Plotting contours in matlab

1 view (last 30 days)
mukesh bisht on 22 Jun 2021
Commented: Bjorn Gustavsson on 22 Jun 2021
Hi
I have a data set (attached) for which I have plotted a scattered plot (Figure2) and a contour plot (Figure1). But, the contour plot doesn't look good, the boundary of contour plot is rough (i.e. not smooth). Please suggest some method to correct the contour plot or some other interpolation method.
My code:
For Scattered plot
x = D(:,1);y = D(:,2);z = D(:,3);pointsize = 10;
figure(2)
scatter(x,y,pointsize,z,'filled')
colormap jet;colorbar
For Contour plot
x = D(:,1);y = D(:,2);z = D(:,3);
xv = linspace((min(x)), max(x)); yv = linspace(min(y), max(y));
[Xm,Ym] = ndgrid(xv, yv);
Zm = griddata(x, y, z, Xm, Ym,'natural');
hold on
figure(1)
contourf(Xm, Ym, Zm, 200, 'LineStyle','none'); colormap jet; colorbar  Bjorn Gustavsson on 22 Jun 2021
Your best bet on removing the "prickly edge" is to artificially interpolate data along the "red-valued" boundary (parabolic, hyperbolic?) to add one (or two) dense perimeters along that edge - if that's appropriate or what you want.
Triangulation-based methods will return some peculiarly shaped triangles in that region from the data you have, that is what it is, it feels counterintuitive and "unfair" but if you zoom in and look at the triangulation you will see what happens (you'll have to do an explicit triangulation with delaunay - likely on a much smaller sub-grid and manually plot the triangulation). To some degree you might change that behaviour somewhat by scaling your x and y-values differently but that's a crutch.

KSSV on 22 Jun 2021
Edited: KSSV on 22 Jun 2021
You have to increase the number of points while using linspace. Now defualt it is 100.
n = 500 ; % change this value
xv = linspace((min(x)), max(x),n);
yv = linspace(min(y), max(y),n);