[SciPy-user] Solving an equation using scipy.optimize.newton

fdu.xiaojf@gmai... fdu.xiaojf@gmai...
Wed Sep 5 20:57:41 CDT 2007


Thanks Dominique and dmitrey,

Dominique Orban wrote:
 >
 >
 > On 9/5/07, *fdu.xiaojf@gmail.com <mailto:fdu.xiaojf@gmail.com>*
 > <fdu.xiaojf@gmail.com <mailto:fdu.xiaojf@gmail.com>> wrote:
 >
 >     Hi all,
 >
 >     I'm trying to solve an equation f(x) = 0 with scipy.optimize.newton .
 >
 >     However the problem isn't so simple. There are bound constraints for my
 >     equation: the equation cannot be evaluated when x is out of [Min,
 >     Max], but
 >     the root is always in the interval of [Min, Max]
 >
 >     When newton() iterates to find a root, it sometimes try to evaluate the
 >     equation with a x out of [Min, Max], and then error occurs.
 >
 >     How to solve this problem ?
 >
 >     I couldn't easily find two points with different signs every time,
 >     so methods
 >     like brentq don't work here.
 >
 >
 > There are variants of Newton's method that can handle bound constraints.
 > However, another way to treat your problem would be to solve the
 > optimization problem:

Are there any trust region optimization method with python interface ?

 >
 > minimize   0
 > subject to   f(x) = 0, and Min <= x <= Max.
 >
 > The objective function of this problem is constant, so any x satisfying
 > the constraints is optimal, and is what you are looking for. Note
 > however that you now have an optimization problem with nonlinear
 > equality constraints.
 >
 > If a solver isn't able to identify a point satisfying the constraints,
 > it will usually guarantee some sort of 'proximity property', i.e., the
 > final iterate will minimize the residual of constraints in some sense.
 >
 > Dominique
 >

I can transform my problem to a bound constrained optimization problem,
but I still have to find an optimization solver who only evaluate my
equation in trust region([Min, Max]). Will lbfgsb or tnc meet the
requirement ?

Regards,



More information about the SciPy-user mailing list