Mercurial > hg > smallbox
view examples/ALPS solvers tests/SMALL_solver_test_ALPS.m @ 199:751fa3bddd30 luisf_dev
Added config file for SMALL_solve (removed the if/else code from SMALL_solve); added headers to both config files;
author | luisf <luis.figueira@eecs.qmul.ac.uk> |
---|---|
date | Tue, 20 Mar 2012 14:28:51 +0000 |
parents | 855025f4c779 |
children |
line wrap: on
line source
%% Example test of solvers from different toolboxes on Sparco problem 6 % % The main purpose of this example is to show how to use SMALL structure % to solve SPARCO compressed sensing problems (1-11) and compare results % from different solvers. % To generate SMALL.Problem part of structure you can use generateProblem % function from Sparco toolbox giving the problem number and any % additional parameters you might want to change. Alternatively, you can % might want to consult sparco documentation to write a problem by % yourself. There are four fields the must be specified in SMALL.Problem % - A, b, sizeA and reconstruct. % % To generate SMALL.solver part of the structure you must specify three % fields: % % SMALL.solver.toolbox - string with toolbox name is needed because % different toolboxes are calling solver % functions in different ways. % SMALL.solver.name - its string representing solver name (e.g. % SolveOMP) % SMALL.solver.param - string that contains optional parameters for % particular solver (all parameters you want to % specify except A, b and size of solution) % % Every call to SMALL_solve function will generate following output: % % SMALL.solver.solution - contains solution vector x % SMALL.solver.reconstructed - vector containing signal reconstructed % from the solution % SMALL.solver.time - time that solver spent to find the solution % % SMALL_plot function plots the SMALL.solver.solution and reconstructed % against original signal. % % In this particular example we are testing SMALL_cgp, SMALL_chol, % SolveOMP form SparseLab and greed_pcgp form Sparsify against "PROB006 % Daubechies basis, Gaussian ensemble measurement basis, piecewise cubic % polynomial signal" from Sparco. % % % Centre for Digital Music, Queen Mary, University of London. % This file copyright 2009 Ivan Damnjanovic. % % This program 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 2 of the % License, or (at your option) any later version. See the file % COPYING included with this distribution for more information. %% fprintf('\n\nExample test of SMALL solvers against their counterparts on Sparco problems.\n\n'); %% % Generate SPARCO problem clear SMALL.Problem = generateProblem(6, 'P', 6, 'm', 2*500,'n',2*1024, 'show'); SMALL.Problem.A = opToMatrix(SMALL.Problem.A, 1); %% i=1; %% % ALPS test SMALL.solver(i) = SMALL_init_solver; SMALL.solver(i).toolbox = 'ALPS'; SMALL.solver(i).name = 'AlgebraicPursuit'; % In the following string all parameters except matrix, measurement vector % and size of solution need to be specified. If you are not sure which % parameters are needed for particular solver type "help <Solver name>" in % MATLAB command line SMALL.solver(i).param=struct(... 'sparsity', 125,... 'memory', 0,... 'mode', 1,... 'iternum', 50,... 'tolerance', 1e-14'); SMALL.solver(i)=SMALL_solve(SMALL.Problem,SMALL.solver(i)); i=i+1; %% % SMALL Conjugate Gradient test SMALL.solver(i)=SMALL_init_solver; SMALL.solver(i).toolbox='SMALL'; SMALL.solver(i).name='SMALL_pcgp'; % In the following string all parameters except matrix, measurement vector % and size of solution need to be specified. If you are not sure which % parameters are needed for particular solver type "help <Solver name>" in % MATLAB command line SMALL.solver(i).param='200, 1e-14'; SMALL.solver(i)=SMALL_solve(SMALL.Problem,SMALL.solver(i)); i=i+1; %% % SolveOMP from SparseLab test SMALL.solver(i)=SMALL_init_solver; SMALL.solver(i).toolbox='SparseLab'; SMALL.solver(i).name='SolveOMP'; % In the following string all parameters except matrix, measurement vector % and size of solution need to be specified. If you are not sure which % parameters are needed for particular solver type "help <Solver name>" in % MATLAB command line SMALL.solver(i).param='200, 0, 0, 0, 1e-14'; SMALL.solver(i)=SMALL_solve(SMALL.Problem, SMALL.solver(i)); i=i+1; %% % SMALL OMP with Cholesky update test SMALL.solver(i)=SMALL_init_solver; SMALL.solver(i).toolbox='SMALL'; SMALL.solver(i).name='SMALL_chol'; % In the following string all parameters except matrix, measurement vector % and size of solution need to be specified. If you are not sure which % parameters are needed for particular solver type "help <Solver name>" in % MATLAB command line SMALL.solver(i).param='200, 1e-14'; SMALL.solver(i)=SMALL_solve(SMALL.Problem, SMALL.solver(i)); i=i+1; %% % greed_pcgp from Sparsify test SMALL.solver(i)=SMALL_init_solver; SMALL.solver(i).toolbox='Sparsify'; SMALL.solver(i).name='greed_pcgp'; % In the following string all parameters except matrix, measurement vector % and size of solution need to be specified. If you are not sure which % parameters are needed for particular solver type "help <Solver name>" in % MATLAB command line SMALL.solver(i).param='''stopCrit'', ''M'', ''stopTol'', 200'; SMALL.solver(i)=SMALL_solve(SMALL.Problem, SMALL.solver(i)); %% SMALL_plot(SMALL);