testsettings.m

Go to the documentation of this file.

namespace testing{


/* (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.
 */

function s = testsettings() {

s = struct;

/* % Test sizes
 * What dimension? */
s.testdim= 10;

/*  How many params? (to use from the ones defined below) */
s.testparams= 2;
s.testinputs= 1;

/* % Model settings */
s.m= models.BaseFullModel;
s.m.T= 3;
s.m.dt= .2;
s.m.Sampler= sampling.RandomSampler;
s.m.Sampler.Samples= 10;
s.m.System.MaxTimestep= s.m.dt;
s.m.ODESolver= solvers.ExplEuler;

a = approx.KernelApprox(s.m.System);
ap = approx.algorithms.VKOGA;
ec = kernels.config.ParamTimeExpansionConfig;
ec.StateConfig= kernels.config.GaussConfig(" G ",1:2);
ec.ParamConfig= kernels.config.GaussConfig(" G ",.1:.1:.3);
ap.ExpConfig= ec;
ap.MaxExpansionSize= 5; /*  Keep test run short! */

a.Algorithm= ap;
/* a.CoeffComp = general.regression.ScalarEpsSVR; */
s.TimeKernel= kernels.NoKernel;
/* a.TimeKernel = kernels.GaussKernel(2); */
s.Kernel= kernels.GaussKernel(2);
s.ParamKernel= kernels.GaussKernel(2);
s.m.Approx= a;

/* % Dynamical System settings/functions */
s.Inputs[1] = @(t)1; /*  Function 1: Constant 1 */

s.Inputs[2] = @(t)sin(4*t); /*  Function 2: some sin(t) */

s.Inputs= s.Inputs(1:s.testinputs);

/*  Used Parameter Space */
s.params(1) = struct(" Name ", " P1 ", " MinVal ", -1, " MaxVal ", 1, " Desired ", 10, " Default ", 0);
s.params(2) = struct(" Name ", " P2 ", " MinVal ", 2, " MaxVal ", 3, " Desired ", 5, " Default ", 2.5);
s.params(3) = struct(" Name ", " P3 ", " MinVal ", 0, " MaxVal ", 10, " Desired ", 6, " Default ", 5);
s.params= s.params(1:s.testparams);

/*  Input conversion */
muidx = randi(s.testparams,s.testdim,1);
s.B= @(t,mu)ones(s.testdim,1);
s.B_p= @(t,mu)ones(s.testdim,1).*mu(muidx);

/*  Initials */
musel = (rand(s.testdim,1)<.5);
s.x0= dscomponents.ConstInitialValue(rand(s.testdim,1));
av = dscomponents.AffineInitialValue;
av.addMatrix([" mu( " num2str(randi(s.testparams)) " ) "],musel .* rand(s.testdim,1));
av.addMatrix(" .5 + mu(1) ",5*rand(s.testdim,1));
s.x0_p= av;

/*  Linear functions */
muidx = randi(s.testparams,s.testdim,1);
A = rand(s.testdim,s.testdim);
s.flin= @(x,t,mu)A*x;
s.flin_p= @(x,t,mu)(A + diag(mu(muidx)))*x;

/*  Nonlinear functions */
randx = randi(s.testdim);
randmu = randi(s.testparams);
s.fnlin_p= @(x,t,mu)(.5+t/2)*sin(x) + x(randx)*mu(randmu);
s.fnlin= @(x,t,mu)(.5+t/2).*sin(x);

end


}
};