Question

After converting a sphere into a patch, I plotted the first vertex of the patch, and realized that the vertex is not on the sphere.

Code

[faces, vertices] = surf2patch(sphere(),'triangle');
sphere();
hold on;
plot3(vertices(1,1), vertices(1,2), vertices(1,3),'*r');

The resulting figure looks like this

How can the vertex not be on the surface of the sphere? Am I misunderstanding something?

Thanks

Answer 1

Explanation

When a function is an argument to another function, Matlab treats this as a function call with a single return value. In the context of our function call above,

[faces, vertices] = surf2patch(sphere(),'triangle');

is equivalent to:

z = sphere();
[faces, vertices] = surf2patch(z,'triangle');

Now recall from the documentation:

[x, y, z] = sphere() calculates a set of 3D coordinates, [x, y, z].

z = sphere() calculates the same [x, y, z] and only returns the z.

sphere() with (nargout = 0) calculates the same [x, y, z], and plots it. This is the color sphere shown in the diagram above.

Examining the source for surf2patch.m shows that when called with a single z argument, it assumes values for x and y automatically. These values are different than the [x, y, z] plotted by calling sphere() with no arguments.

Testing

To see for yourself, trace through the following code with a breakpoint enabled just inside surf2patch.m, and examine the input arguments:

[x,y,z] = sphere()
[facesA, verticesA] = surf2patch(x,y,z,'triangle')
[facesB, verticesB] = surf2patch(sphere(), 'triangle')
[facesC, verticesC] = surf2patch(z, 'triangle')

You see that the output [facesA, verticesA] matches the plot from sphere(), whereas the second and third call to surf2patch() have no input [x, y] values, so they get auto-generated to something different than the original [x,y,z].

Lesson

If you expect multiple return values, assign them explicitly.