mo_anneal Module Reference

Anneal optimization of cost function. More...

Data Types
interface	anneal
	Optimize cost function with simulated annealing. More...
interface	generate_neighborhood_weight
interface	gettemperature
	Find initial temperature for simulated annealing. More...

Functions/Subroutines
real(dp) function, dimension(size(para, 1))	anneal_dp (eval, cost, para, prange, prange_func, temp, dt, nitermax, len, nst, eps, acc, seeds, printflag, maskpara, weight, changeparamode, reflectionflag, pertubflexflag, maxit, undef_funcval, tmp_file, funcbest, history)
real(dp) function	gettemperature_dp (eval, cost, paraset, acc_goal, prange, prange_func, samplesize, maskpara, seeds, printflag, weight, maxit, undef_funcval)
real(dp) function	pargen_anneal_dp (old, dmax, omin, omax, rn)
real(dp) function	pargen_dds_dp (old, perturb, omin, omax, rn)
real(dp) function	dchange_dp (delta, idigit, iszero)
subroutine	generate_neighborhood_weight_dp (truepara, cum_weight, save_state_xor, itotalcounter, nitermax, neighborhood)

Detailed Description

Anneal optimization of cost function.

Minimization of cost function and temperature finding of minima.

Changelog

• Juliane Mai, Mar 2012
  • module implementation
• Juliane Mai, May 2012
  • anneal: sp version
• Juliane Mai, May 2012
  • anneal: documentation
• Juliane Mai, May 2012
  • GetTemperature: sp and dp version
• Juliane Mai, Jun 2012
  • weighted parameter selection
• Juliane Mai, Aug 2012
  • function anneal instead of subroutine
  • using new module get_timeseed as default for seeding
  • new optional for minimization or maximization
  • fixed parameter ranges possible instead of interface range
• Juliane Mai, Nov 2012
  • history of achieved objective function values as optional out only in anneal but not anneal_valid
• Juliane Mai, Jan 2013
  • including DDS features in anneal, i.e. reflection at parameter boundaries, different parameter selection modes (one, all, neighborhood), and different parameter pertubation modes (flexible r=dR (anneal version) or constant r=0.2 (dds version))
  • remove sp versions
  • fixed and flexible parameter ranges are now in one function using optional arguments
  • undef_funcval instead of anneal_valid function
• Juliane Mai, Feb 2013
  • xor4096 optionals combined in save_state
• Arya Prasetya, Dec 2021
  • doxygen documentation anneal and get_temperature

Author

Juliane Mai

Date

Mar 2012

Copyright

Copyright 2005-2024, the CHS Developers, Sabine Attinger: All rights reserved. FORCES is released under the LGPLv3+ license

The UFZ (CHS) FORCES code is free software. You can redistribute it and/or modify it under the terms of the GNU General Public License as published by the free Software Foundation either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You received a copy of the GNU Lesser General Public License along with the UFZ CHS FORCES code. It can be found in the files COPYING and COPYING.LESSER provided with this software. The complete GNU license text can also be found at http://www.gnu.org/licenses/.

Function/Subroutine Documentation

anneal_dp()

real(dp) function, dimension(size(para, 1)) mo_anneal::anneal_dp ( procedure(eval_interface), intent(in), pointer  eval, procedure(objective_interface), intent(in), pointer  cost, real(dp), dimension(:), intent(in)  para, real(dp), dimension(size(para, 1), 2), intent(in), optional  prange, external subroutine(real(dp), dimension(:), intent(in) paraset, integer(i4), intent(in) ipar, real(dp), dimension(2), intent(out) rangepar), optional  prange_func, real(dp), intent(in), optional  temp, real(dp), intent(in), optional  dt, integer(i4), intent(in), optional  nitermax, integer(i4), intent(in), optional  len, integer(i4), intent(in), optional  nst, real(dp), intent(in), optional  eps, real(dp), intent(in), optional  acc, integer(i8), dimension(3), intent(in), optional  seeds, logical, intent(in), optional  printflag, logical, dimension(size(para, 1)), intent(in), optional  maskpara, real(dp), dimension(size(para, 1)), intent(in), optional  weight, integer(i4), intent(in), optional  changeparamode, logical, intent(in), optional  reflectionflag, logical, intent(in), optional  pertubflexflag, logical, intent(in), optional  maxit, real(dp), intent(in), optional  undef_funcval, character(len = *), intent(in), optional  tmp_file, real(dp), intent(out), optional  funcbest, real(dp), dimension(:, :), intent(out), optional, allocatable  history  )

private

Parameters

[in]	para	initial parameter
[in]	prange	lower and upper limit per parameter
[in]	temp	starting temperature (DEFAULT: Get_Temperature)
[in]	dt	geometrical decreement, 0.7<DT<0.999 (DEFAULT: 0.9)
[in]	nitermax	maximal number of iterations (DEFAULT: 1000)
[in]	len	Length of Markov Chain, DEFAULT: max(250, size(para,1))
[in]	nst	Number of consecutive LEN steps! (DEFAULT: 5)
[in]	eps	epsilon decreement of cost function (DEFAULT: 0.01)
[in]	acc	Acceptance Ratio, <0.1 stopping criteria (DEFAULT: 0.1)
[in]	seeds	Seeds of random numbers (DEFAULT: Get_timeseed)
[in]	printflag	If command line output is written (.true.) (DEFAULT: .false.)
[in]	maskpara	true if parameter will be optimized false if parameter is discarded in optimization (DEFAULT: .true.)
[in]	weight	vector of weights per parameter gives the frequency of parameter to be chosen for optimization (DEFAULT: uniform)
[in]	changeparamode	which and how many param. are changed in a step 1 = one parameter 2 = all parameter 3 = neighborhood parameter (DEFAULT: 1_i4)
[in]	reflectionflag	if new parameter values are selected normal distributed and reflected (.true.) or uniform in range (.false.) (DEFAULT: .false.)
[in]	pertubflexflag	if pertubation of normal distributed parameter values is constant 0.2 (.false.) or depends on dR (.true.) (DEFAULT: .true.)
[in]	maxit	Maximization or minimization of function maximization = .true., minimization = .false. (DEFAULT: .false.)
[in]	undef_funcval	objective function value occuring if parameter set leads to invalid model results, e.g. -9999.0_dp! (DEFAULT: not present)
[in]	tmp_file	file for temporal output
[out]	funcbest	minimized value of cost function (DEFAULT: not present)
[out]	history	returns a vector of achieved objective! after ith model evaluation (DEFAULT: not present)

Returns

parameter set minimizing objective

Definition at line 289 of file mo_anneal.f90.

dchange_dp()

real(dp) function mo_anneal::dchange_dp ( real(dp), intent(in)  delta, integer(i4), intent(in)  idigit, integer(i4), intent(in)  iszero  )

private

Definition at line 1319 of file mo_anneal.f90.

generate_neighborhood_weight_dp()

subroutine mo_anneal::generate_neighborhood_weight_dp ( integer(i4), dimension(:), intent(in)  truepara, real(dp), dimension(:), intent(in)  cum_weight, integer(i8), dimension(n_save_state), intent(inout)  save_state_xor, integer(i4), intent(in)  itotalcounter, integer(i4), intent(in)  nitermax, logical, dimension(size(cum_weight)), intent(out)  neighborhood  )

private

Definition at line 1344 of file mo_anneal.f90.

gettemperature_dp()

real(dp) function mo_anneal::gettemperature_dp ( procedure(eval_interface), intent(in), pointer  eval, procedure(objective_interface), intent(in), pointer  cost, real(dp), dimension(:), intent(in)  paraset, real(dp), intent(in)  acc_goal, real(dp), dimension(size(paraset, 1), 2), intent(in), optional  prange, external subroutine(real(dp), dimension(:), intent(in) paraset, integer(i4), intent(in) ipar, real(dp), dimension(2), intent(out) rangepar), optional  prange_func, integer(i4), intent(in), optional  samplesize, logical, dimension(size(paraset, 1)), intent(in), optional  maskpara, integer(i8), dimension(2), intent(in), optional  seeds, logical, intent(in), optional  printflag, real(dp), dimension(size(paraset, 1)), intent(in), optional  weight, logical, intent(in), optional  maxit, real(dp), intent(in), optional  undef_funcval  )

private

Parameters

[in]	paraset	a valid parameter set of the model
[in]	acc_goal	acceptance ratio to achieve
[in]	prange	lower and upper limit per parameter
[in]	samplesize	size of random set the acc_estimate is based on. DEFAULT: Max(250, 20*Number paras)
[in]	maskpara	true if parameter will be optimized. false if parameter is discarded in optimization. DEFAULT: .true.
[in]	seeds	Seeds of random numbers. DEFAULT: time dependent.
[in]	printflag	.true. if detailed temperature estimation is printed. DEFAULT: .false.
[in]	weight	vector of weights per parameter gives the frequency of parameter to be chosen for optimization. DEFAULT: equal weighting
[in]	maxit	Maxim. or minim. of function maximization = .true., minimization = .false. . DEFAULT: .false.
[in]	undef_funcval	objective function value occuring if parameter set leads to invalid model results, e.g. -999.0_dp. DEFAULT: not present

Definition at line 991 of file mo_anneal.f90.

Referenced by mo_anneal::gettemperature::gettemperature_dp().

Here is the caller graph for this function:

pargen_anneal_dp()

real(dp) function mo_anneal::pargen_anneal_dp ( real(dp), intent(in)  old, real(dp), intent(in)  dmax, real(dp), intent(in)  omin, real(dp), intent(in)  omax, real(dp), intent(in)  rn  )

private

Definition at line 1252 of file mo_anneal.f90.

pargen_dds_dp()

real(dp) function mo_anneal::pargen_dds_dp ( real(dp), intent(in)  old, real(dp), intent(in)  perturb, real(dp), intent(in)  omin, real(dp), intent(in)  omax, real(dp), intent(in)  rn  )

private

Definition at line 1292 of file mo_anneal.f90.