R help - Mar 2009 - Fw: Fitting GUMBEL Distribution - CDF function and P P Plot

Dera R Helpers,

I am re-posting my query.

Please guide me.

Maithili


--- On Fri, 3/13/09, Maithili Shiva <maithili_shiva at yahoo.com> wrote:

 I am trying to fit the Gumbel distribution to a data. I am
 using lmom package. I am getting problem in Cumulative
 Distribution Function of Gumbel distribution as I am getting
 it as a series of 0's and 1's thereby affecting the
 P P Plot. My R code is as follows.
 
 
  library(quantreg)
  library(RODBC)
  library(MASS)
  library(actuar)
  library(lmom)
 
  x <-
 
c(986.78,1067.76,1046.47,1034.71,1004.53,1007.89,964.94,1060.24,1188.07,1085.63,988.33,972.71,1177.71,972.48,1203.20,1047.27,1062.95,1113.65,995.97,1093.98)
 
 
 #Estimating the parameters for GUMBEL distribution
 
  N                    <- length(x)
 
  lmom 		    <- samlmu(x); lmom
 
  parameters_of_GUMBEL <- pelgum(lmom); parameters_of_GUMBEL
  
 
  # Parameters are xi =  1019.4003   alpha =   59.5327
 
 
 
  # _____________________ P - P Plot ________________________
  
  e        <- c(1:N)
 
  f        <- c((e-.5)/N)
 
  Fx       <- cdfgum(x, para = parameters_of_GUMBEL)
  
  g        <- sort(Fx)
  
  png(filename = "GUMBEL_P-P.png")
  
  a <- par('bg'= "#FFFFCC")
  
  plot (f,g,bg=a,fg= "#804000",main ="P-P
 Plot", ylab=" Cumulative Distribution
 Function", xlab="i", font.main=2,
 cex.main=1,col="#999966",bty 
"o",col.main="black",col.axis="black",col.lab
 ="black")
 
  abline(rq(g ~ f, tau = .5),col="red")
 
  dev.off()
 
 
 # Fx RETURNS    0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 1 1 1 0 1 
 and Thus plot is not proper
 
  
 
 
 Please guide me as where I am going wrong.
 
 With regards
 
 
 Maithili

Congratulations, you have discovered a bug in the implementation of 
function cdfgum.  A corrected version of package lmom will be submitted 
today and should be available on CRAN within a day or two.

Meanwhile, as a workaround define cdfgum in your own workspace, thus:

cdfgum <- function(x,para) exp(-exp(-(x - para[1])/para[2]))



J. R. M. Hosking


Maithili Shiva wrote:
> Dera R Helpers,
> 
> I am re-posting my query.
> 
> Please guide me.
> 
> Maithili
> 
> 
> --- On Fri, 3/13/09, Maithili Shiva <maithili_shiva at yahoo.com>
wrote:
> 
>  I am trying to fit the Gumbel distribution to a data. I am
>  using lmom package. I am getting problem in Cumulative
>  Distribution Function of Gumbel distribution as I am getting
>  it as a series of 0's and 1's thereby affecting the
>  P P Plot. My R code is as follows.
>  
>  
>   library(quantreg)
>   library(RODBC)
>   library(MASS)
>   library(actuar)
>   library(lmom)
>  
>   x <-
>  
c(986.78,1067.76,1046.47,1034.71,1004.53,1007.89,964.94,1060.24,1188.07,1085.63,988.33,972.71,1177.71,972.48,1203.20,1047.27,1062.95,1113.65,995.97,1093.98)
>  
>  
>  #Estimating the parameters for GUMBEL distribution
>  
>   N                    <- length(x)
>  
>   lmom 		    <- samlmu(x); lmom
>  
>   parameters_of_GUMBEL <- pelgum(lmom); parameters_of_GUMBEL
>   
>  
>   # Parameters are xi =  1019.4003   alpha =   59.5327
>  
>  
>  
>   # _____________________ P - P Plot ________________________
>   
>   e        <- c(1:N)
>  
>   f        <- c((e-.5)/N)
>  
>   Fx       <- cdfgum(x, para = parameters_of_GUMBEL)
>   
>   g        <- sort(Fx)
>   
>   png(filename = "GUMBEL_P-P.png")
>   
>   a <- par('bg'= "#FFFFCC")
>   
>   plot (f,g,bg=a,fg= "#804000",main ="P-P
>  Plot", ylab=" Cumulative Distribution
>  Function", xlab="i", font.main=2,
>  cex.main=1,col="#999966",bty > 
"o",col.main="black",col.axis="black",col.lab
>  ="black")
>  
>   abline(rq(g ~ f, tau = .5),col="red")
>  
>   dev.off()
>  
>  
>  # Fx RETURNS    0 1 1 1 0 0 0 1 1 1 0 0 1 0 1 1 1 1 0 1 
>  and Thus plot is not proper
>  
>   
>  
>  
>  Please guide me as where I am going wrong.
>  
>  With regards
>  
>  
>  Maithili
> 
> ______________________________________________
> R-help at 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.
>