Calling a plotting function inside a loop.

Hello all. I'm trying to make a series of figures using a for loop. I need to do a good deal of processing on the points so I'm using my own function to generate the points for each given set of coefficients then convert from spherical to Cartesian coordinates and plot. The relevant code is as follows:

%The matrices dataexist and resultsmatrix are calculated earlier
c = 1;
for d = 1:26
   if dataexist(d,c) == 1
      figure(d)
      coeffs = resultsmatrix(d,:,c);
      plotsurface(coeffs);
   end
end

function [] = plotsurface(coeffs)
%Cartesian points are calculated based on the coefficients
tri = delaunay(cart(:,1), cart(:,2));
[r,c] = size(tri);
trisurf(tri, cart(:,1), cart(:,2), cart(:,3));

The problem is, this gives me the first surface just fine on Figure 1, but every figure after that is blank. So somehow the function isn't plotting to it's proper figure. I'm wondering if maybe it's all overwriting onto Figure 1 despite my making a new figure before each call to plotsurface but I don't know why it would do that.

Answer 1

[Walter Roberson]

Create an axes within the figure, pass the axes handle down to plotsurface(), and parent your trisurf() against that particular axes by using the 'Parent' property:

thisaxes = axes('Parent',d);
coeffs = resultsmatrix(d,:,c);
plotsurface(coeffs, thisaxes);

function [] = plotsurface(coeffs, thisaxes)
[...]
trisurf(tri, cart(:,1), cart(:,2), cart(:,3), 'Parent', thisaxes);

[B K]

This didn't seem to work.

I have:

for d = 1:26

if dataexist(d,c) == 1

figure(d)

thisaxes = axes('Parent',d);

coeffs = resultsmatrix(d,:,c);

plotsurface(coeffs, thisaxes);

end

end

Then passed in thisaxes into plotsurface as you said. The same problem happens.

[Walter Roberson]

I would suggest you put a breakpoint in plotsurface, and check that the data is non-empty for the additional figures, and check that the axes are distinct. As well, as you single step past the trisurf function you would be able to see whether the plot in the first figure changes.

[B K]

Alright, I tried breakpoints which showed me that it still stubbornly overwrites in figure 1 no matter what I do. I even threw together a version without the loop just to see if I could get it to plot on two different figures by going

figure(1)

trisurf(...)

figure(2)

trisurf(...)

That always works for plot and surf and based on the documentation I don't see why trisurf should behave differently.

[Walter Roberson]

I have verified that with 2008b, trisurf() will first obey any 'Parent' parameter passed in, provided the parent is given as an axis; if no parent is passed in, then it will use gcf to find the current figure, and gca() that to find the current axes on the current figure, and will use that. If for some reason no figure is current, then figure() should be invoked, which will generally return the lowest positive-integer-numbered unused figure.