bug-gsl
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-gsl] Error in gsl_cdf_fdist_Pinv


From: Daniel Falster
Subject: Re: [Bug-gsl] Error in gsl_cdf_fdist_Pinv
Date: Mon, 13 Nov 2006 07:47:20 +1100
User-agent: Thunderbird 1.5.0.7 (Windows/20060909)

Hi Brian,
Sorry, below find example program and output from my windows XP. I am using ver1.8 (30 Mar 06) downloaded from sourceforge.net. Thanks, Daniel

PROGRAM
#include <gsl/gsl_randist.h>
#include <gsl/gsl_cdf.h>
#include <iostream>
using namespace std;

int main(void)
{
for(int df =1; df<350; df+=10)
       cout<<df<<"\t"<<gsl_cdf_fdist_Pinv(0.95,1,df)<<endl;
system("pause");
return 0;
}

OUTPUT:
Microsoft Windows XP [Version 5.1.2600]
F:\_resources\C>Project1
1       161.448
11      4.84434
21      4.32479
31      4.15962
41      4.07855
51      4.03039
61      3.99849
71      3.97581
81      3.95885
91      3.94569
101     3.93519
111     3.92661
121     3.91946
131     3.91343
141     3.90826
151     3.90378
161     3.89987
171     3.89642
181     3.89335
191     3.89061
201     3.88814
211     3.88591
221     3.88388
231     3.88203
241     3.88033
251     3.87877
261     -1.#IND
271     -1.#IND
281     -1.#IND
291     -1.#IND
301     -1.#IND
311     -1.#IND
321     -1.#IND
331     -1.#IND
341     -1.#IND
Press any key to continue . . .


Brian Gough wrote:
At Fri, 10 Nov 2006 12:08:31 +1100,
Daniel Falster wrote:
The function double gsl_cdf_fdist_Pinv (double P, double nu1, double
nu2) is in GSL ver1.8 but hasn't been included in earlier versions. It
returns the inverse of the cumulative distribution function for the
F-distribution. The arguments nu1 and nu2 are the degrees of freedom
(1,..., Inf). When nu1 =1, and nu2 greater than ~ 250 the functions
fails. I have verified this for both windows and unix versions.
For example, for (int y = 1; y<500; y+=20)
                     cout<<y<<"
\t"<<gsl_cdf_fdist_Pinv(0.95,1.0,y)<<endl;
fails for y > 240 on windows XP.

Hello,

Thanks for your email. Could you send the whole example program (with
#include's etc) and its complete output showing how it fails, so I can
reproduce the problem correctly.


--
**********************************
Daniel Falster
Biological Sciences
Macquarie University, Australia
Work: address@hidden
Home: address@hidden
http://daniel.falster.no/
**********************************





reply via email to

[Prev in Thread] Current Thread [Next in Thread]