How to compare between two surface plots in matlab?

A and B are two 2D matrices with dimensions (length(X), length(Y)) which are function of X and Y vectors. I plot A and B in a 3D graph using surf as follows:

figure(1)
surf(X,Y,A)
hold on
surf(X,Y,B)
hold on

But, the generated figure doesn't show the difference in values between A and B clearly.

Is there any other way where I can compare between A and B in a 3D plot and show the differences between them clearly?

Answer 1

[Star Strider]

Perhaps:

figure(1)
surf(X,Y,A, 'FaceAlpha',0.5)
hold on
surf(X,Y,B, 'FaceAlpha',0.5)
hold off

.

[Adi Nor]

Thankyou. This helps me to make my idea little clear. Is there any other way to make it more clear?

[Star Strider]

My pleasure!

It would be worth experimenting with the 'FaceAlpha' property for both surf plots in order to determine what combination works best for you. Without your data (and a clear idea of what you want), I can only guess.

[Adi Nor]

I need to show the difference between the two matrices more clearer because the values of matrix A is obtained from my optimization, wheras the values of B is obtained from the literature and I want to show the siginficant improvenment in the values of A compared to B through plotting the both matrices. I can give you the data:

X = -20:10:80; %length(X)=11
Y = 2:2:8; %length(Y)=4
%my optimized matrix A (4*11)
A = [0.0046    0.0442    0.3160    1.3217    3.5302    5.8383    6.9030    7.1149    7.1335    7.1327    7.0763;
    0.0047    0.0450    0.3329    1.4479    4.1525    6.5922    7.6623    7.8441    7.8713    7.8725    7.8008;
    0.0048    0.0466    0.3562    1.5350    4.5591    7.2824    8.5957    8.8450    8.8786    8.8067    8.8070;
    0.0049    0.0475    0.3795    1.5952    4.7382    7.9092    9.5009    9.8884    9.9292    9.8430    9.8434]
%matrix B from previous work (4*11)
B = [0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.5441    6.4802;
    0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.5441    6.4802;
    0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.4801    6.4802;
    0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.4801    6.4802]

[Star Strider]

Try this:

[X,Y] = ndgrid(-2:0.5:2);
Z1 = X.^2 + Y.^2;
Z2 = X.^2 + Y.^2 + 5;

figure
surf(X, Y, Z1, 'FaceAlpha',0.5)
hold on
surf(X, Y, Z2, 'FaceAlpha',0.5)
plot3((X(:)*[1 1]).', (Y(:)*[1 1]).', [Z1(:) Z2(:)], '-r', 'LineWidth',2)   % Draw Vertical Red Lines Between The Surfaces
hold off
grid on

It plots both surfaces, then draws red lines connecting the points of both of them. (They must be defined over the same (X,Y) grids.) Because off the 'FaceAlpha' properties, the vertical lines are visible for the entire plot.

[Adi Nor]

Is this code applicable for my data listed in the previous comment? what's Z1 and Z2? and why ndgrid(-2:0.5:2)?

I think it needs some modifications?

[Star Strider]

I did not see your previous Comment until now. (I was likely writing mine at the same time.)

My code appears to work with your data (I renamed your ‘X’ and ‘Y’ as ‘Xv’ and ‘Yv’ respectively):

[X,Y] = meshgrid(Xv, Yv);

figure
surf(X, Y, A, 'FaceAlpha',0.5)
hold on
surf(X, Y, B, 'FaceAlpha',0.5)
plot3((X(:)*[1 1]).', (Y(:)*[1 1]).', [A(:) B(:)], '-r', 'LineWidth',2)     % Draw Vertical Red Lines Between The Surfaces
hold off
grid on

producing this plot:

.

[Adi Nor]

Thank you! it works well now.

[Star Strider]

As always, my pleasure!

Answer 2

[Ameer Hamza]

Sometime plotting the difference can be helpful for visualization

figure(1)
surf(X,Y,A-B)

[Adi Nor]

I understand your idea, but I don't need to plot the difference. I need to show it by plotting the two matrices. This is because the values of matrix A is obtained from my optimization, wheras the values of B is obtained from the literature and I want to show the siginficant improvenment in the values of A compared to B through plotting the both matrices.

[Ameer Hamza]

Ok. In that case, I think plotting the estimated values as surface, and the original data points as scattered points will be helpful. For example

figure(1)
surf(X,Y,A)
hold on
scatter3(X(:),Y(:),B(:), 'r+')
hold on

[Adi Nor]

I have an error using your proposed code

Error using scatter3 (line 64)
X, Y and Z must be vectors of the same length.

It should be noted that

X=-20:10:8; %length(X)=11

Y=2:2:8; %length(Y)=4

A and B are a 4*11 2D matrix

[Ameer Hamza]

Ok, If X and Y are vectors then modify the code to this

figure(1)
surf(X,Y,A)
hold on
[Xg, Yg] = meshgrid(X, Y);
scatter3(Xg(:),Yg(:),B(:), 'r+')
hold on

[Adi Nor]

The data is given as follows:

X = -20:10:80; %length(X)=11
Y = 2:2:8; %length(Y)=4
%my optimized matrix A (4*11)
A = [0.0046    0.0442    0.3160    1.3217    3.5302    5.8383    6.9030    7.1149    7.1335    7.1327    7.0763;
    0.0047    0.0450    0.3329    1.4479    4.1525    6.5922    7.6623    7.8441    7.8713    7.8725    7.8008;
    0.0048    0.0466    0.3562    1.5350    4.5591    7.2824    8.5957    8.8450    8.8786    8.8067    8.8070;
    0.0049    0.0475    0.3795    1.5952    4.7382    7.9092    9.5009    9.8884    9.9292    9.8430    9.8434]
%matrix B from previous work (4*11)
B = [0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.5441    6.4802;
    0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.5441    6.4802;
    0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.4801    6.4802;
    0.0047    0.0445    0.3077    0.8624    2.4259    4.9777    6.2328    6.5002    6.5413    6.4801    6.4802]

[Ameer Hamza]

Try the code in my last comment.

[Adi Nor]

Thank you! .. your solution is acceptable for me.