similar to: problems in resampling time series, block length, trend.test

Hi, I have a time series (say "x") of 13 years showing an evident increase. I want to exclude two observations (the fourth and 10th), so I do: > trend.test(x[-c(4,10)]) where: > x[-c(4,10)] [1] 7 37 79 72 197 385 636 705 700 1500 1900 and I get: Spearman's rank correlation rho data: x[-c(4, 10)] and time(x[-c(4, 10)]) S = 4, p-value < 2.2e-16

My goal is to remove signals trend without any a-priori knowledge of the trend type, if any. Some signals are very noisy and non-stationary (example attached). I have experimented with a number of techniques. Staring at the results, I can hardly tell which method is best. I am attaching the result of function "local.trend" as I cannot understand it. Nor I can make a sense of the returned

The editorial in Volume 1/2 of R News requests that we cite articles in R News by using the ISSN 1609-3631 and the URL http://cran.R-project.org/doc/Rnews/. Could someone provide a sample BibTeX entry, say for Brian Ripley's article on "Installing R under Windows" from that issue? I am using the natbib package, if that matters. I would start with @Article{R:Ripley:2001, author

Dear R-users, tsboot fails when I try to perform a block bootstrap on seaKen (package wq): these commands: require(wq) require(datasets) boot.block.sen <- function(data){seaKen(data)[[1]]} tsboot(sunspot.month, boot.block.sen, R=1999, l=12, sim="fixed") return: Error in seaKen(data) : x must be a 'ts' Any suggestion on how might I change seaKen in order to use it with

Dear Lillian, I tried to estimate parameters for time series regression using time series bootstrapping as described on page 434 in Davison & Hinkley (1997) - bootstrap methods and their application. This approach is based on an AR process (ARIMA model) with a regression term (compare also with page 414 in Venable & Ripley (2002) - modern applied statistics with S) I rewrote the code

Fritz, I have slightly adapted (didn't work before) "tsboot" from the "boot" library to the current time series conventions of R. The following patch will do that. I suggest to apply this patch to the file "boot/R/bootfuns.q" of the "boot" library at CRAN. best Adrian --- bootfuns.orig.q Thu Dec 9 10:07:23 1999 +++ bootfuns.q Thu Dec 9 10:06:51 1999

Fritz, I have slightly adapted (didn't work before) "tsboot" from the "boot" library to the current time series conventions of R. The following patch will do that. I suggest to apply this patch to the file "boot/R/bootfuns.q" of the "boot" library at CRAN. best Adrian --- bootfuns.orig.q Thu Dec 9 10:07:23 1999 +++ bootfuns.q Thu Dec 9 10:06:51 1999

I'm trying to bootstrap some regression coefficients so that I can estimate confidence intervals, but boot is not producing results. Can anybody suggest what I'm doing wrong here? > SpecPress <- ts(rnorm(501)) > Returns <- ts(rnorm(501)) > BootData <- data.frame(cbind(SpecPress, Returns)) > boot.specpress <- function(data, indices, maxit=20){ + data <-

Hello everybody, when I create an object of class boot with the function tsboot() from the package boot and try to compute several types of confidence intervals with boot.ci("object of class boot created with tsboot") I obtain the warning message, that "BCa-intervals are not defined for time-series bootstraps". Does that hold in general? Or is it just not defined in

Hi, It would be nice if we can get ISSN for wiki.centos.org. Any plans? I know assigning ISSN for blogs is being held now. If we have ISSN it might help to get more people to contribute. Just the hint. David

Hi I just installed the pastecs package and I am wondering: is there an english (or german) translation of the file pastecs.pdf? If not, is there an explanation somewhere of the object of type 'turnpoints' as a result of turnpoints(), especially the "info" field? Thanks, Rainer -- NEW EMAIL ADDRESS AND ADDRESS: Rainer.Krug at uct.ac.za RKrug at sun.ac.za WILL BE

Hello, I'm trying to get a few turnpoints for a financial time series. There is a function in pastecs that does that. However, I get a large number of turnpoints: library(pastecs) data(EuStockMarkets) dax <-EuStockMarkets[,1] plot(dax) turnp <-turnpoints(dax) summary(turnp) #gives 925 peaks/pits!!! How can specify to get only 30 turnpoints? Second question: the extract function

Does anyone know how p-values can be generated if tsboot (stationary bootstrap) for time series is performed? That would be of great help. Thanks a lot for your comments. Markus [[alternative HTML version deleted]]

Hi, I have a question to post. But I have a problem finding the appropriate forum. Some of the FAQ on pastecs are answered by R-help@stat.math.ethz.chmailing list. I searched to see if this mailing list requires registration or not. There was no information about that. I am a registered member of R- help general mailing list but I am aware it is better to post my question to a specific group in

Sorry for entering late in the thread, but I was absent last week. I am developping an alternate GUI for R (and also Splus, Octave, Scilab, Matlab, Ox & Mathematica). The project started 3 years ago, but it is progressing slowly, since I am currently the only developer on the project and had to do it during my extra time. Now, it is going a little faster because I got some time and money

In function daystoyears in package pastecs, I get this (wrong?) result with 1995: > daystoyears(1,datemin="1/1/1997",dateformat="m/d/Y") 1997.001 > daystoyears(1,datemin="1/1/1995",dateformat="m/d/Y") 1994.999 Any insights? Thanks Angel

Hi R-user, I was trying to make a raster map with WGS84 projection in R, but I could not make it. I found one data set in Google that data is almost the same format as of mine. I wanted to make a raster map of temperature with 1 degree spatial resolution for the global scale. I could make it in GIS software but I do have many variables (to be many raster images) and ultimately I am importing them

Dear list users How is it possible to visualise both a linear trend line and a quadratic trend line on a plot of two variables? Here my almost working exsample. data(Duncan) attach(Duncan) plot(prestige ~ income) abline(lm(prestige ~ income), col=2, lwd=2) Now I would like to add yet another trend line, but this time a quadratic one. So I have two trend lines. One linear trend line

Hi R-User I was trying to make a raster map with WGS84 projection in R, but I could not make it. I found one data set in Google that data is almost the same format as of mine. I wanted to make a raster map of temperature with 1 degree spatial resolution for the global scale. I could make it in the GIS software but I do have many variables (to be many raster images) and ultimately I am

Hi, I'm trying to analyse some time series data on dissolved organic nitrogen. Because it has gaps in it, I try to interpolate (linear) with the regul method from the pastecs package. I have a number of stations with measurement series in a matrix constructed from a data frame with the unstack method (temp) I also used the unstack method to make a similar matrix for the time points