Re: [Bug-gsl] lambert W function

[Brian Gough]

At Wed, 25 Apr 2007 17:35:34 +0200,
Katrin Wolff wrote:
> The Lambert W(x) function gives errors for some values x around 1.6. If I 
> re-insert the value into the function with high precision the error persists, 
> if I truncate after the 16th digit and fill up with zeros it does not. The 
> error message is
> 
> gsl: lambert.c:224: ERROR: gsl_sf_lambert_W0_e(x, &result)
> Default GSL error handler invoked.
> doitflag.sh: line 5: 21853 Abort trap
> 
> and example values to produce it are
> 
> 1.6014351509876314505476102567627094686031341552734
> 1.6849341956993197921832461361191235482692718505859

Can you send an example program which shows the problem so we can
reproduce it. Thanks.

-- 
Brian Gough

Network Theory Ltd,
Publishing Free Software Manuals --- http://www.network-theory.co.uk/