Tests.m

Go to the documentation of this file.

namespace models{
namespace burgers{


/* (Autoinserted by mtoc++)
 * This source code has been filtered by the mtoc++ executable,
 * which generates code that can be processed by the doxygen documentation tool.
 *
 * On the other hand, it can neither be interpreted by MATLAB, nor can it be compiled with a C++ compiler.
 * Except for the comments, the function bodies of your M-file functions are untouched.
 * Consequently, the FILTER_SOURCE_FILES doxygen switch (default in our Doxyfile.template) will produce
 * attached source files that are highly readable by humans.
 *
 * Additionally, links in the doxygen generated documentation to the source code of functions and class members refer to
 * the correct locations in the source code browser.
 * However, the line numbers most likely do not correspond to the line numbers in the original MATLAB source files.
 */

class Tests {
    public: /* ( Static ) */


        static function d = getMoRePasDir() {
            d = fullfile(KerMor.App.DataDirectory," test_Burgers_MoRePaS ");
            if exist(d," file ") ~= 7
                [~,~] = mkdir(d);
            end
        }


        static function  createMoRePaSIIPlots(models.Burgers.Burgers m,colvec<double> mu) {
            d = models.burgers.Tests.getMoRePasDir;

/*              types = {'fig','pdf','jpg'};
 *             types = {'png','jpg'}; */
            types = [" jpg "];
            pm = PlotManager;
            pm.NoTitlesOnSave= true;
            pm.SingleSize= [720 540];
            pm.LeaveOpen= false;
            pm.UseFileTypeFolders= true;

            /* % Different dimension m' plots (model independent)
 *             load(fullfile(d,'minreqmdashorders.mat'));
 *             models.burgers.Tests.morepas12_plotDiffMeanReqErrs([6 7],pm,v500,v1500,v500_maxorder,v1500_maxo);
 *             h = legend(gca(pm.Figures(1)),'d=500','d=1500','d=500 (maxo)','d=1500 (maxo)');
 *             set(h,'Location','Best');
 *             h = legend(gca(pm.Figures(2)),'d=500','d=1500','d=500 (maxo)','d=1500 (maxo)');
 *             set(h,'Location','Best');
 *             pm.done;
 *             pm.FilePrefix = 'burgers';
 *             pm.savePlots(d,types,[],true);
 *             return; */

            /* % Model preparations
             *m = r.FullModel; */
            m.Approx.Order= [12 5];
            r = m.buildReducedModel(100);

            if nargin < 2
                mu = .0400574;
            end
            inputidx = 1;

            /* % Simulation pics
 *             m.PlotAzEl = [-49 34];
 *             [~,y] = m.simulate(mu, 1);
 *             [~,y1] = m.simulate(m.Data.ParamSamples(:,1), 1);
 *             [~,y2] = m.simulate(m.Data.ParamSamples(:,end), 1);
 *             r.ErrorEstimator.Enabled = false;
 *             [t,yr] = r.simulate(mu, 1);
 *             r.ErrorEstimator.Enabled = true;
 *             
 *             pm.FilePrefix = [m.SaveTag '_sim'];
 *             m.plot(t, y1, pm.nextPlot('full_sol_mumin',...
 *                 sprintf('Full solution for \\mu=%g',m.Data.ParamSamples(:,1))));
 *             
 *             m.plot(t, y2, pm.nextPlot('full_sol_mumax',...
 *                 sprintf('Full solution for \\mu=%g',m.Data.ParamSamples(:,end))));
 *             %view(-24,44);
 *             m.plot(t, y1-y2, pm.nextPlot('full_sol_muvar',...
 *                 sprintf('Model variance over parameter range [%g, %g]',...
 *                 m.Data.ParamSamples(:,1),m.Data.ParamSamples(:,end))));
 *             %view(-26,22);
 *             m.plot(t, y, pm.nextPlot('full_sol',...
 *                 sprintf('Full solution for \\mu=%g',mu)));
 *             m.plot(t, y, pm.nextPlot('red_sol',...
 *                 sprintf('Reduced solution for \\mu=%g',mu)));
 *             m.plot(t, y-yr, pm.nextPlot('abs_err','Absolute error'));
 * %             m.plot(t, abs((y-yr)./y), pm.nextPlot('rel_err','Relative error'));
 *             pm.done;
 *             pm.savePlots(d,types,[]); */

            /*  Plot for different M,M' values
 *             pm.FilePrefix = sprintf('%s_m_mdash_errors',m.SaveTag);
 *             savefile = fullfile(d,pm.FilePrefix);
 *             if exist([savefile '.mat'],'file')
 *                 load([savefile '.mat']);
 *             else
 *                 errordata = testing.DEIM.computeDEIMErrors(...
 *                     m.Approx, m.Data.ApproxTrainData);
 *                 relerrs = [1e-1 1e-2 1e-4 1e-6];
 *                 [t, valtrue] = testing.DEIM.getMinReqErrorOrdersTable(errordata, relerrs, m.Data.ApproxTrainData.xi.m);
 *                 t.saveToFile([savefile '.tex']);
 *                 [tmaxo, valmaxorder] = testing.DEIM.getMinReqErrorOrdersTable(errordata, relerrs, m.Data.ApproxTrainData.xi.m, m.Approx.MaxOrder);
 *                 tmaxo.saveToFile([savefile '_true.tex']);
 *                 save([savefile '.mat'],'errordata', 'relerrs', 't', 'tmaxo','valtrue','valmaxorder');
 *             end
 *             
 *             t.display;
 *             tmaxo.display;
 *             testing.DEIM.plotDEIMErrs(errordata, pm);
 *             figure(2); view(28,12);
 *             figure(6); view(-34,46);
 *             pm.done;
 *             pm.savePlots(d,types,[1 2 6],true); */

            /* % Estimator analysis
 *             ea = EstimatorAnalyzer(r);
 *             ea.LineWidth = 2;
 *             ea.MarkerSize = 12;
 *             ea.NumMarkers = 4;
 *             ea.PlotStartIndex = 4;
 *             e = r.ErrorEstimator; */

            /* % Error estimator check for M,M' DEIM approx error est ---------------------------
             * Using true log lip const
 *             e.UseTrueLogLipConst = true;
 *             
 *             est = ea.getDefaultEstStruct;
 *             est(end+1).Name = 'Ref: True local const'; % Expensive versions
 *             est(end).Estimator = e.clone;
 *             est(end).Estimator.UseTrueDEIMErr = true;
 *             est(end).MarkerStyle = 'p';
 *             est(end).LineStyle = '-';
 *             est(end).Color = [0 0.5 0];
 *             ref1 = est(end);
 *             est = testing.DEIM.getDEIMEstimators_ErrOrders(r,est,[1 2 3 5 10]);
 *             ea.Est = est;
 *             ea.SaveTexTables = fullfile(d,'table_mdash.tex');
 *             [errs, relerrs, ctimes] = ea.compute(mu,1);
 *             ea.createPlots(errs, relerrs, ctimes, pm);
 *             pm.createZoom(1,[.7 1 .9*min(errs(:,end)) 1.1*max(errs(:,end))]);
 *             pm.done;
 *             pm.FilePrefix = sprintf('%s_err_mdash_trueloglip',m.SaveTag);
 *             pm.savePlots(d,types,[1 4],true); */

/*              %% Error estimator check for M,M' DEIM approx error est ---------------------------
 *             % Using full jac log norm
 *             e.UseTrueLogLipConst = false;
 *             e.UseFullJacobian = true;
 *             ea.SaveTexTables = fullfile(d,'table_mdash.tex');
 *             
 *             est = ea.getDefaultEstStruct;
 *             est(end+1).Name = 'L_G[J(y)]'; % Expensive versions
 *             est(end).Estimator = e.clone;
 *             est(end).Estimator.UseTrueDEIMErr = true;
 *             est(end).MarkerStyle = 'p';
 *             est(end).LineStyle = '-';
 *             est(end).Color = [.9 .6 0];
 *             ref2 = est(end);
 *             est(end+1) = ref1; % Expensive versions
 *             %est = testing.DEIM.getDEIMEstimators_ErrOrders(r, est, [1 2 3 5 10]);
 *             ea.Est = est;
 *             [errs, relerrs, ctimes] = ea.compute(mu,1);
 *             ea.createPlots(errs, relerrs, ctimes, pm);
 *             pm.createZoom(1,[.7 1 .9*min(errs(:,end)) 1.1*max(errs(:,end))]);
 *             pm.done;
 *             pm.FilePrefix = sprintf('%s_err_mdash_fulljac_lognorm',m.SaveTag);
 *             pm.savePlots(d,types,[1 4],true);
 * 
 *             
 *             %% JacMDEIM & SimTrans error plots ------------------------------------------------
 *             r.System.f.Order = [12 10];
 *             e.UseTrueLogLipConst = false;
 *             e.UseFullJacobian = false;
 *             e.UseTrueDEIMErr = false;
 *             
 *             % Expensive versions
 *             est = ea.getDefaultEstStruct;
 *             ref1.Estimator.UseTrueDEIMErr = false;
 *             %ref1.Name = 'Reference #1 m''';
 *             est(end+1) = ref1;
 *             ref2.Estimator.UseTrueDEIMErr = false;
 *             %ref2.Name = 'Reference #2 m''';
 *             est(end+1) = ref2;
 *             est = testing.DEIM.getDEIMEstimators_MDEIM_ST(r,est,[1 3 10],[1 3 10]);
 *             ea.Est = est;
 *             ea.NumMarkers = 3;
 *             ea.SaveTexTables = fullfile(d,'table_jac_st.tex');
 *             [errs, relerrs, ctimes] = ea.compute(mu,1);
 *             ea.createPlots(errs, relerrs, ctimes, pm);
 *             pm.createZoom(1,[.65 1 .9*min(errs(:,end)) 1.1*max(errs(:,end))]);
 *             pm.done;
 *             pm.FilePrefix = 'err_jac_st';
 *             pm.savePlots(d,types,[1 4],true,[true false]); */
        }
        static function  morepas12_plotDiffMeanReqErrs(indices,pm,varargin) {
            for k = 1:length(indices)
                idx = indices(k);
                h = pm.nextPlot(sprintf(" diffmeanreqerrs_%d ",idx),...
                        " Minimum/Mean required m ""  values for relative errors ",...
                        " DEIM order m "," m ""  values ");
                hold(h," on ");
                n = length(varargin);
                l = LineSpecIterator(n);
                for i=1:n
                    d = varargin[i];
                    d = d(1:138,[1 idx]);
                    sel = round(d(:,1)) == d(:,1);
                    d = d(sel,:);
                    plot(h,d(:,1),d(:,2)," Color ",l.nextColor," LineWidth ",2);
                end
            end
        }


        static function [res , m ] = testBurgers_DEIM_versions(dim,version) {
            if nargin < 2
                version = 1;
                if nargin < 1
                    dim = 20;
                end
            end
            m = models.burgers.Burgers(dim, version);

            if dim > 1000
                m.Data.useFileTrajectoryData;
            end

            /* % Sampling - manual */
            s = sampling.ManualSampler;
            s.Samples= logspace(log10(0.04),log10(0.08),10);
            m.Sampler= s;

            /* % Approx */
            a = approx.DEIM(m.System);
            a.MaxOrder= 40;
            m.Approx= a;

            m.offlineGenerations;
            a.Order= [5 2];
            r = m.buildReducedModel;
            [t,y] = r.simulate(r.getRandomParam);
            m.plot(t,y);
            res = true;
        }


        static function res = test_BurgersModels() {
            models.burgers.Tests.testBurgers_DEIM_versions(50, 1);
            models.burgers.Tests.testBurgers_DEIM_versions(50, 2);
            res = true;
        }

    

};
}
}