"Array indices must be positive integers or logical values".Error in (line 17) tij_lm(l, m) = min([i(l + 1) - i(l), i(l) - i(l - 1), j(m + 1) - j(m), j(m) - j(m - 1)]) / 2;

Getting "Array indices must be positive integers or logical values". Error

Error in cross_correlation (line 17)

tij_lm(l, m) = min([i(l + 1) - i(l), i(l) - i(l - 1), j(m + 1) - j(m), j(m) - j(m - 1)]) / 2;

Heres my code, why do I get the error?

% Example data for time series i and j

i = [1, 2, 3, 4, 5];

j = [1.5, 3, 4.5];

% Call the cross_correlation function with the example data

[Cxy, Jij, tij_lm] = cross_correlation(i, j);

function [Cxy, Jij, tij_lm] = cross_correlation(i, j)

% Inputs:

% i - time series data for variable i

% j - time series data for variable j

% Compute the length of time series i and j

Si = length(i);

Sj = length(j);

% Initialize Jij and tij_lm matrices

Jij = zeros(Si, Sj);

tij_lm = zeros(Si, Sj);

% Compute Jij and tij_lm for each l and m

for l = 1:Si

for m = 1:Sj

tij_lm(l, m) = min([i(l + 1) - i(l), i(l) - i(l - 1), j(m + 1) - j(m), j(m) - j(m - 1)]) / 2;

if 0 < i(l) - j(m) && i(l) - j(m) < tij_lm(l, m)

Jij(l, m) = 1;

elseif i(l) == j(m)

Jij(l, m) = 0.5;

else

Jij(l, m) = 0;

end

end

end

% Compute the cross correlation value C(x/y)

Cxy = sum(sum(Jij));

end

[Dyuman Joshi]

Indexing in MATLAB starts with 1

When l and m are 1, l-1 and m-1 are 0, which throws the error you are receiving.

""Array indices must be positive integers ...

for l = 1:Si
    for m = 1:Sj
        tij_lm(l, m) = min([i(l + 1) - i(l), i(l) - i(l - 1), j(m + 1) - j(m), j(m) - j(m - 1)]) / 2;

Since I don't know what you want to do with this line of code, I can't suggest any changes.

[Arif Hoq]

i = [1, 2, 3, 4, 5];
j = [1.5, 3, 4.5];
l=1:5;

i(l) - i(l - 1)

when l=1

i(1)=1

i(1-1)= i(0) % which shows an error.

also same here

j(m) - j(m - 1)

index should be an integer.

Can you please explain what you want here? please give an example.

tij_lm(l, m) = min([i(l + 1) - i(l), i(l) - i(l - 1), j(m + 1) - j(m), j(m) - j(m - 1)]) / 2;

Answer 1

[Arif Hoq]

if you start your index from 2 then, this can be one approach:

% Example data for time series i and j
i = [1, 2, 3, 4, 5];
j = [1.5, 3, 4.5];
 
% Call the cross_correlation function with the example data
[Cxy, Jij, tij_lm] = cross_correlation(i, j)

Cxy = 0.5000

Jij = 5×3

0 0 0 0 0 0 0 0.5000 0 0 0 0 0 0 0

tij_lm = 5×3

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

 
function [Cxy, Jij, tij_lm] = cross_correlation(x, y)
  % Inputs:
  %   i - time series data for variable i
  %   j - time series data for variable j
  
  % Compute the length of time series i and j
  xlength = length(x);
  ylength = length(y);
  
  % Initialize Jij and tij_lm matrices
  Jij = zeros(xlength, ylength);
  tij_lm = zeros(xlength, ylength);
  
  % Compute Jij and tij_lm for each l and m
  for i = 2:xlength
    for j = 2:ylength
      tij_lm(i, j) = min([x(i) - x(i), x(i) - x(i - 1), y(j) - y(j), y(j) - y(j - 1)]) / 2;

      if 0 < x(i) - y(j) && x(i) - y(j) < tij_lm(i, j)
        Jij(i, j) = 1;
      elseif x(i) == y(j)
        Jij(i, j) = 0.5;
      else
        Jij(i, j) = 0;
      end
    end
  end
  
  % Compute the cross correlation value C(x/y)
  Cxy = sum(sum(Jij));
end