Mercurial > repos > public > sbplib
view +noname/calculateErrors.m @ 774:66eb4a2bbb72 feature/grids
Remove default scaling of the system.
The scaling doens't seem to help actual solutions. One example that fails in the flexural code.
With large timesteps the solutions seems to blow up. One particular example is profilePresentation
on the tdb_presentation_figures branch with k = 0.0005
author | Jonatan Werpers <jonatan@werpers.com> |
---|---|
date | Wed, 18 Jul 2018 15:42:52 -0700 |
parents | 1201eb16557e |
children |
line wrap: on
line source
% [discr, trueSolution] = schemeFactory(m) % where trueSolution should be a timeSnapshot of the true solution a time T % T is the end time % m are grid size parameters. % N are number of timesteps to use for each gird size % timeOpt are options for the timeStepper % errorFun is a function_handle taking 2 or 3 arguments, errorFun(trueSolution, approxSolution), errorFun(trueSolution, approxSolution, discr) function e = calculateErrors(schemeFactory, T, m, N, errorFun, timeOpt) %TODO: Ability to choose paralell or not assertType(schemeFactory, 'function_handle'); assertNumberOfArguments(schemeFactory, 1); assertScalar(T); assert(length(m) == length(N), 'Vectors m and N must have the same length'); assertType(errorFun, 'function_handle'); if ~ismember(nargin(errorFun), [2,3]) error('sbplib:noname:calculateErrors:wrongNumberOfArguments', '"%s" must have 2 or 3, found %d', toString(errorFun), nargin(errorFun)); end default_arg('timeOpt', struct()); e = zeros(1,length(m)); parfor i = 1:length(m) done = timeTask('m = %3d ', m(i)); [discr, trueSolution] = schemeFactory(m(i)); timeOptTemp = timeOpt; timeOptTemp.k = T/N(i); ts = discr.getTimestepper(timeOptTemp); ts.stepTo(N(i), true); approxSolution = discr.getTimeSnapshot(ts); switch nargin(errorFun) case 2 e(i) = errorFun(trueSolution, approxSolution); case 3 e(i) = errorFun(trueSolution, approxSolution, discr); end fprintf('e = %.4e', e(i)) done() end fprintf('\n') end %% Example error function % u_true = grid.evalOn(dr.grid, @(x,y)trueSolution(T,x,y)); % err = u_true-u_false; % e(i) = norm(err)/norm(u_true); % % e(i) = sqrt(err'*d.H*d.J*err/(u_true'*d.H*d.J*u_true));