Trying to plot a radial contour

I am trying to plot a 2x2 contour subplot, it is supposed to be circular, but I either get lines or no plot at all or an error saying my z needs to be a 2x2 array.

function F = ratio_rp_bf(R,theta,bD)

for r = 1:length(R)

  for b = 1:length(bD)

    F(b,r) = (16/3).*(1./bD(b)).*(1./R(r)).^3.*abs(sqrt((1./R(r)).^4-2.*(1./R(r)).^2.*cos(2.*theta(r))+1));

  end

end

bD = [0.1, 1.0, 10, 100];
R = (1:1.25:5);

c = R;
a = 1/2*bD;
b = abs(sqrt(c.^2 - a.^2));

theta = abs(atan(b./a));

[F] = ratio_rp_bf(R,theta,bD);

              for b = 1:length(bD)

               figure(1)
               subplot(2,2,b)
               x = R.*cos(theta);
               y = R.*sin(theta);
               z = F;
               [x,y,z] = meshgrid(x,y,z);
               [C,h] = contour(R,theta,bD,x,y,z);
               clabel(C,h);
               grid on
  end

Answer 1

[Thorsten]

You try to visualize the contours of a 2 x 2 image. That should most likely look degraded. And note that for the values chosen each row is simply 1/10 of the previous row:

F(b+1, r) = F(b, r)/10

Is this correct?

Answer 2

[Walter Roberson]

Why are you trying to pass 6 parameters in to contour()? contour expects at most 4 numeric parameters (excluding name/value pairs)

[Kelly]

I was following my professor's example. He didn't really explain why and it's the first time we've used the contour command, so I'm still trying to get a grasp on it. Because our original data is in polar coordinates, I use x and y to convert to cartesian coordinates, and z is supposed to equal our computations for F. When I try to plot using x,y,z, I get an error saying z needs to be a 2x2 array, which I understand, but my data sets aren't 2x2 arrays.

[Walter Roberson]

When you use

[x,y,z] = meshgrid(x,y,z);

you make it harder to debug, as you are overwriting the variables you are meshgrid()'ing.

Before you do the meshgrid(), what are size(x), size(y), size(z) ?

Why are you looping over bD when nothing in your loop is affected by the looping other than which subplot you are plotting in?

[Kelly]

x, y, and z are 1x1. R is 1x1, theta is 1x5 and bD is 1x2. bD is somehow supposed to be the z axis, I don't know how though. R is supposed to be plotted along the x axis, and theta the y. z is equal to the computations.

[Walter Roberson]

How can bD possibly be 1 x 2 when just above you have

bD = [0.1, 1.0, 10, 100];

which is 1 x 4 ??

[Kelly]

I don't know. R, x, and y should be 1x4 as well, but when I use the size command MATLAB gives me a different answer. But when I double click the variables in the workspace window, it shows the appropriate dimensions.

[Walter Roberson]

Have you tried restarting MATLAB? Then put in a breakpoint at the meshgrid line, and examine the size() and print the values of the variables to see how many entries show up and what they are.

[Kelly]

I have, and the size() is the same, but the values print as 1x4 like they are supposed to do, and the computations are correct. I checked them by hand.

Answer 3

[Image Analyst]

I'm not really sure I follow what you're trying to do, nor do I know what sort of output ratio_rp_bf() creates, but if you just want to draw dircles, see the FAQ:

http://matlab.wikia.com/wiki/FAQ#How_do_I_create_a_circle.3F

If F is an image that you created that's supposed to be a circular cone so that contours on it would be concentric circles, then please upload that image since I don't have ratio_rp_bf() which creates it.

[Walter Roberson]

The ratio_rp_bf function is at the very top of the posting.

[Image Analyst]

Oh you're right. But then wouldn't that make this a recursive function, since it's calling itself? Maybe I'm just not following it correctly because of the bad formatting. Where is the actual end of the ratio_rp_bf() function?

[Kelly]

%this is the function the main program is calling to run all the computations F, for the length of bD, using R and theta which are dependent variables and which vary directly.

function F = ratio_rp_bf(R,theta,bD)

for r = 1:length®

for b = 1:length(bD)

F(b,r) = (16/3).*(1./bD(b)).*(1./R(r)).^3.*abs(sqrt((1./R(r)).^4-2.*(1./R(r)).^2.*cos(2.*theta(r))+1));

end

end

% This is the main program. After all the computations have been done, the output should be a 2x2 contour subplot in the form of a circle for each bD. The bD should cause a break to form in the plotted circle.

bD = [0.1, 1.0, 10, 100];

R = (1:1.25:5);

c = R;

a = 1/2*bD;

b = abs(sqrt(c.^2 - a.^2));

theta = abs(atan(b./a)); %finds the angle theta for each given radii (R)

[F] = ratio_rp_bf(R,theta,bD);% calls function to perform computaions

              for b = 1:length(bD)

               figure(1)
               subplot(2,2,b)
               x = R.*cos(theta); % conversion from polar coordinates to cartesian.
               y = R.*sin(theta);
               z = (x.^2+y.^2)<=R.^2;
               [C,h] = contour(x,y,z,F);
               clabel(C,h);
               grid on
  end

I get an error stating z needs to be size 2x2 or greater.

[Kelly]

According to my professor, contour(R,theta,bD) should plot the contours of the expression F and it should be radial with breaks at each bD.

[Walter Roberson]

One of you has misinterpreted. The three-argument form of contour() is always X, Y, Z coordinates. To use specific contour levels, you need the two-input or four-input forms of contour(). See http://www.mathworks.com/help/matlab/ref/contour.html

Note that you have not included F in your contour() call.