how to add method to extend built-in class

Question

Erik Johnson 2013-5-7

2
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/74976-how-to-add-method-to-extend-built-in-class

评论： Jan Siegmund 2020-4-23

Prior to R2008a, it was possible to extend a class and a new method by just creating, somewhere in the path, the file @classname/methodname.m. Starting with R2008a, this still works ONLY if the class is defined using the older formulation; for classes defined using the newer classdef introduced in R2008a, how can I add a new method for a standard class (e.g., ss or sym) (without the complications introduced by subclassing)?

For example, if x and y are symbolic values, the minimum of the two can be written x*heaviside(y-x)+y*heaviside(x-y). I had a pre-R2008a function @sym/min.m that (in its simplest implementation) had:

function out = min(x,y)
out = heaviside(y-x).*x + heaviside(x-y).*y;
% a more complicated version handled matrix inputs,
% the dimension over which to minimize, and so forth

My @sym method directory could live anywhere on my path. Once the sym class was implemented with classdef, this no longer worked (consistent with the documentation at http://www.mathworks.com/help/matlab/matlab_oop/organizing-classes-in-folders.html). (Though, strangely, MATLAB's help command still sees methods in my @sym folder.)

The only solution I can find is to define, earlier on the path, a file min.m that uses an if isa(...) formulation to use my min code if one of the arguments is symbolic, and the built-in min otherwise -- something like:

function [varargout] = min(varargin)
varargout = cell(1,max(1,nargout));
if nargin==2 && (isa(varargin{1},'sym') || isa(varargin{2},'sym'))
    x = varargin{1};
    y = varargin{2};
    varargout{1} = heaviside(y-x).*x + heaviside(x-y).*y;
else
    [varargout{:}] = builtin('min',varargin{:});
end;

This seems like an overly-complicated solution, and it gives constant messages: Warning: Function /Users/username/matlab/min.m has the same name as a MATLAB builtin. We suggest you rename the function to avoid a potential name conflict.

An alternate idea was to move my @sym/min.m file into MATLAB's toolbox directory (which would be matlabroot/toolbox/symbolic/symbolic/@sym/min.m in this case); however, MATLAB seems to ignore this file, and this seems like a bad idea in any case.

Subclassing and creating my own symbolic class is possible, but more complicated than necessary for a simple extension like this, and would require recasting every symbolic variable to my subclass before calling the subclassed min method.

What is the alternate solution?

2 个评论
显示无隐藏无

Rudy 2014-7-23

Dear Erik,

I am facing the same problem right now. I have to update the code of a tool box I just inherited. Did you find a way to solve this problem?

Jan Siegmund 2020-4-23

Strangely, matlab still recommends the "@sym" folder:

https://de.mathworks.com/matlabcentral/answers/103129-how-do-i-overload-built-in-matlab-functions

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Rudy 2014-7-29

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/74976-how-to-add-method-to-extend-built-in-class#answer_146786

在 MATLAB Online 中打开

Since I had no luck trying to find a solution, I had to find one on my own. This is the method I came up with.

The toolbox had three new method for the zpk class. I created a new class, called sdzpk, and declared it to be a subclass of the built in zpk class. Then, wherever any of the new methods were used, I first converted the object to the new class before passing it to the method.

The following code may ilustrate that better:

Class definition file:

    classdef sdzpk < zpk & sdlti
        methods (Access = public)
            function obj = sdzpk(varargin)
                % Constructor method. Long code here to perform data validation
                % and pass information to the zpk constructor
                obj = obj@zpk(args{:});
            end
            % Existing methods
            % This are the old files I inherited. No need to edit them.
           tsys = ctranspose(sys);
           sys = delay2z(sys);
           sysr = minreal(sys,tol);
           F = minreals(F,tol);
           FP = setpoles(F,p);
           F = symmetr(F,type,tol);
           F = z2zeta(F,tol);
        end       
    end