R help - Apr 2005 - Inverse of the Laplace Transform/Gaver Stehfest algorithm

Hi there,

Is there an implementation of the Gaveh Stehfest algorithm in R 
somewhere ? Or some other inversion ?

Thanks,
Tolga

Tolga Uzuner wrote:
> Hi there,
>
> Is there an implementation of the Gaveh Stehfest algorithm in R 
> somewhere ? Or some other inversion ?
>
> Thanks,
> Tolga
>
Well, at least here is Zakian's algorithm, for anyone who needs it:

Zakian<-function(Fs,t){
# Fs is the function to be inverted and evaluated at t
    a = c(12.83767675+1.666063445i, 
12.22613209+5.012718792i,10.93430308+8.409673116i, 
8.776434715+11.92185389i,5.225453361+15.72952905i)
    K = c(-36902.08210+196990.4257i, 
61277.02524-95408.62551i,-28916.56288+18169.18531i, 
+4655.361138-1.901528642i,-118.7414011-141.3036911i)
    ssum = 0.0
#   Zakian's method does not work for t=0. Check that out.
    if(t == 0){
        print("ERROR:   Inverse transform can not be calculated for
t=0")
        print("WARNING: Routine zakian() exiting.")
        return("Error")}
#   The for-loop below is the heart of Zakian's Inversion Algorithm.
    for(j in 1:5){ssum = ssum + Re(K[j]*Fs(a[j]/t))}
 
    return (2.0*ssum/t)
}

#   InvLap(1/(s-1))=exp(t)
#   check if Zakian(function(s){1/(s-1)},1)==exp(1)
lapfunc<-function(s){1.0/(s-1.0)}

#   Function Zakian(functobeinverted,t) is invoked.
Zakian(lapfunc,1.0)

How do I suppress the following ?

Warning messages:
1: the condition has length > 1 and only the first element will be used
in: if (strike == forward) atmvol(forward, t, alpha, beta, rho, upsilon)
else {
2: the condition has length > 1 and only the first element will be used
in: if (x(z) == 0) 1 else z/x(z)

Tolga Uzuner wrote:
> How do I suppress the following ?
>
> Warning messages:
> 1: the condition has length > 1 and only the first element will be used
> in: if (strike == forward) atmvol(forward, t, alpha, beta, rho, upsilon)
> else {
> 2: the condition has length > 1 and only the first element will be used
> in: if (x(z) == 0) 1 else z/x(z)
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! 
> http://www.R-project.org/posting-guide.html
>
>
>
Maybe better to understand what generates the warning!

To assure you are uninformed, say
options(warn=-1)

Kjetil

-- 

Kjetil Halvorsen.

Peace is the most effective weapon of mass construction.
               --  Mahdi Elmandjra




-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.

On 5/15/05, Tolga Uzuner <tolga at coubros.com>
wrote:
> How do I suppress the following ?
> 
> Warning messages:
> 1: the condition has length > 1 and only the first element will be used
> in: if (strike == forward) atmvol(forward, t, alpha, beta, rho, upsilon)
> else {
> 2: the condition has length > 1 and only the first element will be used
> in: if (x(z) == 0) 1 else z/x(z)
Check out ?suppressWarnings

Kjetil Brinchmann Halvorsen wrote:
> Tolga Uzuner wrote:
> 
>> How do I suppress the following ?
>>
>> Warning messages:
>> 1: the condition has length > 1 and only the first element will be
used
>> in: if (strike == forward) atmvol(forward, t, alpha, beta, rho,
upsilon)
>> else {
>> 2: the condition has length > 1 and only the first element will be
used
>> in: if (x(z) == 0) 1 else z/x(z)
>>
>> ______________________________________________
>> R-help at stat.math.ethz.ch mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide! 
>> http://www.R-project.org/posting-guide.html
>>
>>
>>
> Maybe better to understand what generates the warning!
Yes! In both cases you should really look why you are using *conditions 
of length > 1*! And if this is intended, you certainly want to use 
"ifelse()" rather than "if(){} else{}".

Uwe Ligges

> To assure you are uninformed, say
> options(warn=-1)
> 
> Kjetil
>

>>>>> "UweL" == Uwe Ligges <ligges at
statistik.uni-dortmund.de>
>>>>>     on Sun, 15 May 2005 16:41:22 +0200 writes:
    UweL> Kjetil Brinchmann Halvorsen wrote:
    >> Tolga Uzuner wrote:
    >> 
    >>> How do I suppress the following ?
    >>> 
    >>> Warning messages: 1: the condition has length > 1 and
    >>> only the first element will be used in: if (strike =   
>>> forward) atmvol(forward, t, alpha, beta, rho, upsilon)
    >>> else { 2: the condition has length > 1 and only the
    >>> first element will be used in: if (x(z) == 0) 1 else
    >>> z/x(z)
    >>> 
    >>> ______________________________________________
    >>> R-help at stat.math.ethz.ch mailing list
    >>> https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do
    >>> read the posting guide!
    >>> http://www.R-project.org/posting-guide.html
    >>> 
    >>> 
    >>> 
    >> Maybe better to understand what generates the warning!

    UweL> Yes! In both cases you should really look why you are
    UweL> using *conditions of length > 1*! 

yes, indeed!
This is a bug almost always (my subjective probability : 0.995)

Maybe you want to use
     if(any(...)) 
or   if(all(...)) 
instead of the current  if(...)  

    UweL> And if this is intended, you certainly want to use
"ifelse()" rather
    UweL> than "if(){} else{}".

(from my above guess, the probability for this would be about 1:200)

Martin

    >> To assure you are uninformed, say options(warn=-1)
    >> 
    >> Kjetil