Mercurial > hg > ltpda

comparison m-toolbox/classes/@matrix/tdfit.m @ 0:f0afece42f48

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Import.
author Daniele Nicolodi <nicolodi@science.unitn.it>
date Wed, 23 Nov 2011 19:22:13 +0100
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:f0afece42f48
1 % TDFIT fit a MATRIX of transfer function SMODELs to a matrix of input and output signals.
2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3 %
4 % DESCRIPTION: TDFIT fits a MATRIX of transfer function SMODELs to a set of
5 % input and output signals. It uses ao\tdfit as the core algorithm.
6 %
7 %
8 % CALL: b = tdfit(outputs, pl)
9 %
10 % INPUTS: outputs - an array of MATRIXs representing the outputs of a system,
11 % one per each experiment.
12 % pl - parameter list (see below)
13 %
14 % OUTPUTs: b - a pest object containing the best-fit parameters,
15 % goodness-of-fit reduced chi-squared, fit degree-of-freedom
16 % covariance matrix and uncertainties. Additional
17 % quantities, like the Information Matrix, are contained
18 % within the procinfo. The best-fit model can be evaluated
19 % from pest\eval.
20 %
21 % <a href="matlab:utils.helper.displayMethodInfo('matrix', 'tdfit')">Parameters Description</a>
22 %
23 % EXAMPLES:
24 %
25 % VERSION: $Id: tdfit.m,v 1.9 2011/04/08 08:56:31 hewitson Exp $
26 %
27 % HISTORY: 09-02-2011 G. Congedo
28 %
29 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30
31 function varargout = tdfit(varargin)
32
33 % Check if this is a call for parameters
34 if utils.helper.isinfocall(varargin{:})
35 varargout{1} = getInfo(varargin{3});
36 return
37 end
38
39 import utils.const.*
40 utils.helper.msg(msg.PROC3, 'running %s/%s', mfilename('class'), mfilename);
41
42 % Collect input variable names
43 in_names = cell(size(varargin));
44 for ii = 1:nargin,in_names{ii} = inputname(ii);end
45
46 % Collect all ltpdauoh objects
47 [mtxs, mtxs_invars] = utils.helper.collect_objects(varargin(:), 'matrix', in_names);
48 [pl, invars] = utils.helper.collect_objects(varargin(:), 'plist');
49
50 % combine plists
51 pl = parse(pl, getDefaultPlist());
52
53 if nargout == 0
54 error('### tdfit cannot be used as a modifier. Please give an output variable.');
55 end
56
57 % combine plists
58 pl = parse(pl, getDefaultPlist());
59
60 % Extract necessary parameters
61 inputs = pl.find('inputs');
62 TFmodels = pl.find('models');
63 WhFlts = pl.find('WhFlts');
64 inNames = pl.find('innames');
65 outNames = pl.find('outnames');
66 P0 = pl.find('P0');
67 pnames = pl.find('pnames');
68
69 % Checks
70 if ~isa(mtxs,'matrix')
71 error('### Please, give the system outputs as an array of MATRIXs per each experiment.');
72 end
73 if ~isa(inputs,'collection') && any(~isa(inputs.objs,'matrix'))
74 error('### Please, give the system inputs as a COLLECTION of MATRIXs per each experiment.');
75 end
76 if ~isa(TFmodels,'ssm') && ~isa(TFmodels,'matrix') && any(~isa(TFmodels.objs,'smodel'))
77 error('### Please, give the system transfer functions as a MATRIX of SMODELs, or a single SSM model.');
78 end
79 if ~isa(WhFlts,'matrix') && any(~isa(WhFlts.objs,'filterbank'))
80 error('### Please, give the system inputs as a MATRIX of FILTERBANKs.');
81 end
82
83 % Define constants
84 Nexp = numel(mtxs);
85
86 % Prepare objects for fit
87 outputs2fit = prepare2fit(mtxs,'outputs',Nexp);
88 inputs2fit = prepare2fit(inputs,'inputs',Nexp);
89 WhFlts2fit = prepare2fit(WhFlts,'filters',Nexp);
90 if ~isa(TFmodels,'ssm')
91 models2fit = prepare2fit(TFmodels,'models',Nexp);
92 else
93 models2fit = TFmodels;
94 end
95 inNames2fit = prepare2fit(inNames,'inNames',Nexp);
96 outNames2fit = prepare2fit(outNames,'outNames',Nexp);
97
98 % fit plist
99 fitpl = pl.pset('inputs', inputs2fit,...
100 'models', models2fit,...
101 'WhFlts', WhFlts2fit,...
102 'inNames', inNames2fit,...
103 'outNames', outNames2fit,...
104 'P0', P0,...
105 'pnames', pnames);
106
107 % do fit
108 params = tdfit(outputs2fit, fitpl);
109
110 % Make output pest
111 out = copy(params,1);
112
113 % Set Name and History
114 mdlname = char(TFmodels(1).name);
115 for kk=2:numel(TFmodels)
116 mdlname = strcat(mdlname,[',' char(TFmodels(kk).name)]);
117 end
118 out.name = sprintf('tdfit(%s)', mdlname);
119 out.addHistory(getInfo('None'), pl, mtxs_invars(:), [mtxs(:).hist]);
120
121 % Set outputs
122 if nargout > 0
123 varargout{1} = out;
124 end
125 end
126
127 %--------------------------------------------------------------------------
128 % Included Functions
129 %--------------------------------------------------------------------------
130
131 function obj2fit = prepare2fit(obj,type,Nexp)
132 switch type
133 case 'outputs'
134 obj2fit = obj(1).objs;
135 case 'inputs'
136 obj2fit = obj.objs{1}.objs;
137 case 'filters'
138 obj2fit = obj.objs;
139 case 'inNames'
140 obj2fit = obj;
141 case 'outNames'
142 obj2fit = obj;
143 end
144 if exist('obj2fit','var')
145 if size(obj2fit)~=[numel(obj2fit),1]
146 obj2fit = obj2fit';
147 end
148 for ii=2:Nexp
149 switch type
150 case 'outputs'
151 obj2cat = obj(ii).objs;
152 case 'inputs'
153 obj2cat = obj.objs{ii}.objs;
154 case 'filters'
155 obj2cat = obj.objs;
156 case 'inNames'
157 obj2cat = obj;
158 case 'outNames'
159 obj2cat = obj;
160 end
161 if size(obj2cat)~=[numel(obj2cat),1]
162 obj2cat = obj2cat';
163 end
164 obj2fit = [obj2fit;obj2cat];
165 end
166 end
167 if strcmp(type,'models')
168 obj2fit = smodel();
169 % obj2fit.setXvar();
170 sz = size(obj.objs);
171 obj2fit = repmat(obj2fit,Nexp*sz);
172 for ii=1:Nexp
173 obj2fit((1:sz(1))+(ii-1)*sz(1),(1:sz(2))+(ii-1)*sz(2)) = obj.objs;
174 end
175 end
176 end
177
178 %--------------------------------------------------------------------------
179 % Get Info Object
180 %--------------------------------------------------------------------------
181 function ii = getInfo(varargin)
182 if nargin == 1 && strcmpi(varargin{1}, 'None')
183 sets = {};
184 pl = [];
185 else
186 sets = {'Default'};
187 pl = getDefaultPlist;
188 end
189 % Build info object
190 ii = minfo(mfilename, 'matrix', 'ltpda', utils.const.categories.sigproc, '$Id: tdfit.m,v 1.9 2011/04/08 08:56:31 hewitson Exp $', sets, pl);
191 ii.setModifier(false);
192 end
193
194 %--------------------------------------------------------------------------
195 % Get Default Plist
196 %--------------------------------------------------------------------------
197 function plout = getDefaultPlist()
198 persistent pl;
199 if exist('pl', 'var')==0 || isempty(pl)
200 pl = buildplist();
201 end
202 plout = pl;
203 end
204
205 function pl = buildplist()
206
207 pl = plist();
208
209 % Inputs
210 p = param({'Inputs', 'A COLLECTION of MATRIXs, one per each experiment, containing the input A0s.'}, paramValue.EMPTY_DOUBLE);
211 pl.append(p);
212
213 % Models
214 p = param({'Models', 'A MATRIX of transfer function SMODELs.'}, paramValue.EMPTY_DOUBLE);
215 pl.append(p);
216
217 % PadRatio
218 p = param({'PadRatio', ['PadRatio is defined as the ratio between the number of zero-pad points '...
219 'and the data length.<br>'...
220 'Define how much to zero-pad data after the signal.<br>'...
221 'Being <tt>tdfit</tt> a fft-based algorithm, no zero-padding might bias the estimation, '...
222 'therefore it is strongly suggested to do that.']}, 1);
223 pl.append(p);
224
225 % Whitening Filters
226 p = param({'WhFlts', 'A MATRIX of FILTERBANKs containing the whitening filters per each output AO.'}, paramValue.EMPTY_DOUBLE);
227 pl.append(p);
228
229 % Parameters
230 p = param({'Pnames', 'A cell-array of parameter names to fit.'}, paramValue.EMPTY_CELL);
231 pl.append(p);
232
233 % P0
234 p = param({'P0', 'An array of starting guesses for the parameters.'}, paramValue.EMPTY_DOUBLE);
235 pl.append(p);
236
237 % LB
238 p = param({'LB', ['Lower bounds for the parameters.<br>'...
239 'This improves convergency. Mandatory for Monte Carlo.']}, paramValue.EMPTY_DOUBLE);
240 pl.append(p);
241
242 % UB
243 p = param({'UB', ['Upper bounds for the parameters.<br>'...
244 'This improves the convergency. Mandatory for Monte Carlo.']}, paramValue.EMPTY_DOUBLE);
245 pl.append(p);
246
247 % Algorithm
248 p = param({'ALGORITHM', ['A string defining the fitting algorithm.<br>'...
249 '<tt>fminunc</tt>, <tt>fmincon</tt> require ''Optimization Toolbox'' to be installed.<br>'...
250 '<tt>patternsearch</tt>, <tt>ga</tt>, <tt>simulannealbnd</tt> require ''Genetic Algorithm and Direct Search'' to be installed.<br>']}, ...
251 {1, {'fminsearch', 'fminunc', 'fmincon', 'patternsearch', 'ga', 'simulannealbnd'}, paramValue.SINGLE});
252 pl.append(p);
253
254 % OPTSET
255 p = param({'OPTSET', ['An optimisation structure to pass to the fitting algorithm.<br>'...
256 'See <tt>fminsearch</tt>, <tt>fminunc</tt>, <tt>fmincon</tt>, <tt>optimset</tt>, for details.<br>'...
257 'See <tt>patternsearch</tt>, <tt>psoptimset</tt>, for details.<br>'...
258 'See <tt>ga</tt>, <tt>gaoptimset</tt>, for details.<br>'...
259 'See <tt>simulannealbnd</tt>, <tt>saoptimset</tt>, for details.']}, paramValue.EMPTY_STRING);
260 pl.append(p);
261
262 % SymDiff
263 p = param({'SymDiff', 'Use symbolic derivatives or not. Only for gradient-based algorithm or for LinUnc option.'}, paramValue.NO_YES);
264 pl.append(p);
265
266 % DiffOrder
267 p = param({'DiffOrder', 'Symbolic derivative order. Only for SymDiff option.'}, {1, {1,2}, paramValue.SINGLE});
268 pl.append(p);
269
270 % FitUnc
271 p = param({'FitUnc', 'Fit parameter uncertainties or not.'}, paramValue.YES_NO);
272 pl.append(p);
273
274 % UncMtd
275 p = param({'UncMtd', ['Choose the uncertainties estimation method.<br>'...
276 'For multi-channel fitting <tt>hessian</tt> is mandatory.']}, {1, {'hessian', 'jacobian'}, paramValue.SINGLE});
277 pl.append(p);
278
279 % LinUnc
280 p = param({'LinUnc', 'Force linear symbolic uncertainties.'}, paramValue.YES_NO);
281 pl.append(p);
282
283 % GradSearch
284 p = param({'GradSearch', 'Do a preliminary gradient-based search using the BFGS Quasi-Newton method.'}, paramValue.NO_YES);
285 pl.append(p);
286
287 % MonteCarlo
288 p = param({'MonteCarlo', ['Do a Monte Carlo search in the parameter space.<br>'...
289 'Useful when dealing with high multiplicity of local minima. May be computer-expensive.<br>'...
290 'Note that, if used, P0 will be ignored. It also requires to define LB and UB.']}, paramValue.NO_YES);
291 pl.append(p);
292
293 % Npoints
294 p = param({'Npoints', 'Set the number of points in the parameter space to be extracted.'}, 100000);
295 pl.append(p);
296
297 % Noptims
298 p = param({'Noptims', 'Set the number of optimizations to be performed after the Monte Carlo.'}, 10);
299 pl.append(p);
300
301 end
302 % END