Re: quadratic programming (qp) constraints -- always enforced?

[Leo Butler]

fork <address@hidden> writes:

> Hi all, 
>
> When I call qp() with A_in and b to define constraints, I am getting a 
> solution
> returned with info.solveiter=3 and info.info=0 but which is outside the
> constraints.  (I start with a wacky x0, if that matters.)
>
> Is this even possible?
>
> I haven't run a small test case yet, since it seems like it is probably my 
> fault
> in getting the constraint matrix screwed up, but I thought I would check the
> collective wisdom first.

An example would be illuminating. My guess is that the bug is in your
specification, not the qp code. Leo