Path: | rdoc/multimin.rdoc |
Last Update: | Sun Nov 14 22:53:48 +0000 2010 |
This chapter describes routines for finding minima of arbitrary multidimensional functions. The library provides low level components for a variety of iterative minimizers and convergence tests. These can be combined by the user to achieve the desired solution, while providing full access to the intermediate steps of the algorithms. Each class of methods uses the same framework, so that you can switch between minimizers at runtime without needing to recompile your program. Each instance of a minimizer keeps track of its own state, allowing the minimizers to be used in multi-threaded programs.
Contents:
The problem of multidimensional minimization requires finding a point x such that the scalar function, takes a value which is lower than at any neighboring point. For smooth functions the gradient g = \nabla f vanishes at the minimum. In general there are no bracketing methods available for the minimization of n-dimensional functions. The algorithms proceed from an initial guess using a search algorithm which attempts to move in a downhill direction.
Algorithms making use of the gradient of the function perform a one-dimensional line minimisation along this direction until the lowest point is found to a suitable tolerance. The search direction is then updated with local information from the function and its derivatives, and the whole process repeated until the true n-dimensional minimum is found.
The Nelder-Mead Simplex algorithm applies a different strategy. It maintains n+1 trial parameter vectors as the vertices of a n-dimensional simplex. In each iteration step it tries to improve the worst vertex by a simple geometrical transformation until the size of the simplex falls below a given tolerance.
Both types of algorithms use a standard framework. The user provides a high-level driver for the algorithms, and the library provides the individual functions necessary for each of the steps. There are three main phases of the iteration. The steps are,
Each iteration step consists either of an improvement to the line-minimisation in the current direction or an update to the search direction itself. The state for the minimizers is held in a GSL::MultiMin::FdfMinimizer or a GSL::MultiMin::FMinimizer object.
Note that the minimization algorithms can only search for one local minimum at a time. When there are several local minima in the search area, the first minimum to be found will be returned; however it is difficult to predict which of the minima this will be. In most cases, no error will be reported if you try to find a local minimum in an area where there is more than one.
It is also important to note that the minimization algorithms find local minima; there is no way to determine whether a minimum is a global minimum of the function in question.
These method create a minimizer of type type for an n-dimension function. The type is given by a string, or by a Ruby constant.
ex:
include GSL::MultiMin m1 = FdfMinimizer.alloc(FdfMinimizer::CONJUGATE_FR, 2) m2 = FdfMinimizer.alloc("steepest_descent", 4) m3 = FMinimizer.alloc(FMinimizer::NMSIMPLEX, 3) m4 = FMinimizer.alloc("nmsimplex", 2)
This method initializes the minimizer self to minimize the function fdf (the GSL::MultiMin::Function_fdf class, see below) starting from the initial point x (GSL::Vector). The size of the first trial step is given by step_size (Vector). The accuracy of the line minimization is specified by tol.
This method initializes the minimizer self to minimize the function func, starting from the initial point x (Vector). The size of the initial trial steps is given in vector step_size.
These return the name of the minimizer self.
You must provide a parametric function of n variables for the minimizers to operate on. You may also need to provide a routine which calculates the gradient of the function. In order to allow for general parameters the functions are defined by the classes, GSL::MultiMin::Function_fdf and GSL::MultiMin::Function.
See example below.
include GSL::MultiMin my_f = Proc.new { |v, params| x = v[0]; y = v[1] p0 = params[0]; p1 = params[1] 10.0*(x - p0)*(x - p0) + 20.0*(y - p1)*(y - p1) + 30.0 } my_df = Proc.new { |v, params, df| x = v[0]; y = v[1] p0 = params[0]; p1 = params[1] df[0] = 20.0*(x-p0) df[1] = 40.0*(y-p1) } my_func = Function_fdf.alloc(my_f, my_df, 2) my_func.set_params([1.0, 2.0]) # parameters
See example below.
include GSL::MultiMin np = 2 my_f = Proc.alloc { |v, params| x = v[0]; y = v[1] p0 = params[0]; p1 = params[1] 10.0*(x - p0)*(x - p0) + 20.0*(y - p1)*(y - p1) + 30.0 } my_func = Function.alloc(my_f, np) my_func.set_params([1.0, 2.0]) # parameters
These methods perform a single iteration of the minimizer self. If the iteration encounters an unexpected problem then an error code will be returned. The minimizer maintains a current best estimate of the minimum at all times. This information can be accessed with the following methods,
These method return the current best estimate of the location of the minimum, the value of the function at that point, its gradient, and minimizer specific characteristic size for the minimizer self.
This method resets the minimizer self to use the current point as a new starting point.
A minimization procedure should stop when one of the following conditions is true:
The handling of these conditions is under user control. The methods below allow the user to test the precision of the current result.
These method test the norm of the gradient g against the absolute tolerance epsabs. The gradient of a multidimensional function goes to zero at a minimum. The tests return GSL::SUCCESS if the following condition is achieved,
|g| < epsabs
and returns GSL::CONTINUE otherwise. A suitable choice of epsabs can be made from the desired accuracy in the function for small variations in x. The relationship between these quantities is given by \delta f = g \delta x.
These method test the minimizer specific characteristic size (if applicable to the used minimizer) against absolute tolerance epsabs. The tests return (GSL::SUCCESS if the size is smaller than tolerance, otherwise GSL::CONTINUE is returned.
#!/usr/bin/env ruby require("gsl") include GSL::MultiMin my_f = Proc.new { |v, params| x = v[0]; y = v[1] p0 = params[0]; p1 = params[1] 10.0*(x - p0)*(x - p0) + 20.0*(y - p1)*(y - p1) + 30.0 } my_df = Proc.new { |v, params, df| x = v[0]; y = v[1] p0 = params[0]; p1 = params[1] df[0] = 20.0*(x-p0) df[1] = 40.0*(y-p1) } my_func = Function_fdf.alloc(my_f, my_df, 2) my_func.set_params([1.0, 2.0]) # parameters x = Vector.alloc(5.0, 7.0) # starting point minimizer = FdfMinimizer.alloc("conjugate_fr", 2) minimizer.set(my_func, x, 0.01, 1e-4) iter = 0 begin iter += 1 status = minimizer.iterate() status = minimizer.test_gradient(1e-3) if status == GSL::SUCCESS puts("Minimum found at") end x = minimizer.x f = minimizer.f printf("%5d %.5f %.5f %10.5f\n", iter, x[0], x[1], f) end while status == GSL::CONTINUE and iter < 100
#!/usr/bin/env ruby require("gsl") include GSL::MultiMin np = 2 my_f = Proc.new { |v, params| x = v[0]; y = v[1] p0 = params[0]; p1 = params[1] 10.0*(x - p0)*(x - p0) + 20.0*(y - p1)*(y - p1) + 30.0 } my_func = Function.alloc(my_f, np) my_func.set_params([1.0, 2.0]) # parameters x = Vector.alloc([5, 7]) ss = Vector.alloc(np) ss.set_all(1.0) minimizer = FMinimizer.alloc("nmsimplex", np) minimizer.set(my_func, x, ss) iter = 0 begin iter += 1 status = minimizer.iterate() status = minimizer.test_size(1e-2) if status == GSL::SUCCESS puts("converged to minimum at") end x = minimizer.x printf("%5d ", iter); for i in 0...np do printf("%10.3e ", x[i]) end printf("f() = %7.3f size = %.3f\n", minimizer.fval, minimizer.size); end while status == GSL::CONTINUE and iter < 100