interp2 from 6x6 map to 500x500 map

2 vues (au cours des 30 derniers jours)
anto
anto le 28 Jan 2023
Commenté : anto le 30 Jan 2023
I have this matrix which represents a profile of velocity which my model gave as an output:
V = [55.4710 58.8408 64.2922 64.3100 58.8997 55.4584;...
31.4855 33.5256 39.2656 39.2657 33.5190 31.4851;...
32.2817 38.7198 44.8627 44.8602 38.7216 32.2812;...
32.3167 38.7370 44.8789 44.8765 38.7389 32.3162;...
31.5531 33.4961 39.2318 39.2320 33.4895 31.5529;...
55.7294 59.0097 64.4591 64.4769 59.0688 55.7165];
which physically represents nodal values in a mesh, which visualized is :
the coordinates of the vertical and horizontal line forming the mesh are:
N_mesh = 6;
dx = 1*(0:1/N_mesh:1);
dy = 1*(0:1/N_mesh:1);
I want to use the interpolation "interp2" with the "nearest" option, to create a map for example 500x500.
I looked for documentation but can't seem to use the correct things to successfully do what I have in mind.
For example I tried using 'meshgrid', but since the dx has 7 points, and the map has 6 points the image is not centered...
I tried this:
dx = 1*(0:1/(N_mesh-1):1); %defining nodal points
dy = 1*(0:1/(N_mesh-1):1);
[X,Y] = meshgrid(dx,dy);
Xq = linspace(0,1,500); %defining points for the bigger map
Yq = linspace(0,1,500);
[Xq,Yq] = meshgrid(Xq,Yq);
V_500x500 = interp2(X,Y,V,Xq,Yq,"nearest"); %interpolating
figure() %visualizing the result
contourf(V_500x500)
colorbar
title('interp2 output')
But the image is not centered at the extremes...
the problem is that the original dx are dx = 1*(0:1/N_mesh:1); , but for using meshgrid I have to use dx = 1*(0:1/(N_mesh-1):1);
I just wanted the first image I attached being in the bigger map.
Does anybody know how to get the centered image with interp2?
Any help would be greatly appreciated.

Réponse acceptée

Jan
Jan le 28 Jan 2023
Of course, 1*(0:1/N_mesh:1) has N_mesh+1 points.
"But the image is not centered at the extremes..." - What does this mean?
By the way, you can omit meshgrid and work with index vectors directly:
V = [55.4710 58.8408 64.2922 64.3100 58.8997 55.4584;...
31.4855 33.5256 39.2656 39.2657 33.5190 31.4851;...
32.2817 38.7198 44.8627 44.8602 38.7216 32.2812;...
32.3167 38.7370 44.8789 44.8765 38.7389 32.3162;...
31.5531 33.4961 39.2318 39.2320 33.4895 31.5529;...
55.7294 59.0097 64.4591 64.4769 59.0688 55.7165];
vx = 1:6;
wx = linspace(1, 7, 501);
wx = floor(wx(1:end-1));
W = interp2(vx, vx.', V, wx, wx.');
figure;
vy = linspace(0, 1, numel(vx));
imagesc(vy, vy.', V);
figure;
wy = linspace(0, 1, numel(wx));
imagesc(wy, wy, W);
  1 commentaire
anto
anto le 30 Jan 2023
OK, yes I needed the matrix which you called W, a "centered" (in the sense of properly rescale the 6x6 map) 500x500 matrix in which I can use it for further analysys. Thanks a lot

Connectez-vous pour commenter.

Plus de réponses (0)

Catégories

En savoir plus sur Interpolating Gridded Data dans Help Center et File Exchange

Produits


Version

R2022b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by