try kernel smoothing via the density() function. And take a look at ecdf(). HTH, Stephan sendona essile schrieb: How can I plot the graph of a density of a sample with an unknown distribution? I can provide any sample size which is required. I want to have a smooth density graph of my

Re: [R] splitting a string up

strsplit(1041281__2009_08_20_.lev, split=_)[[1]][1] HTH, Stephan stephen sefick schrieb: x - 1041281__2009_08_20_.lev I would like to split this string up and only extract the leading numbers. 1041281 to use as a label for a data column in a bigger for loop function to read in data.

Re: [R] Several simple but hard tasks to do with R

Hi Rakknar, I believe that the menu command File - Save to File (or some such, I use the German version) in the R GUI for Windows (I'm unclear on your OS) has not yet been suggested. This writes a file containing the entire R session. Does this create the kind of log you are looking for?

Re: [R] Why are there small circles in my plot

Hi, Mao Jianfeng schrieb: plot(c(min(myda$traits),max(myda$traits)),c(-0.03,0.5), xlab='State', ylab='ylab') Here, you are plotting two data points: (min(myda$traits),-0.03) and (max(myda$traits),0.5). Try this: plot(c(min(myda$traits),max(myda$traits)),c(-0.03,0.5), xlab='State',

Re: [R] Collinearity in Linear Multiple Regression

Hi Alex, I personally have had more success with the (more complicated) collinearity diagnostics proposed by Belsley, Kuh Welsch in their book Regression Diagnostics than with Variance Inflation Factors. See also: Belsley, D. A. A Guide to using the collinearity diagnostics. Computational

Re: [R] Collinearity in Linear Multiple Regression

Hi Tim, Variance proportions (and condition indices) are exactly the tools described in Belsley, Kuh Welsch, Regression Diagnostics - see my previous post. Good to see I'm not the only one to use them! BKW also describe in detail how to calculate all this using SVD, so you don't need to use

Re: [R] quadratic programming

Hi, The CRAN Task View on Optimization may help: http://stat.ethz.ch/CRAN/web/views/Optimization.html HTH, Stephan barbara.r...@uniroma1.it schrieb: Devo risolvere un problema di minimo vincolato con vincoli di uguaglianza e un altro con vincoli di uguaglianza e disuguaglianza. Cosa posso

Re: [R] Deleting rows or cols that do not meet cut off

f[rowSums(f=1)0,colSums(f=1)0] Judging from your result, you want less than or equal to 1. HTH, Stephan Crosby, Jacy R schrieb: How can I delete both rows and columns that do not meet a particular cut off value. Example: d - rbind(c(0,1,6,4), + c(2,5, 7,5), + c(3,

Re: [R] Deleting rows or cols that do not meet cut off

Hi Wacek, Wacek Kusnierczyk schrieb: ... or gain a bit on performance by doing the threshold comparison on the whole matrix just once at once: dd = d = 1 d[apply(d, 1, any), apply(d, 2, any)] d[apply(dd, 1, any), apply(dd, 2, any)] Or not? Cheers, Stephan

Re: [R] bug when subtracting decimals?

Hi, Gavin Simpson wrote: I bemoan the apparent inability of those asking such questions to use the resources provided to solve these problems for themselves... Looking at all the people who quite obviously do NOT read the posting guide and provide commented, minimal, self-contained,

Re: [R] Citing R

citation() HTH, Stephan Tom Backer Johnsen schrieb: What is the correct citation to R in BibTeX format? I have looked in the R pages but so far without any luck. Tom __ R-help@r-project.org mailing list

Re: [R] Time series forecasting

Hi Perry, my impression after a very cursory glance: this looks like noise. Perhaps you should think a little more about your series - what kind of seasonality there could be (is this weekly data? or monthly?), whether the peaks and troughs could be due to some kind of external driver,

Re: [R] How to generate natural cubic spline in R?

Hi, if you are looking for *natural* cubic splines (linear beyond the outer knots), you could use rcs() in Frank Harrell's Design package. HTH, Stephan David Winsemius schrieb: If one enters: ??spline ... You get quite a few matches. The one in the stats functions that probably answers

Re: [R] Does R support double-exponential smoothing?

Hi, ets() in Hyndman's forecast package allows you to specify which one of the many smoothing variants (additive/multiplicative season, damped trend, additive/multiplicative errors) you want. HTH, Stephan minben schrieb: I want to use double-exponential smoothing to forecast time series

Re: [R] model selection using ANOVA

Hi Alina, your approach sounds problematic - you can always get a smaller RSS if you add terms to your model, so your approach will always go for larger models, and you will end up overfitting. Consider information criteria, e.g., AIC or BIC, which penalize larger models. References for AIC

Re: [R] How to generate natural cubic spline in R?

Hi David, David Winsemius schrieb: The splinefun documentation indicates that natural is one of the types of cubic spline options available. That sounds good, didn't know that... rcs() has the advantage of coming with a book (Harrell's Regression Modeling Strategies). Does rcs actually do

Re: [R] add variable in for loop

Hi Paul, you do *not* want to do this, it takes too long and may lead to rounding errors. Vectorize everything, e.g., use sum(meanrotation). And look into ?apply, and google for the R Inferno. And no, there is no +=... Good luck! Stephan pgseye schrieb: Hi, I'm learning to write some

Re: [R] text matching and substitution

Hi Simeon, ?gsub HTH, Stephan simeon duckworth schrieb: I am trying to simplify a text variable by matching and replacing it with a string in another vector so for example in colours - paste(letters,colours(),stuff,LETTERS) find and replace with (red,blue,green,gray,yellow,other) -

Re: [R] text matching and substitution

) } On Sat, Mar 28, 2009 at 9:45 AM, Stephan Kolassa stephan.kola...@gmx.dewrote: Hi Simeon, ?gsub HTH, Stephan simeon duckworth schrieb: I am trying to simplify a text variable by matching and replacing it with a string in another vector so for example in colours - paste(letters,colours

Re: [R] text matching and substitution

red xx xxx xx xx xxx xxx xx blue xx xx xx xx x x xx xx xx xx red red xx xx xx xx xx xx xx xx xx xx xx xx x x x x which i'd like to replace with red blue red other other thanks On Sat, Mar 28, 2009 at 2:38 PM, Stephan Kolassa stephan.kola...@gmx.dewrote: Hi Simeon, I'm

Re: [R] ApEn (Approximate Entropy), Total Corr, Information Interaction

Hi Vishal, re 1]: Ben Bolker very kindly shared an R reimplementation of Kaplan's Matlab code a little while ago: http://www.nabble.com/Approximate-Entropy--to21144062.html#a21149402 Best wishes Stephan Vishal Belsare schrieb: Is there any existing implementation in R/S of : 1] Pincus

Re: [R] using a noisy variable in regression (not an R question)

Hi Juliet, Juliet Hannah schrieb: One simple thing to try would be to form categories Simple but problematic. Frank Harrell put together a wonderful page detailing all the issues with categorizing continuous data: http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/CatContinuous So:

Re: [R] using a noisy variable in regression (not an R question)

Hi Juliet, Juliet Hannah schrieb: I should have emphasized, I do not intend to categorize -- mainly because of all the discussions I have seen on R-help arguing against this. Sorry that we all jumped on this ;-) I just thought it would be problematic to include the variable by itself. Take

Re: [R] How to comment in R

Hi Mihai, one (very bad style) way would be if (FALSE) { comment comment comment } But putting a # in front of every line is easier to spot in the code. HTH, Stephan mihai.mira...@bafin.de schrieb: Hi everybody, I use for the moment # at the begining of each line for comments. Is

Re: [R] R system command on windows

?shell HTH, Stephan Aurelie Labbe, Dr. schrieb: Hi, I am trying to use the R command system under windows (XP). If I try the simple command system(mkdir toto) to create a directory toto, it tells me that it cannot find the command mkdir... Does anybody knows how it works ? Is it a path

Re: [R] subset problem (reducing from six to two levels)

Hi, does this help? http://www.nabble.com/factor-question-to18638814.html#a18638814 HTH, Stephan Ine schrieb: Hi all, I have got a seemingly simple problem (I am an R starter) with subsetting my data set, but cannot figure out the solution: I want to subset a data set from six to two

  1   2   >