CODEWITHSUNDEEP
matlab
Umer Farooq
Umer Farooq

Reputation: 7486

Array addition error

I wrote the following code for convolution in Matlab, but it gives me errors at notified lines. I am a beginner in Matlab so please bear with me answering this question.

function [y] = UmerConv(x,h)

xlen=length(x);
hlen=length(h);

p=1;
for j=1:xlen
    for k=1:hlen
        uinput{p}=x(j)*h(k);
        p=p+1;
    end
end

for i=1:hlen
    if(i==1 || i==hlen)
        y{i}=uinput(i);   // error
    else
        y{i}=uinput(i)+uinput(i+2); // error
    end
end

Thanks

Upvotes: 1

Views: 58

Answers (2)

ely
ely

Reputation: 77484

You assign values into uinput with cell array syntax {}, but then later you index them with regular array syntax with uinput(i). You have to keep it consistent. Using the curly braces {} makes the array a cell array, which is indexed differently than a regular array (which uses just parentheses).

You then also make the same assignment choice with the variable y, using the cell array syntax when it seems you'll probably want just regular array syntax.

The corrected code should probably be:

function [y] = UmerConv(x,h)

xlen=length(x);
hlen=length(h);

p=1;
for j=1:xlen
    for k=1:hlen
        uinput(p) = x(j)*h(k); % <-- Changed the {} syntax
        p=p+1;
    end
end

for i=1:hlen
    if(i==1 || i==hlen)
        y(i) = uinput(i);  % <-- Now you access both with () instead of {}
    else
        y(i) = uinput(i) + uinput(i+2); % <-- Same here.
    end
end

At each place where I noted a correction in a comment, you could alternately stick with solely the {} syntax, and then everything will work but the arrays will be cell arrays. Usually this is not desired for a such a numerical computation.

Upvotes: 2

Manlio
Manlio

Reputation: 10864

You should use y(i) instead that y{i} to access the i-th position of a vector.

Upvotes: 0

Related Questions