Re: [Bug-gsl] bug in gsl_root_test_delta()

[Brian Gough]

Dave Spiegel writes:
 > I think that I have found a bug in the function gsl_root_test_delta.
 > When x1 and x0 are both 0.0, gsl_root_test_delta returns -2, indicating 
 > failure to converge.

Hello, 

Thanks for the bug report.  The result is a bit unintuitive for x1==0
but it is following the inequality in the documentation.

For roots close to zero, add a small epsabs term like GSL_DBL_MIN to
work around it.  I'll consider changing the behaviour in future to
make it more obvious.

-- 
best regards, 

Brian Gough

Network Theory Ltd,
Publishing the GSL Manual - http://www.network-theory.co.uk/gsl/manual/