How to use a for loop to get previous values to calculate next one

I am having trouble with my foor loop. I want to get the previous value to use in the next calculated value.

x(1) = A+B*(-K*cl^(k-1));
for k=0:4
    x(k) = A*x(k-1) + B*(-K*cl^(k-1));
end

Error I keep getting is:

Unable to perform assignment because the indices on the left side are not compatible with the size of the right side.

Answer 1

[Deepak]

Hi @Holly Janes,

I understand that you have written a MATLAB code to generate an array “x” that calculates and stores the values at next index from the previous one, but you are getting the error:

“Unable to perform assignment because the indices on the left side are not compatible with the size of the right side”.

The error occurs because, in MATLAB, array indices start at 1 instead of 0. Therefore, the loop should begin at (k = 2), since the value at index 1 is already assigned before the loop starts. Additionally, to optimize the code, you can initialize the array "x" with zeros before the loop by using the "zeros" function.

Here is the updated MATLAB code for the same:

% Preallocate x with the correct size 
x = zeros(1, 5); 

% Initial value 
x(1) = A + B * (-K * cl^(1-1)); 

% Loop to calculate subsequent values 
for k = 2:5 
    x(k) = A * x(k-1) + B * (-K * cl^(k-1)); 
end 
disp(x);

Please find attached the documentation of “array indexing” and “zeros” function for reference:

www.mathworks.com/help/matlab/math/array-indexing.html

www.mathworks.com/help/matlab/ref/zeros.html

I hope this was helpful.

[Walter Roberson]

If we assume that

x(1) = A+B*(-K*cl^(k-1));

worked, then A must have been scalar, and either the other elements are scalar or else they happen to be just the right size for matrix multiplication to reduce down to scalars.

But then when we examine

for k=0:4
    x(k) = A*x(k-1) + B*(-K*cl^(k-1));

the A will be retrieved first; it must be scalar for the previous statement to have worked. Then x(k-1) would be evaluated, and because k = 0 first, that would be x(-1) . x(-1) would fail with "Array indices must be positive integers or logical values." rather than an error about a mismatch in the number of elements.

So although you are correct to point out indexing at 0 is a problem with the code, it is not the most immediate error.

Answer 2

[Walter Roberson]

@Deepak correctly points out problems with indexing at 0 or -1. But that error would have priority . Because that error does not occur, we can tell that the real problem is in the statement

x(1) = A+B*(-K*cl^(k-1));

Possibly A is non-scalar. Or possibly B*(-K*cl^(k-1)) is non-scalar.