MATLAB: Mesh plot with only every nth line plotted

3-d plotmeshnth line

Hello,

is it possible to reduce the number of plotted lines in a mesh-plot without loss of accuracy? I know, I can reduce the number of data points. But in this case the accuracy of the mesh-plot gets quite bad (see exsample below). Is there any way to say "mesh" to show only every nth line of the wireframe?

% [Matlab]
[DataX,DataY,DataZ] = peaks(100);
subplot(1,2,1);
mesh(DataX,DataY,DataZ,'EdgeColor','k');
subplot(1,2,2);
mesh(DataX(1:10:end,1:10:end),DataY(1:10:end,1:10:end),...
    DataZ(1:10:end,1:10:end),'EdgeColor','k');
% [/Matlab]

Best regards

Guido

[Edit] Really bad mistake in writing corrected. [/Edit]

Best Answer

There isn't a nice automatic way to do it, but it's not too hard. You need 3 surfaces. One for the white background, one for the row edges, and one for the column edges. It would look something like this:

rskip = round(linspace(1,size(DataZ,1),32));
cskip = round(linspace(1,size(DataZ,2),32));
surf(DataX,DataY,DataZ,'FaceColor','white','EdgeColor','none');
hold on
surf(DataX(rskip,:),DataY(rskip,:),DataZ(rskip,:),'FaceColor','none','MeshStyle','row');
surf(DataX(:,cskip),DataY(:,cskip),DataZ(:,cskip),'FaceColor','none','MeshStyle','column');
hold off

The reason for this is that you want to reduce the resolution between edges, but not along edges.

Related Solutions

MATLAB: Print ignores z order of stacked contours

Solved it! The trick is to use the contour map, that is returned. Since I only want the outline of the surface, I first plot the contourf with 0 levels and take the contour map from that plot. Makes things a lot easier.

A = peaks(10);
A(A<=0) = nan;
% generate the outline
[C, ~] = contourf(A, 0, 'linewidth', 1);
% plot the contour
[~, h] = contourf(A, 50, 'linecolor', 'none');
colormap(flipud(gray));
% set background to red to make it transparent later
set(gca, 'color', 'red');
% fill off and on to apply the red background to holes
set(h, 'fill', 'off');
set(h, 'fill', 'on');
% obtain all lines of the outline
CC = C;
clear C;
n = 0;
sFrom = 2;
sTo   = 1;
while true
    n = n+1;
    sTo = sFrom+CC(2, sFrom-1)-1;
    C{n} = CC(:, sFrom:sTo);
    sFrom = sTo+2;
    if sTo == size(CC, 2)
        break;
    end % if
end % while
% now plot the outlines into the contour
hold on
for n = 1:numel(C)
    plot(C{n}(1,:), C{n}(2,:), 'k', 'linewidth', 2);
end % for
hold off
% save the figure
print(gcf, '-r300', '-dpng', 'peaks.png');
% use external tool mogrify to make everything in red transparent
system('mogrify -transparent red peaks.png');

This generates the desired result:

MATLAB: How to turn off antialiasing in MATLAB R2014b

Starting in MATLAB R2014b, figures use an antialiasing technique to reduce the appearance of jagged lines. However, the smoothing technique sacrifices some sharpness for smoothness and, in some cases, lines might appear fuzzy. To turn off antialiasing, set the GraphicsSmoothing property for the figure to 'off'. For example:

set(gcf,'GraphicsSmoothing','off')

These images show the difference between two lines, one with GraphicsSmoothing set to on and one with GraphicsSmoothing set to off.

If your graph contains mostly vertical or horizontal lines that appear uneven in thickness, use the AlignVertexCenters property to eliminate the uneven appearance. By default, this property is set to 'off'. To eliminate the uneven appearance, set the AlignVertexCenters property for the lines to 'on'.

For example, create a plot of horizontal lines and set the AlignVertexCenters property for all of the lines to 'on';. Keep the GraphicsSmoothing property set to 'on' (the default).

p = plot([1:10; 1:10],[1 2])
set(p,'AlignVertexCenters','on')

Best Answer

Related Solutions

MATLAB: Print ignores z order of stacked contours

MATLAB: How to turn off antialiasing in MATLAB R2014b

Related Question