Re: midpoint rule for numerical integration

[Brian Gough <bjg at network-theory dot co dot uk>]

Daniel Rohe writes:
 > Hi everyone,
 >  although methods like e.g. midpoint rule are not exactly state of
 > the art I would find it useful to have a gsl-implementation for
 > certain special cases or checks.  I could myself contribute the
 > very simple low-level routine for an open-midpoint-rule which in
 > particular has the advantage of using already evaluated points for
 > the next iteration/refinement. however I would have to rely on
 > someone to integrate it into a higher level gsl function.
 >  what do you think?
 >  daniel

It could be useful to have a simple integrator as a check.  

Feel free to work on it independently, following the GSL design
document and existing GSL integration api, and post your progress on
this list. e.g. url for the source.  It should be possible initially
to write it as a single file which can be dropped into any application
as an add-on to GSL.

If there's favorable feedback we can see about putting it in GSL when
it is stable and fully-tested.

Brian