Re: [Bug-gsl] About GSL

[Patrick Alken]

Thanks for your report, I've fixed the spelling error on the git repository.

For the (n % 2) vs (n & 1) issue, there is a good discussion at:

http://stackoverflow.com/questions/160930/how-do-i-check-if-an-integer-is-even-or-odd

The consensus seems to be that:

1) Most modern compilers will produce identical output for the two cases
2) n & 1 may not work on negative numbers, while n % 2 is guaranteed to 
work correctly

On 09/02/2014 10:01 AM, Guillaume Gomez wrote:
> Hello,
>
> I'm currently binding the GSL into Rust (when this part is finished, I'll
> try to totally port the code). You can follow the development on github on
> this link if you want : https://github.com/GuillaumeGomez/rust-GSL
>
> But here's what's made contact you : while I was looking at the source, I
> saw that sometimes, you use code like that :
>
> int n;
> if (n % 2 == 0)
>
> Why this instead of :
>
> int n;
> if (n & 1 == 0)
>
> ? The second way is a little faster if I'm not wrong.
>
> Another thing : for example in the function gsl_integration_qawc, there is
> a little mistake in an error message :
>
> "tolerance cannot be acheived with given epsabs and epsrel". Acheived
> should be achieved if I'm not wrong.
>
> Thanks in advance for you answer.
>
> Cordially.