I have built a function which depends on two variables; I would like to get a 3dplot to see the behaviour of my function for different values of the two inputs

I have built a function which depends on two variables; I would like to get a 3dplot to see the behaviour of my function for different values of the two inputs. In particular, it seems that the mistake is in this line, when the multiplication by "xu" is performed:

xu = [0.00001:(T/n):(T-0.00001)];
[lambda0,lambda1] = meshgrid(0:.1:3, 0:.1:3);
z = 2*(lambda0.*lambda1.*xu.*(T - xu)).^(0.5);

any solution for the problem?

[Stephen23]

What are the "two inputs" ? T, n, xu, lambda0, or lambda1 ? You omitted to actually tell us which values you want to plot over.

[Amedeo De Vivo]

Hi Stephen! lambda0 and lambda1 are the variables of interest.

Answer 1

[Walter Roberson]

Your lambda0 and lambda1 are 31 x 31 matrices. Your xu is a row vector of length n, which might or might not be 31. What result are you expecting when you multiply a 31 x 31 matrix by a row vector of length n ?

[Amedeo De Vivo]

Hi Walter!

In my original code lambda0 and lambda1 were scalar values and everything worked fine with the multiplication by "xu": as I am switching to a 3d framework, I thought the ".*" multiplication was sufficient. Are you suggesting me to create lambda0 and lambda1 matrices of the same length of "xu" and multiply by the transpose of "xu"??

[Walter Roberson]

.* requires that the two operands be the same size (unless one is scalar) and gives a result that is that size.

Meshgrid all three variables at the same time to get 3d arrays to use.

[Stephen23]

Something like this:

[lambda0,lambda1,XU] = ndgrid(0:.1:3, 0:.1:3, xu);
Z = 2*(lambda0.*lambda1.*XU.*(T - XU)).^(0.5);

[Walter Roberson]

Right. Notice that really you have a function of 3 variables here, the two lambda and xu, so you will need a 4 dimensional display routine, not a 3d display routine.