运算符重载元素数组

Question

I have a built a class and would like to add Operator Overloading for minus . I wrote the code and it works just fine if I input only one element, however the function I wrote is supposed to work for multiple elements.

When I try to input an array, (array of elements)-(one element) , it should "deduct" def from each element of arr and return a new array of elements.

This is the code:

  function this=minus(obj_arr,filt)
        [lena, lenb]=size(obj_arr);
        if (lena==1)&&(lenb>1)
            obj_arr=obj_arr';
            len=lena;
        elseif (lena>1)&&(lenb==1)
            len=lenb;
        elseif (lena==1)&&(lenb==1)
            len=lena;
        else
            error('at least one dimention must be 1')
        end

        [lenfa,lenfb]=size(filt);
        if (lenfa~=1)||(lenfb~=1)
            error('filter must be 1x1')
        end


        difference=mean(filt.filtered_pf_arr);
        [~,max_val]=max(difference);
        for i=1:len-1

            temp_filt_pf=obj_arr(i).filtered_pf_arr;
            [num_calls,num_fr]=size(temp_filt_pf);
            for ii=1:num_calls
                new_temp_filf_pf=zeros(num_calls,num_fr);
                temp_mean_filt_pf=temp_filt_pf(ii,:);

                max_val_temp=temp_mean_filt_pf(max_val);
                few_def=difference.*max_val_temp;
                new_temp_filf_pf(ii,:)=temp_filt_pf(ii,:)-few_def;
            end
            obj_arr(i).filtered_pf_arr=new_temp_filf_pf;

        end
        this=obj_arr;

It works when I input one command window

arr- an array of elements of the class
elementx- an element of class

elementA-elementB
classname.minus(arr,elementx)
arr(1)-elementx

It doesn't work when I input

arr-elementx

error: No method 'minus' with matching signature found for class 'one_session'.

Answer 1

After this bit:

[lena, lenb]=size(obj_arr);
if (lena==1)&&(lenb>1)
    obj_arr=obj_arr';
    len=lena;
elseif (lena>1)&&(lenb==1)
    len=lenb;
elseif (lena==1)&&(lenb==1)
    len=lena;
else
    error('at least one dimention must be 1')
end

len is always 1. Next, you have a loop:

for i=1:len-1

that will never run. You are taking i from the array 1:(1-1) == 1:0 = [] . So the loop is always skipped.

You probably meant for len to have the length of the input vector. You can instead use isvector and length :

if ~isvector(obj_arr)
    error('at least one dimension must be 1')
end
len = length(obj_arr);

Note that length(obj_arr) is the same as max(size(obj_arr)) . isvector can be implemented as nnz(size(obj_arr)>1)==1 (ie number of dimensions larger than 1 is 1) or, assuming a 2D matrix as you did in your code, any(size(obj_arr)==1) .

Next, make sure the loop runs through all elements of the vector:

for i=1:len

---

For efficiency reasons, you should define your function as

function obj_arr = minus(obj_arr,filt)

This will allow MATLAB to better optimize certain calls to your function, avoiding unnecessary copies when working in-place (ie arr=arr-1 ).