Re: [R] fExtreme packages

From: Spencer Graves <>
Date: Fri 03 Feb 2006 - 15:07:39 EST


          The problem is not you. I just tried:


          I run R under XEmacs, and this command caused my version of XEmacs to hang. After waiting 15 minutes, I was able to break out of it. However, I did not get the intended result. This is tragically typical of my personal experience with Rmetrics packages.

          The Rmetrics project is breathtaking in scope but deficient in execution. I believe it works well for Prof. Diethelp Wuertz's classes at the Swiss Federal Institute of Technology, and he is to be commended for making it available for others. Indeed, he has done a superhuman job with what he has created. However, the Rmetrics packages would be much more useful if Prof. Wuertz could find a way to recruit a team of collaborators who would prioritize needed improvements and attack them as a team, in roughly the way the R Project itself is managed. I sincerely hope that these comments are not interpreted as criticisms of Prof. Wuertz; he certainly has accomplished more than I could have. Rather, I hope only to suggest a possible vision of potentially broader, more collaborative effort.

          If you want to learn time series (and financial time series) in R, I will recommend to you the path that I have found most useful in my own stidy of time series capabilities in R:

  1. Read the chapter on time series in Venable and Ripley (2002) Modern Applied Statistics with S, 4th ed. (Springer). This book is my primary reference for almost anything R other than "help", "", "RSiteSearch" and the other free documentation. The chapter on time series is an excellent introduction to basic R time series capabilities.
  2. I'd follow this by studying carefully the "R Help Desk" article on "Date and Time Classes in R" by Gabor Grothendieck and Thomas Petzoldt in R News, vol. 4/1, June 2004.
  3. Install the "zoo" package and work through the two "vignettes" associated therewith. (Note: The "vignette" documentation is a bit sparse. If you work through the examples NOT using Emacs, you will find that you can create a "*.R" file containing the R commands in the vignette. I've found this a very powerful way to learn. For a way to do this with Emacs, try RSiteSearch for that.)
  4. Depending on your interests, you might follow this with the "dse" bundle.
  5. Subscribe to "R-sig-finance" if you don't already. Questions like this might get a better reception from this specialized list. However, I don't have adequate experience to say that. Thus, I would suggest persistence. Some questions generate a genuine "feeding frenzy" of responses, while others never get answered. To increase your chances of moving away from "unanswered" towards "feeding frenzy", I encourage you to follow the posting guide! "". Your question below suggests you may have already done so, but I felt obligated to repeat the suggestion nonetheless. If you don't get an answer in 24 hours, feel free to post a similar question. However, please reword your question to clarify and simplify the issue as much as possible.
  6. I'm currently trying to create a companion package for Ruey Tsay (2005) Analysis of Financial Time Series, 2nd ed. (Wiley). I've found this book to be quite useful for study of financial time series, though not for study of R capabilities in that area.
  7. Finally, continue doing what you already are doing. I think R has substantial capabilities for what you want, but they are not unfortunately well organized.
	  hope this helps.
	  spencer graves wrote:

> Hello,
> I am a new user of R. I am trying to use the packages fBasics and fExtremes
> when i am running the examples I get few error. Could someone tell me what is
> happenig? Thank you beforehand.
> from Fbasics packages:
> xmpfBasics()
> Error in file(file, "r") : unable to open connection
> In addition: Warning message:
> cannot open file '/usr/lib/R/library/fBasics/demoIndex'


> Error in file(file, "r", encoding = encoding) :
> unable to open connection
> In addition: Warning message:
> cannot open file 'fBasics-xmpTools'
> from fExtremes:
> data(danish)
> xmpExtremes("\nStart: Simulate a GPD Distributed Sample > ")
> x = gpdSim(model = list(shape = 0.25, location = 0, scale = 1), n = 1000)
> Error in rgpd(n = n, xi = model$shape, mu = model$location, beta =
> model$scale) : unused argument(s) (mu ...)
> fit = gpdFit(danish, threshold = 10, type = "mle")
>> par(mfrow = c(1, 1)) 
>> gpdtailPlot(fit, main = "Danish Fire: GPD Tail Estimate", col = 

> "steelblue4")
> Error in as.numeric(gpd.obj$upper.exceed) :
> argument "gpd.obj" is missing, with no default
> ______________________________________________
> mailing list
> PLEASE do read the posting guide! mailing list PLEASE do read the posting guide! Received on Fri Feb 03 15:15:26 2006

This archive was generated by hypermail 2.1.8 : Fri 03 Feb 2006 - 20:29:24 EST