Property assignment is not allowed when the object is empty. Use subscripted assignment to create an array element.

Hello!

I was developed a software with app designer and connect all values in a constructor class (this connection is running). But when I click in Run button, I need to access the function Positive_Time(load, tnt, analysis) in Class_Load and presents this error. Can you help me?

Property assignment is not allowed when the object is empty.  Use subscripted
assignment to create an array element.
Error in Class_Load/Positive_Time (line 222)
            load.tnt = tnt;

Here my Class_TNT

classdef Class_TNT
    
    properties (SetAccess = public, GetAccess = public)
        Z            = 0;
        W            = 0;
        type         = 0;
        negative     = 0;
    end

methods
    
    function tnt = Class_TNT(W, Z, type, negative)
        if (nargin > 0)
            tnt.Z            = Z;
            tnt.W            = W;
            tnt.type         = type;
            tnt.negative     = negative;
        else
            tnt.Z            = 5;
            tnt.W            = 10;
            tnt.type         = 1;
            tnt.negative     = 1;
            
        end
    end
end
end

Here, my Class_Load

 classdef Class_Load
            
    properties (SetAccess = public, GetAccess = public)
        
        tnt      Class_TNT              % TNT's properties
        analysis Class_TypeAnalysis     % Type Analysis propertie
        td      = 0;    % Variable about the time that is during the 
        
    end
    
    methods
        
        function load = Class_Load(tnt, analysis)
            if (nargin > 0)
                load.Positive_Time(load, tnt, analysis);
            else
                load.tnt = Class_TNT();   % Empty constructor
            end
            
        end
        
    end

methods
    function load = Positive_Time(load, tnt, analysis)
            
            load.tnt = tnt;
            load.analysis = analysis;
            
            type          = load.tnt.type;
            Z             = load.tnt.Z;
            W             = load.tnt.W;
                        
            switch type
                
                case 1
                    
                    if (0.20993 < Z && Z <= 1.0355)
                        load.td = 0.001 * ( 2.2064 * Z^3 - 0.3363 * Z^2 - ...
                            0.5644 * Z + 0.3756 ) * ( W^(1/3) );
                    elseif (4.001 < Z)
                        load.td = 0.001 * ( 1.5602 * log(Z) +1.2416 ) * ...
                            ( W^(1/3) );
                    end
                    
                case 2
                    
                    if (0.2210 < Z && Z <= 0.7790)
                        load.td = 0.001 * ( 4.2668 * Z^3 - 1.9736 * Z^2 + ...
                            0.132 * Z + 0.2145 ) * ( W^(1/3) );
                    elseif (3.7612 < Z)
                        load.td = 0.001 * ( - 0.0029 * Z^2 + 0.2159 * Z + ...
                            2.1382 ) * ( W^(1/3) );
                    end
                    
            end
    end
end
end

[per isakson]

load.Positive_Time(load, tnt, analysis);

should be

load.Positive_Time( tnt, analysis );

That's the only problem I found.

[Ana Reis]

I did. But the problemn not desapear

[Matt J]

I gather from your post that you don't know what the load object contains at the moment that the error occurs? You should know - it's the kind of information that routine use of the debugger would give you:

https://www.mathworks.com/help/matlab/ref/dbstop.html#buzezti-1

[Ana Reis]

thank you for the answer, but I always use breakpoint...

[Image Analyst]

Of course, everyone uses breakpoints.

load() is a built in function. It's a very, very bad idea to call your returned structure "load" or any other built-in function name. Pick another name than load.

Attach your class file, and your test file that calls it if you want more help.

[Walter Roberson]

        function load = Class_Load(tnt, analysis)
            if (nargin > 0)
                load.Positive_Time(load, tnt, analysis);

You have no class property named load so the load in the argument does not refer to any class property or any input variable. You have an output variable named load but you have not assigned to it, so it does not exist yet. Therefore the load refers to the load() function, which would default to loading matlab.mat if it can find it, returning the loaded variables in the form of a struct. But then load.Positive_Time does not exist to pass the result into.

Answer 1

[per isakson]

I fail to reproduce the error you report. However, that should not surprise since your description of how you execute the classes is vague.

My first test

tnt = Class_TNT( 17, 7, 1, 1 );
ana = Class_TypeAnalysis;
ladda = Class_Ladda( tnt, ana )

outputs

ladda = 
  Class_Ladda with properties:
         tnt: [0×0 Class_TNT]
    analysis: [0×0 Class_TypeAnalysis]
          td: 0
>> 

No error, an instance of Class_Ladda is created, but the values of the properties tnt and analysis are empty.

My second test

I replaced the statements

this.Positive_Time( tnt, analysis );

in the constructor of Class_Ladda by

this = this.Positive_Time( tnt, analysis );

since it's a value class. Now the output looks ok

ladda = 
  Class_Ladda with properties:

         tnt: [1×1 Class_TNT]
    analysis: [1×1 Class_TypeAnalysis]
          td: 0.0110
>> 

The classes are

classdef Class_TNT
    properties (SetAccess = public, GetAccess = public)
        Z            = 0;
        W            = 0;
        type         = 0;
        negative     = 0;
    end
    methods
        function tnt = Class_TNT( W, Z, type, negative )
            if (nargin > 0)
                tnt.Z            = Z;
                tnt.W            = W;
                tnt.type         = type;
                tnt.negative     = negative;
            else
                tnt.Z            = 5;
                tnt.W            = 10;
                tnt.type         = 1;
                tnt.negative     = 1;
            end
        end
    end
end

classdef Class_Ladda
    properties ( SetAccess = public, GetAccess = public )
        tnt      Class_TNT              % TNT's properties
        analysis Class_TypeAnalysis     % Type Analysis propertie
        td      = 0;    % Variable about the time that is during the
    end
    methods
        function this = Class_Ladda( tnt, analysis )
            if (nargin > 0)
                this.Positive_Time( tnt, analysis );
            else
                this.tnt = Class_TNT();   % Empty constructor
            end
        end
    end
    methods
        function this = Positive_Time( this, tnt, analysis )
            this.tnt = tnt;
            this.analysis = analysis;
            type          = this.tnt.type;
            Z             = this.tnt.Z;
            W             = this.tnt.W;
            
            switch type
                case 1
                    if (0.20993 < Z && Z <= 1.0355)
                        this.td = 0.001 * ( 2.2064 * Z^3 - 0.3363 * Z^2 - ...
                            0.5644 * Z + 0.3756 ) * ( W^(1/3) );
                    elseif (4.001 < Z)
                        this.td = 0.001 * ( 1.5602 * log(Z) +1.2416 ) * ...
                            ( W^(1/3) );
                    end
                case 2
                    if (0.2210 < Z && Z <= 0.7790)
                        this.td = 0.001 * ( 4.2668 * Z^3 - 1.9736 * Z^2 + ...
                            0.132 * Z + 0.2145 ) * ( W^(1/3) );
                    elseif (3.7612 < Z)
                        this.td = 0.001 * ( - 0.0029 * Z^2 + 0.2159 * Z + ...
                            2.1382 ) * ( W^(1/3) );
                    end
                otherwise
                    error( 'Positive_Time:Unknown', 'Invalid type: %d', type )
            end
        end
    end
end

I've replaced "load" by "ladda"

[Ana Reis]

Thank you soo much!!!!! You saved me! Ans I'm sorry to put vague classes. I need to learn how to write the correct question in Answers sites like these.

[per isakson]

Thank you for confiming that my answer worked for you!

I guess you have seen TUTORIAL: how to ask a question (on Answers) and get a fast answer. IMO:

• TL;NR
• hard to fullfill all the recommendation
• however, it's worth reading

There are many types of questions. Regarding "I need help, my code throws an error":

• To spot errors by code inspections is difficult (except for trivial syntax errors) and tedious
• Provide code and an instruction, which makes it easy to reproduce the error. Code, in excess of a screenful, should be attached with help of the paper-clip icon. Don't forget appropriate values of input arguments.
• Template: 1) A very short background. 2) On my R2018b/Win10 this little script produces the following error. 3) The files, File1.m and File2.m are attached. 4) Some comments which shows your ideas on the error and that you made an effort.
• Always include the full error message (as you did).
• A Minimal working example is always good, but is often easier said than done. Making a MWE is often a way to solve the problem and get deeper insight.
• Never ever just reply "Doesn't work:(" to an answer. (You didn't.)

Finally, I opened your Class_Load in the editor and clicked the green triangle and got

>> Class_Load
ans = 
  Class_Load with properties:
         tnt: [1×1 Class_TNT]
    analysis: [0×0 Class_TypeAnalysis]
          td: 0
>> 

in the command window.

[Ana Reis]

yes yes! I'm going to read this. Thank you for the help and for this TUTORIAL to ask questions.