[Help-gsl] Comment on documentation of multiroots/fsolver code

[John Pye]

Hi there

I had a comment on the behaviour of the multiroots code, and what I
thought seemed like a gap in the documentation.

When running gsl_multiroot_fsolver_iterate, I was able to converge on a
solution to my problem. Having passed a *pointer* to my vector of
variables 'x', I assumed that this vector would have been updated at the
end of the calculation to contain the solution vector. However, the
value had not been updated. The solution vector was in the internal s->x
vector instead.

Can I suggest that there might need to be a section added somewhere near
section 34.8 in the manual, perhaps titled "Retrieving the solution
vector" or something like that. It should be made clear whether or not
s->x always contains the final solution vector, and whether the vector
s->f always contains the residual vector corresponding to that 'x', or not.

I'm looking here at the version 1.9 documentation. If this has recently
been fixed, please ignore this.

Cheers
JP