how to split one dimensional array into 3 equal chunks?

Hi, I have the attached data array as an example, I need to split it into 3 equal chunks! Then I need to apply that for any other array with the same type but with different length.

[Adam Danz]

There's one variable stored in your mat file so it would be a lot simpler to help you if you just describe that variable rather than require multiple people do download a mat file, load it in matlab, etc.

The variable Cb is a column vector of length 4897 which is not divisible by 3 so it is impossible to divide it into 3 chunks of equal length.

[sana3 sal]

So, suppose it is any other array that has divisible length!?

Answer 1

[Star Strider]

Try this:

s = load('Cb.mat');
Cb = s.Cb;

var3c = @(oldvar) mat2cell(oldvar(:), [fix(numel(oldvar)/3) *[1, 1], numel(oldvar)-2*fix(numel(oldvar)/3)], 1);     % Create New Matrix From Original Vector

newCb = var3c(Cb)

newCb =

    3×1 cell array

      {1632×1 double}
      {1632×1 double}
      {1633×1 double}

You cannot divide your vector equally, although you can get close. The ‘var3c’ anonymous function works with any length vector. It returns a cell array of 3 column vectors, even if the original vector is a row vector.

[sana3 sal]

It is okay, there is no problem in dividing the vector into {1632×1 double} {1632×1 double} {1633×1 double} It worked with me thanks. But now , how i can access each cell alone? for example,i need to find the standard deviation for each part?

[Star Strider]

My pleasure.

Use the cellfun (link) function:

newCbstd = cellfun(@std, newCb)

newCbstd =
   0.521806807562178
   0.061049922231889
   0.096887917396344

[sana3 sal]

I used this, is it fine?

S1=std(newCb{1,:})
S2=std(newCb{2,:})
S3=std(newCb{3,:})

[sana3 sal]

Any way, thank you alot you saved my life ♥

[Star Strider]

As always, my pleasure!

If it gives you the same result as I got, that would work.

Answer 2

[Adam Danz]

This will take a column vector whose length is divisible by 3 and reshape it into a matrix with 3 columns of equal length. Each column of the new matrix will be one of your chunks.

 Cb = (1:300)'; 
 n = 3;
 Cbmat = reshape(Cb, [], n);

Cbmat(:,1) is chunk 1, Cbmat(:,2) is chunk 2, etc.

This will break if Cb is not divisible by n.