Requires more input arguments to run
2 views (last 30 days)
Show older comments
I just bought MatLab for a school project a few hours ago, and I've never used this software before. I need to run the code my professor has posted, but the message "'convection_hat' requires more input arguments to run" comes up. Is there something wrong with the code she posted? I'd really appreciated any help, as I'm totally lost at the moment. Here is the code:
function d_theta_hat = convection_hat( t, theta_hat_vec )
global F N tol del_x del_y lap_inv;
% theta_hat = Fourier transform of Theta given in ODE call.
% t = time
% return Fourier Transform of -u \cdot \grad theta
% where v = -\lap \inv P ( \theta \cdot \grad \inv \theta )
% and u = v / \norm{\grad v}_{L^2}
theta_hat = reshape( theta_hat_vec, N, N );
% g = \theta \grad\inv \theta
% To convolve accurately we need to "circularly" convolve, which matlab
% doesn't have in 2D. Instead we ifft, multipliy and fft back.
linv_theta_hat = lap_inv .* theta_hat;
theta = ifft2( theta_hat, 'symmetric' );
g_hat(:, :, 1) = fft2( theta .* ifft2( del_x .* linv_theta_hat, 'symmetric' ) );
g_hat(:, :, 2) = fft2( theta .* ifft2( del_y .* linv_theta_hat, 'symmetric' ) );
% div_g_hat = (\divergence g)^\hat
div_g_hat = del_x .* g_hat(:, :, 1) + del_y .* g_hat( :, :, 2 );
% v = -\lap\inv P g
v_hat(:, :, 1) = -lap_inv .* ( g_hat(:,:,1) - del_x .* lap_inv .* div_g_hat );
v_hat(:, :, 2) = -lap_inv .* ( g_hat(:,:,2) - del_y .* lap_inv .* div_g_hat );
v_hat_norm1 = del_x.*v_hat(:,:,1);
v_hat_norm2 = del_x.*v_hat(:,:,2);
v_hat_norm3 = del_y.*v_hat(:,:,1);
v_hat_norm4 = del_y.*v_hat(:,:,2);
norm_v_hat = sqrt( norm( v_hat_norm1(:), 2 )^2 + ...
norm( v_hat_norm2(:), 2 )^2 + norm( v_hat_norm3(:), 2 )^2 + ...
norm( v_hat_norm4(:), 2 )^2 );
norm_v = norm_v_hat / N^2; % L2 conversion factor for DFT.
% Compare norm_v relative to norm_g
linv_g_hat( :, :, 1 ) = lap_inv .* g_hat(:, :, 1);
linv_g_hat( :, :, 2 ) = lap_inv .* g_hat(:, :, 2);
norm_linv_g = norm( linv_g_hat(:), 2 ) / N^2;
if norm_v < norm_linv_g *tol;
disp( sprintf( ...
'Possible saddle point: t=%.3f, norm_v=%.9f, norm_g=%.9f', ...
t, norm_v, norm_linv_g ) );
end
u(:, :, 1) = F*ifft2( v_hat(:, :, 1), 'symmetric' ) / norm_v;
u(:, :, 2) = F*ifft2( v_hat(:, :, 2), 'symmetric' ) / norm_v;
d_theta_hat = -fft2( ...
u(:,:,1) .* ifft2( del_x.*theta_hat, 'symmetric' ) + ...
u(:,:,2) .* ifft2( del_y.*theta_hat, 'symmetric' ) );
% Finally we need our output to be used in ode23 or ode45 functions, so we
% need reshape it.
d_theta_hat = d_theta_hat(:);
end
3 Comments
per isakson
on 1 Aug 2014
Edited: per isakson
on 1 Aug 2014
Your professor obviously chose not to make it that simple for you. Didn't she provide some additionally instruction?
Accepted Answer
Hikaru
on 1 Aug 2014
What did you type before you get that error message? You need to define the values for t and theta_hat_vec
4 Comments
Hikaru
on 1 Aug 2014
Technically, you also need to know the values for the global variables (F, N, tol, del_x, del_y, and lap_inv) before this code could run.
Good luck with the school project.
More Answers (0)
See Also
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!