octave: scripts/optimization/fminunc.m annotate

annotate scripts/optimization/fminunc.m @ 26376:00f796120a6d stable

maint: Update copyright dates in all source files.

author	John W. Eaton <jwe@octave.org>
date	Wed, 02 Jan 2019 16:32:43 -0500
parents	804e18e3e320
children	0a62d9a6aa2d 757a7119e319

rev	line source
26376 00f796120a6d maint: Update copyright dates in all source files. John W. Eaton <jwe@octave.org> parents: 26138 diff changeset	1 ## Copyright (C) 2008-2019 VZLU Prague, a.s.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	2 ##
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	3 ## This file is part of Octave.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	4 ##
24534 194eb4bd202b maint: Update punctuation for GPL v3 license text. Rik <rik@octave.org> parents: 23220 diff changeset	5 ## Octave is free software: you can redistribute it and/or modify it
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	6 ## under the terms of the GNU General Public License as published by
24534 194eb4bd202b maint: Update punctuation for GPL v3 license text. Rik <rik@octave.org> parents: 23220 diff changeset	7 ## the Free Software Foundation, either version 3 of the License, or
22755 3a2b891d0b33 maint: Standardize Copyright formatting. Rik <rik@octave.org> parents: 22323 diff changeset	8 ## (at your option) any later version.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	9 ##
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	10 ## Octave is distributed in the hope that it will be useful, but
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	11 ## WITHOUT ANY WARRANTY; without even the implied warranty of
22755 3a2b891d0b33 maint: Standardize Copyright formatting. Rik <rik@octave.org> parents: 22323 diff changeset	12 ## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
3a2b891d0b33 maint: Standardize Copyright formatting. Rik <rik@octave.org> parents: 22323 diff changeset	13 ## GNU General Public License for more details.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	14 ##
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	15 ## You should have received a copy of the GNU General Public License
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	16 ## along with Octave; see the file COPYING. If not, see
24534 194eb4bd202b maint: Update punctuation for GPL v3 license text. Rik <rik@octave.org> parents: 23220 diff changeset	17 ## <https://www.gnu.org/licenses/>.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	18 ##
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	19 ## Author: Jaroslav Hajek <highegg@gmail.com>
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	20
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	21 ## -- texinfo --
20852 516bb87ea72e 2015 Code Sprint: remove class of function from docstring for all m-files. Rik <rik@octave.org> parents: 20735 diff changeset	22 ## @deftypefn {} {} fminunc (@var{fcn}, @var{x0})
516bb87ea72e 2015 Code Sprint: remove class of function from docstring for all m-files. Rik <rik@octave.org> parents: 20735 diff changeset	23 ## @deftypefnx {} {} fminunc (@var{fcn}, @var{x0}, @var{options})
516bb87ea72e 2015 Code Sprint: remove class of function from docstring for all m-files. Rik <rik@octave.org> parents: 20735 diff changeset	24 ## @deftypefnx {} {[@var{x}, @var{fval}, @var{info}, @var{output}, @var{grad}, @var{hess}] =} fminunc (@var{fcn}, @dots{})
10793 be55736a0783 Grammarcheck the documentation from m-files. Rik <octave@nomad.inbox5.com> parents: 10687 diff changeset	25 ## Solve an unconstrained optimization problem defined by the function
be55736a0783 Grammarcheck the documentation from m-files. Rik <octave@nomad.inbox5.com> parents: 10687 diff changeset	26 ## @var{fcn}.
14895 e0525ecf156e Add new function fminsearch.m Andy Adler <andy@analyti.ca> parents: 14868 diff changeset	27 ##
20165 f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	28 ## @var{fcn} should accept a vector (array) defining the unknown variables, and
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	29 ## return the objective function value, optionally with gradient.
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	30 ## @code{fminunc} attempts to determine a vector @var{x} such that
20165 f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	31 ## @code{@var{fcn} (@var{x})} is a local minimum.
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	32 ##
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	33 ## @var{x0} determines a starting guess. The shape of @var{x0} is preserved in
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	34 ## all calls to @var{fcn}, but otherwise is treated as a column vector.
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	35 ##
26137 1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	36 ## @var{options} is a structure specifying additional parameters which
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	37 ## control the algorithm. Currently, @code{fminunc} recognizes these options:
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	38 ## @qcode{"AutoScaling"}, @qcode{"FinDiffType"}, @qcode{"FunValCheck"},
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	39 ## @qcode{"GradObj"}, @qcode{"MaxFunEvals"}, @qcode{"MaxIter"},
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	40 ## @qcode{"OutputFcn"}, @qcode{"TolFun"}, @qcode{"TolX"}, @qcode{"TypicalX"}.
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	41 ##
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	42 ## If @qcode{"AutoScaling"} is @qcode{"on"}, the variables will be
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	43 ## automatically scaled according to the column norms of the (estimated)
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	44 ## Jacobian. As a result, @qcode{"TolFun"} becomes scaling-independent. By
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	45 ## default, this option is @qcode{"off"} because it may sometimes deliver
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	46 ## unexpected (though mathematically correct) results.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	47 ##
26137 1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	48 ## If @qcode{"GradObj"} is @qcode{"on"}, it specifies that @var{fcn}--when
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	49 ## called with two output arguments---also returns the Jacobian matrix of
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	50 ## partial first derivatives at the requested point.
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	51 ##
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	52 ## @qcode{"MaxFunEvals"} proscribes the maximum number of function evaluations
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	53 ## before optimization is halted. The default value is
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	54 ## @code{100 * number_of_variables}, i.e., @code{100 * length (@var{x0})}.
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	55 ## The value must be a positive integer.
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	56 ##
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	57 ## @qcode{"MaxIter"} proscribes the maximum number of algorithm iterations
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	58 ## before optimization is halted. The default value is 400.
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	59 ## The value must be a positive integer.
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	60 ##
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	61 ## @qcode{"TolX"} specifies the termination tolerance for the unknown variables
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	62 ## @var{x}, while @qcode{"TolFun"} is a tolerance for the objective function
1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	63 ## value @var{fval}. The default is @code{1e-6} for both options.
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	64 ##
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	65 ## For a description of the other options, see @code{optimset}.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	66 ##
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	67 ## On return, @var{x} is the location of the minimum and @var{fval} contains
20165 f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	68 ## the value of the objective function at @var{x}.
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	69 ##
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	70 ## @var{info} may be one of the following values:
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	71 ##
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	72 ## @table @asis
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	73 ## @item 1
10793 be55736a0783 Grammarcheck the documentation from m-files. Rik <octave@nomad.inbox5.com> parents: 10687 diff changeset	74 ## Converged to a solution point. Relative gradient error is less than
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	75 ## specified by @code{TolFun}.
10821 693e22af08ae Grammarcheck documentation of m-files Rik <octave@nomad.inbox5.com> parents: 10793 diff changeset	76 ##
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	77 ## @item 2
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	78 ## Last relative step size was less than @code{TolX}.
10821 693e22af08ae Grammarcheck documentation of m-files Rik <octave@nomad.inbox5.com> parents: 10793 diff changeset	79 ##
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	80 ## @item 3
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	81 ## Last relative change in function value was less than @code{TolFun}.
10821 693e22af08ae Grammarcheck documentation of m-files Rik <octave@nomad.inbox5.com> parents: 10793 diff changeset	82 ##
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	83 ## @item 0
18816 e275d15c27b5 doc: Periodic spellcheck of documentation. Rik <rik@octave.org> parents: 18609 diff changeset	84 ## Iteration limit exceeded---either maximum number of algorithm iterations
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	85 ## @code{MaxIter} or maximum number of function evaluations @code{MaxFunEvals}.
923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	86 ##
923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	87 ## @item -1
18816 e275d15c27b5 doc: Periodic spellcheck of documentation. Rik <rik@octave.org> parents: 18609 diff changeset	88 ## Algorithm terminated by @code{OutputFcn}.
10821 693e22af08ae Grammarcheck documentation of m-files Rik <octave@nomad.inbox5.com> parents: 10793 diff changeset	89 ##
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	90 ## @item -3
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	91 ## The trust region radius became excessively small.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	92 ## @end table
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	93 ##
21546 f7f97d7e9294 doc: Wrap m-file docstrings to 79 characters + newline (80 total). Rik <rik@octave.org> parents: 21178 diff changeset	94 ## Optionally, @code{fminunc} can return a structure with convergence
f7f97d7e9294 doc: Wrap m-file docstrings to 79 characters + newline (80 total). Rik <rik@octave.org> parents: 21178 diff changeset	95 ## statistics (@var{output}), the output gradient (@var{grad}) at the
f7f97d7e9294 doc: Wrap m-file docstrings to 79 characters + newline (80 total). Rik <rik@octave.org> parents: 21178 diff changeset	96 ## solution @var{x}, and approximate Hessian (@var{hess}) at the solution
f7f97d7e9294 doc: Wrap m-file docstrings to 79 characters + newline (80 total). Rik <rik@octave.org> parents: 21178 diff changeset	97 ## @var{x}.
9627 5bcfa0b346e8 fix extra outputs in fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9623 diff changeset	98 ##
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	99 ## Application Notes: If the objective function is a single nonlinear equation
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	100 ## of one variable then using @code{fminbnd} is usually a better choice.
20165 f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	101 ##
22092 d3cef63f79ac * fminunc.m: Fix incorrect reference to fminunc as fminsearch (bug #48479) Mike Miller <mtmiller@octave.org> parents: 21758 diff changeset	102 ## The algorithm used by @code{fminunc} is a gradient search which depends
20165 f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	103 ## on the objective function being differentiable. If the function has
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	104 ## discontinuities it may be better to use a derivative-free algorithm such as
f1d0f506ee78 doc: Update more docstrings to have one sentence summary as first line. Rik <rik@octave.org> parents: 19833 diff changeset	105 ## @code{fminsearch}.
14895 e0525ecf156e Add new function fminsearch.m Andy Adler <andy@analyti.ca> parents: 14868 diff changeset	106 ## @seealso{fminbnd, fminsearch, optimset}
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	107 ## @end deftypefn
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	108
13027 b9a89ca0fb75 prevent optimization functions from setting ans in workspace at startup John W. Eaton <jwe@octave.org> parents: 12578 diff changeset	109 ## PKG_ADD: ## Discard result to avoid polluting workspace with ans at startup.
b9a89ca0fb75 prevent optimization functions from setting ans in workspace at startup John W. Eaton <jwe@octave.org> parents: 12578 diff changeset	110 ## PKG_ADD: [~] = __all_opts__ ("fminunc");
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	111
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	112 function [x, fval, info, output, grad, hess] = fminunc (fcn, x0, options = struct ())
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	113
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	114 ## Get default options if requested.
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	115 if (nargin == 1 && strcmp (fcn, "defaults"))
26138 804e18e3e320 Reenable query of optimization options (bugs #54952 and #55089). Kai T. Ohlhus <k.ohlhus@gmail.com> parents: 26137 diff changeset	116 x = struct ("AutoScaling", "off", "FunValCheck", "off",
804e18e3e320 Reenable query of optimization options (bugs #54952 and #55089). Kai T. Ohlhus <k.ohlhus@gmail.com> parents: 26137 diff changeset	117 "FinDiffType", "forward", "GradObj", "off",
804e18e3e320 Reenable query of optimization options (bugs #54952 and #55089). Kai T. Ohlhus <k.ohlhus@gmail.com> parents: 26137 diff changeset	118 "MaxFunEvals", [], "MaxIter", 400, "OutputFcn", [],
804e18e3e320 Reenable query of optimization options (bugs #54952 and #55089). Kai T. Ohlhus <k.ohlhus@gmail.com> parents: 26137 diff changeset	119 "TolFun", 1e-6, "TolX", 1e-6, "TypicalX", []);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	120 return;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	121 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	122
21171 2935d56203a4 Fix regressions caused by ismatrix definition change (partial fix bug #47036). Rik <rik@octave.org> parents: 20165 diff changeset	123 if (nargin < 2 \|\| nargin > 3 \|\| ! isnumeric (x0))
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	124 print_usage ();
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	125 endif
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	126
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	127 if (ischar (fcn))
26118 7502fce4cd3a str2func: eliminate optional second "global" argument John W. Eaton <jwe@octave.org> parents: 25054 diff changeset	128 fcn = str2func (fcn);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	129 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	130
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	131 xsz = size (x0);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	132 n = numel (x0);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	133
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	134 has_grad = strcmpi (optimget (options, "GradObj", "off"), "on");
26137 1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	135 cdif = strcmpi (optimget (options, "FinDiffType", "forward"), "central");
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	136 maxiter = optimget (options, "MaxIter", 400);
26137 1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	137 maxfev = optimget (options, "MaxFunEvals", 100*n);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	138 outfcn = optimget (options, "OutputFcn");
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	139
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	140 ## Get scaling matrix using the TypicalX option. If set to "auto", the
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	141 ## scaling matrix is estimated using the Jacobian.
10201 5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	142 typicalx = optimget (options, "TypicalX");
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	143 if (isempty (typicalx))
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	144 typicalx = ones (n, 1);
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	145 endif
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	146 autoscale = strcmpi (optimget (options, "AutoScaling", "off"), "on");
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	147 if (! autoscale)
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	148 dg = 1 ./ typicalx;
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	149 endif
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	150
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	151 funvalchk = strcmpi (optimget (options, "FunValCheck", "off"), "on");
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	152
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	153 if (funvalchk)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	154 ## Replace fcn with a guarded version.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	155 fcn = @(x) guarded_eval (fcn, x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	156 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	157
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	158 ## These defaults are rather stringent. I think that normally, user
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	159 ## prefers accuracy to performance.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	160
10201 5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	161 tolx = optimget (options, "TolX", 1e-7);
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	162 tolf = optimget (options, "TolFun", 1e-7);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	163
9623 bc0739d02724 update initial TR step for fsolve and fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9464 diff changeset	164 factor = 0.1;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	165 ## FIXME: TypicalX corresponds to user scaling (???)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	166 autodg = true;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	167
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	168 niter = 1;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	169 nfev = 0;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	170
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	171 x = x0(:);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	172 info = 0;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	173
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	174 ## Initial evaluation.
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	175 fval = fcn (reshape (x, xsz));
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	176 n = length (x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	177
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	178 if (! isempty (outfcn))
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	179 optimvalues.iter = niter;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	180 optimvalues.funccount = nfev;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	181 optimvalues.fval = fval;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	182 optimvalues.searchdirection = zeros (n, 1);
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	183 state = "init";
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	184 stop = outfcn (x, optimvalues, state);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	185 if (stop)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	186 info = -1;
22785 9c6661004167 error if break statement is in script file separate from loop (bug #39168) John W. Eaton <jwe@octave.org> parents: 22755 diff changeset	187 return;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	188 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	189 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	190
21099 52af4092f863 For optimization scripts, correctly choose tolerance (eps) based on class of fun and X0. Rik <rik@octave.org> parents: 20852 diff changeset	191 if (isa (x0, "single") \|\| isa (fval, "single"))
52af4092f863 For optimization scripts, correctly choose tolerance (eps) based on class of fun and X0. Rik <rik@octave.org> parents: 20852 diff changeset	192 macheps = eps ("single");
52af4092f863 For optimization scripts, correctly choose tolerance (eps) based on class of fun and X0. Rik <rik@octave.org> parents: 20852 diff changeset	193 else
52af4092f863 For optimization scripts, correctly choose tolerance (eps) based on class of fun and X0. Rik <rik@octave.org> parents: 20852 diff changeset	194 macheps = eps ("double");
52af4092f863 For optimization scripts, correctly choose tolerance (eps) based on class of fun and X0. Rik <rik@octave.org> parents: 20852 diff changeset	195 endif
52af4092f863 For optimization scripts, correctly choose tolerance (eps) based on class of fun and X0. Rik <rik@octave.org> parents: 20852 diff changeset	196
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	197 nsuciter = 0;
9199 399884c9d4a1 import the step adaptation strategy from fsolve to fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9084 diff changeset	198 lastratio = 0;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	199
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	200 grad = [];
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	201
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	202 ## Outer loop.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	203 while (niter < maxiter && nfev < maxfev && ! info)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	204
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	205 grad0 = grad;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	206
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	207 ## Calculate function value and gradient (possibly via FD).
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	208 if (has_grad)
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	209 [fval, grad] = fcn (reshape (x, xsz));
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	210 grad = grad(:);
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	211 nfev += 1;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	212 else
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	213 grad = __fdjac__ (fcn, reshape (x, xsz), fval, typicalx, cdif)(:);
9212 6feb27c38da1 support central differences in fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9207 diff changeset	214 nfev += (1 + cdif) * length (x);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	215 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	216
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	217 if (niter == 1)
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	218 ## Initialize by identity matrix.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	219 hesr = eye (n);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	220 else
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	221 ## Use the damped BFGS formula.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	222 y = grad - grad0;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	223 sBs = sumsq (w);
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	224 Bs = hesr' * w;
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	225 sy = y' * s;
9633 ecc2c556f844 simplify damped BFGS formula Jaroslav Hajek <highegg@gmail.com> parents: 9631 diff changeset	226 theta = 0.8 / max (1 - sy / sBs, 0.8);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	227 r = theta * y + (1-theta) * Bs;
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	228 hesr = cholupdate (hesr, r / sqrt (s' * r), "+");
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	229 [hesr, info] = cholupdate (hesr, Bs / sqrt (sBs), "-");
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	230 if (info)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	231 hesr = eye (n);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	232 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	233 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	234
10201 5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	235 if (autoscale)
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	236 ## Second derivatives approximate the Hessian.
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	237 d2f = norm (hesr, "columns").';
10201 5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	238 if (niter == 1)
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	239 dg = d2f;
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	240 else
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	241 ## FIXME: maybe fixed lower and upper bounds?
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	242 dg = max (0.1*dg, d2f);
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	243 endif
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	244 endif
5c66978f3fdf support TypicalX and AutoScaling in fsolve/fminunc, don't autoscale by default Jaroslav Hajek <highegg@gmail.com> parents: 9899 diff changeset	245
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	246 if (niter == 1)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	247 xn = norm (dg .* x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	248 ## FIXME: something better?
9628 73e6ad869f08 further correct initial TR step strategy Jaroslav Hajek <highegg@gmail.com> parents: 9627 diff changeset	249 delta = factor * max (xn, 1);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	250 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	251
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	252 ## FIXME: why tolfnxn? If abs (e) ~ abs(x) * eps is a vector
21751 b571fc85953f maint: Use two spaces after period to indicate sentence break. Rik <rik@octave.org> parents: 21578 diff changeset	253 ## of perturbations of x, then norm (hesre) <= epsxn, i.e., by
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	254 ## tolf ~ eps we demand as much accuracy as we can expect.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	255 if (norm (grad) <= tolfnxn)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	256 info = 1;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	257 break;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	258 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	259
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	260 suc = false;
9207 25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	261 decfac = 0.5;
25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	262
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	263 ## Inner loop.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	264 while (! suc && niter <= maxiter && nfev < maxfev && ! info)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	265
9631 00958d0c4e3c split __dogleg__ > __doglegm__ Jaroslav Hajek <highegg@gmail.com> parents: 9628 diff changeset	266 s = - __doglegm__ (hesr, grad, dg, delta);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	267
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	268 sn = norm (dg .* s);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	269 if (niter == 1)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	270 delta = min (delta, sn);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	271 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	272
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	273 fval1 = fcn (reshape (x + s, xsz))(:);
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	274 nfev += 1;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	275
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	276 if (fval1 < fval)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	277 ## Scaled actual reduction.
19833 9fc020886ae9 maint: Clean up m-files to follow Octave coding conventions. Rik <rik@octave.org> parents: 19697 diff changeset	278 actred = (fval - fval1) / (abs (fval1) + abs (fval));
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	279 else
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	280 actred = -1;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	281 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	282
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	283 w = hesr * s;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	284 ## Scaled predicted reduction, and ratio.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	285 t = 1/2 * sumsq (w) + grad'*s;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	286 if (t < 0)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	287 prered = -t/(abs (fval) + abs (fval + t));
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	288 ratio = actred / prered;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	289 else
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	290 prered = 0;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	291 ratio = 0;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	292 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	293
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	294 ## Update delta.
14868 5d3a684236b0 maint: Use Octave coding conventions for cuddling parentheses in scripts directory Rik <octave@nomad.inbox5.com> parents: 14552 diff changeset	295 if (ratio < min (max (0.1, 0.8*lastratio), 0.9))
9207 25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	296 delta *= decfac;
25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	297 decfac ^= 1.4142;
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	298 if (delta <= 10machepsxn)
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	299 ## Trust region became uselessly small.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	300 info = -3;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	301 break;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	302 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	303 else
9199 399884c9d4a1 import the step adaptation strategy from fsolve to fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9084 diff changeset	304 lastratio = ratio;
9207 25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	305 decfac = 0.5;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	306 if (abs (1-ratio) <= 0.1)
9207 25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	307 delta = 1.4142*sn;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	308 elseif (ratio >= 0.5)
9207 25f50d2d76b3 improve TR updating strategy for fminunc and fsolve Jaroslav Hajek <highegg@gmail.com> parents: 9199 diff changeset	309 delta = max (delta, 1.4142*sn);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	310 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	311 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	312
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	313 if (ratio >= 1e-4)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	314 ## Successful iteration.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	315 x += s;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	316 xn = norm (dg .* x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	317 fval = fval1;
20735 418ae0cb752f Replace ++,-- with in-place operators for performance. Rik <rik@octave.org> parents: 20715 diff changeset	318 nsuciter += 1;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	319 suc = true;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	320 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	321
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	322 niter += 1;
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	323
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	324 ## FIXME: should outputfcn be called only after a successful iteration?
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	325 if (! isempty (outfcn))
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	326 optimvalues.iter = niter;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	327 optimvalues.funccount = nfev;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	328 optimvalues.fval = fval;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	329 optimvalues.searchdirection = s;
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	330 state = "iter";
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	331 stop = outfcn (x, optimvalues, state);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	332 if (stop)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	333 info = -1;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	334 break;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	335 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	336 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	337
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	338 ## Tests for termination conditions. A mysterious place, anything
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	339 ## can happen if you change something here...
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	340
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	341 ## The rule of thumb (which I'm not sure M*b is quite following)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	342 ## is that for a tolerance that depends on scaling, only 0 makes
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	343 ## sense as a default value. But 0 usually means uselessly long
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	344 ## iterations, so we need scaling-independent tolerances wherever
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	345 ## possible.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	346
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	347 ## The following tests done only after successful step.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	348 if (ratio >= 1e-4)
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	349 ## This one is classic. Note that we use scaled variables again,
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	350 ## but compare to scaled step, so nothing bad.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	351 if (sn <= tolx*xn)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	352 info = 2;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	353 ## Again a classic one.
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	354 elseif (actred < tolf)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	355 info = 3;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	356 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	357 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	358
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	359 endwhile
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	360 endwhile
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	361
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	362 ## When info != 1, recalculate the gradient and Hessian using the final x.
20694 a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	363 if (nargout > 4 && (info == -1 \|\| info == 2 \|\| info == 3))
20715 5b7643257978 Remove trailing whitespace at end of lines. Rik <rik@octave.org> parents: 20695 diff changeset	364 grad0 = grad;
20694 a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	365 if (has_grad)
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	366 [fval, grad] = fcn (reshape (x, xsz));
20694 a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	367 grad = grad(:);
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	368 else
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	369 grad = __fdjac__ (fcn, reshape (x, xsz), fval, typicalx, cdif)(:);
20694 a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	370 endif
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	371
20694 a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	372 if (nargout > 5)
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	373 ## Use the damped BFGS formula.
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	374 y = grad - grad0;
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	375 sBs = sumsq (w);
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	376 Bs = hesr' * w;
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	377 sy = y' * s;
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	378 theta = 0.8 / max (1 - sy / sBs, 0.8);
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	379 r = theta * y + (1-theta) * Bs;
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	380 hesr = cholupdate (hesr, r / sqrt (s' * r), "+");
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	381 hesr = cholupdate (hesr, Bs / sqrt (sBs), "-");
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	382 endif
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	383 ## Return the gradient in the same shape as x
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	384 grad = reshape (grad, xsz);
20694 a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	385 endif
a7dbc4fc3732 fminunc.m: Correctly calculate gradient and hessian even when fminunc fails (bug #40101) José Vallet <jose.vallet@aalto.fi> parents: 20165 diff changeset	386
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	387 ## Restore original shapes.
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	388 x = reshape (x, xsz);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	389
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	390 if (nargout > 3)
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	391 output.iterations = niter;
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	392 output.successful = nsuciter;
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	393 output.funcCount = nfev;
6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	394 endif
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	395
9627 5bcfa0b346e8 fix extra outputs in fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9623 diff changeset	396 if (nargout > 5)
5bcfa0b346e8 fix extra outputs in fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9623 diff changeset	397 hess = hesr'*hesr;
5bcfa0b346e8 fix extra outputs in fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9623 diff changeset	398 endif
5bcfa0b346e8 fix extra outputs in fminunc Jaroslav Hajek <highegg@gmail.com> parents: 9623 diff changeset	399
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	400 endfunction
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	401
18609 923614060f1d fminunc.m: Improve documentation. Rik <rik@octave.org> parents: 18608 diff changeset	402 ## A helper function that evaluates a function and checks for bad results.
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	403 function [fx, gx] = guarded_eval (fun, x)
21758 ffad2baa90f7 maint: Use newlines to make code more readable. Rik <rik@octave.org> parents: 21751 diff changeset	404
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	405 if (nargout > 1)
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	406 [fx, gx] = fun (x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	407 else
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	408 fx = fun (x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	409 gx = [];
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	410 endif
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	411
14383 07c55bceca23 Fix guarded_eval() subfunction in fminunc (bug #35534). Olaf Till <olaf.till@uni-jena.de> parents: 14138 diff changeset	412 if (! (isreal (fx) && isreal (gx)))
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	413 error ("fminunc:notreal", "fminunc: non-real value encountered");
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	414 elseif (any (isnan (fx(:))))
11587 c792872f8942 all script files: untabify and strip trailing whitespace John W. Eaton <jwe@octave.org> parents: 11523 diff changeset	415 error ("fminunc:isnan", "fminunc: NaN value encountered");
14386 59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	416 elseif (any (isinf (fx(:))))
59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	417 error ("fminunc:isinf", "fminunc: Inf value encountered");
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	418 endif
21758 ffad2baa90f7 maint: Use newlines to make code more readable. Rik <rik@octave.org> parents: 21751 diff changeset	419
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	420 endfunction
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	421
14386 59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	422
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	423 %!function f = __rosenb__ (x)
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	424 %! n = length (x);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	425 %! f = sumsq (1 - x(1:n-1)) + 100 * sumsq (x(2:n) - x(1:n-1).^2);
13305 63463570d9fe Add %!endfunction block keyword to test.m Rik <octave@nomad.inbox5.com> parents: 13027 diff changeset	426 %!endfunction
19833 9fc020886ae9 maint: Clean up m-files to follow Octave coding conventions. Rik <rik@octave.org> parents: 19697 diff changeset	427 %!
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	428 %!test
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	429 %! [x, fval, info, out] = fminunc (@__rosenb__, [5, -5]);
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	430 %! tol = 2e-5;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	431 %! assert (info > 0);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	432 %! assert (x, ones (1, 2), tol);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	433 %! assert (fval, 0, tol);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	434 %!test
20695 6faaab833605 fminunc.m: Clean up function to meet Octave coding standards. Rik <rik@octave.org> parents: 20694 diff changeset	435 %! [x, fval, info, out] = fminunc (@__rosenb__, zeros (1, 4));
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	436 %! tol = 2e-5;
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	437 %! assert (info > 0);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	438 %! assert (x, ones (1, 4), tol);
b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	439 %! assert (fval, 0, tol);
19833 9fc020886ae9 maint: Clean up m-files to follow Octave coding conventions. Rik <rik@octave.org> parents: 19697 diff changeset	440
9fc020886ae9 maint: Clean up m-files to follow Octave coding conventions. Rik <rik@octave.org> parents: 19697 diff changeset	441 ## Test FunValCheck works correctly
26137 1ae11ca7dceb fminunc.m: Change algorithm defaults to match Matlab. Rik <rik@octave.org> parents: 26118 diff changeset	442 %!assert (fminunc (@(x) x^2, 1, optimset ("FunValCheck", "on")), 0, 1e-6)
14386 59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	443 %!error <non-real value> fminunc (@(x) x + i, 1, optimset ("FunValCheck", "on"))
59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	444 %!error <NaN value> fminunc (@(x) x + NaN, 1, optimset ("FunValCheck", "on"))
59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	445 %!error <Inf value> fminunc (@(x) x + Inf, 1, optimset ("FunValCheck", "on"))
59aab666f2bf Extend "FunValCheck" option to optimization routines to detect Inf values. Rik <octave@nomad.inbox5.com> parents: 14383 diff changeset	446
9084 b7210faa3ed0 implement fminunc using factored trust-region BFGS Jaroslav Hajek <highegg@gmail.com> parents: diff changeset	447
9899 9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	448 ## Solve the double dogleg trust-region minimization problem:
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	449 ## Minimize 1/2norm(rx)^2 subject to the constraint norm(d.*x) <= delta,
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	450 ## x being a convex combination of the gauss-newton and scaled gradient.
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	451
21578 683a1beee538 maint: Use "FIXME:" for all code blocks needing further attention. Rik <rik@octave.org> parents: 21546 diff changeset	452 ## FIXME: error checks
683a1beee538 maint: Use "FIXME:" for all code blocks needing further attention. Rik <rik@octave.org> parents: 21546 diff changeset	453 ## FIXME: handle singularity, or leave it up to mldivide?
9899 9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	454
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	455 function x = __doglegm__ (r, g, d, delta)
21758 ffad2baa90f7 maint: Use newlines to make code more readable. Rik <rik@octave.org> parents: 21751 diff changeset	456
9899 9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	457 ## Get Gauss-Newton direction.
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	458 b = r' \ g;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	459 x = r \ b;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	460 xn = norm (d .* x);
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	461 if (xn > delta)
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	462 ## GN is too big, get scaled gradient.
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	463 s = g ./ d;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	464 sn = norm (s);
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	465 if (sn > 0)
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	466 ## Normalize and rescale.
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	467 s = (s / sn) ./ d;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	468 ## Get the line minimizer in s direction.
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	469 tn = norm (r*s);
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	470 snm = (sn / tn) / tn;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	471 if (snm < delta)
10549 95c3e38098bf Untabify .m scripts Rik <code@nomad.inbox5.com> parents: 10201 diff changeset	472 ## Get the dogleg path minimizer.
9899 9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	473 bn = norm (b);
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	474 dxn = delta/xn; snmd = snm/delta;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	475 t = (bn/sn) * (bn/xn) * snmd;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	476 t -= dxn * snmd^2 - sqrt ((t-dxn)^2 + (1-dxn^2)*(1-snmd^2));
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	477 alpha = dxn*(1-snmd^2) / t;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	478 else
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	479 alpha = 0;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	480 endif
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	481 else
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	482 alpha = delta / xn;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	483 snm = 0;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	484 endif
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	485 ## Form the appropriate convex combination.
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	486 x = alpha * x + ((1-alpha) * min (snm, delta)) * s;
9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	487 endif
21758 ffad2baa90f7 maint: Use newlines to make code more readable. Rik <rik@octave.org> parents: 21751 diff changeset	488
9899 9f25290a35e8 more private function and subfunction changes John W. Eaton <jwe@octave.org> parents: 9758 diff changeset	489 endfunction