How to use property from one class in to define a property in another class?

Hi All,

I have defined one class

classdef AgentVars
    properties
        set_o = linspace(-0.2,0.2,20) % set of observations
    end   
end

I want to use the the set_o property in another class. For example

classdef Agent
    properties
        set_o = AgentVars.set_o
    end
end

However, currently I am not able to access a property from the other class. How can I do that? Thank you

Answer 1

[Rik]

I don't know what the 'correct' answer would be, but you can also set the value in the constructor:

classdef Agent
    properties
        set_o = NaN
    end
    methods
        function obj=Agent
            obj.set_o=AgentVars.set_o
        end
    end
end

You could also inherit from a superclass, but that only works if you want to inherit from a single class.

[Prashanti Ganesh]

Thanks. But when I run it, the value of set_o is still NaN. It should be the value that was assigned in the AgentVars class. How can I get that?

[Rik]

Please post the exact code you're using and attach the classdef m-files.

[Prashanti Ganesh]

Here are the two classes I am using. I want to access the set_o from the AgentVars class to use it for the Agent class. When I try to run it, this is the error I get.

The property 'set_o' in class 'AgentVars' must be accessed from a class instance because it is not a Constant property.

[Rik]

Looking at your code structure it makes most sense to inherit the parameters, making AgentVars a superclass. Is there any reason you don't?

Alternatively, you could do as the error message suggests: make the properties in AgentVars Constant.

After a few more edits (putting calculated properties in the constructor, it runs without error:

a=Agent

a = 
  Agent with properties:

                  set_o: [-0.2000 -0.1789 -0.1579 -0.1368 -0.1158 -0.0947 -0.0737 -0.0526 -0.0316 -0.0105 0.0105 0.0316 0.0526 0.0737 0.0947 0.1158 0.1368 0.1579 0.1789 0.2000]
                  kappa: 0.0800
                  sigma: 0.0400
                    c_t: 1
                   beta: 100
                  agent: 1
                 lambda: NaN
               eval_ana: 1
                  alpha: 0.1000
    task_agent_analysis: NaN
                  q_0_0: 0.5000
                    o_t: NaN
                    d_t: NaN
                    a_t: NaN
                    r_t: NaN
                   pi_0: NaN
                   pi_1: NaN
                 E_mu_t: 0.5000
                  v_a_0: NaN
                  v_a_1: NaN
                  v_a_t: NaN
                  p_a_t: NaN
                      t: NaN
              p_o_giv_u: [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
         p_o_giv_u_norm: [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0]
                      C: NaN
                    q_o: NaN
                    q_1: NaN
                      d: NaN
                      G: NaN
                  p_d_t: [NaN NaN]
                    Q_t: [2×2 double]
                product: [1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1]
                     mu: [100×100 double]
                   p_mu: [1×100 double]
              mu_for_ev: [100×1 double]

[Prashanti Ganesh]

Thanks for this. I have not made AgentVars into a superclass because I was not aware of this method. I will try it. Could you please explain how to make the properties a constant.

[Prashanti Ganesh]

I figured how to make it a constant. Could please help me understand what are calculated properties?

[Rik]

As far as I know that isn't an official term. Take a look at the file I attached: the values of some properties are calculated based on the values of other properties. Because those aren't yet initialized, they can't be accessed in the properties block. That is why they are calculated in the constructor.

[Prashanti Ganesh]

Thanks so much. It makes sense now.