jml.optimization
Class QPWithBoundConstraints

java.lang.Object
  extended by jml.optimization.QPWithBoundConstraints

public class QPWithBoundConstraints
extends java.lang.Object

Quadratic programming with bound constraints:

min 2 \ x' * Q * x + c' * x s.t. l <= x <= u

Version:
1.0, Feb. 25th, 2013
Author:
Mingjie Qian

Constructor Summary
QPWithBoundConstraints()
           
 
Method Summary
static void main(java.lang.String[] args)
           
static QPSolution solve(org.apache.commons.math.linear.RealMatrix Q, org.apache.commons.math.linear.RealMatrix c, double l, double u, double epsilon)
          Solve this bound constrained QP problem.
static QPSolution solve(org.apache.commons.math.linear.RealMatrix Q, org.apache.commons.math.linear.RealMatrix c, double l, double u, double epsilon, org.apache.commons.math.linear.RealMatrix x0)
          Solve this bound constrained QP problem.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

QPWithBoundConstraints

public QPWithBoundConstraints()
Method Detail

main

public static void main(java.lang.String[] args)
Parameters:
args -

solve

public static QPSolution solve(org.apache.commons.math.linear.RealMatrix Q,
                               org.apache.commons.math.linear.RealMatrix c,
                               double l,
                               double u,
                               double epsilon)
Solve this bound constrained QP problem.

Parameters:
Q - the positive semi-definite matrix
c - the linear coefficient vector
l - lower bound
u - upper bound
epsilon - convergence precision
Returns:
a QPSolution instance containing the optimizer and the optimum

solve

public static QPSolution solve(org.apache.commons.math.linear.RealMatrix Q,
                               org.apache.commons.math.linear.RealMatrix c,
                               double l,
                               double u,
                               double epsilon,
                               org.apache.commons.math.linear.RealMatrix x0)
Solve this bound constrained QP problem.

Parameters:
Q - the positive semi-definite matrix
c - the linear coefficient vector
l - lower bound
u - upper bound
epsilon - convergence precision
x0 - staring point if not null
Returns:
a QPSolution instance containing the optimizer and the optimum