Data splitting and saving to different variables

How to split a matrix data column wise and saving each column in different variable using loop. say M is a 3*8 matrix. I want to store these data of M columnwise into a 8 different column matrix, m1,m2,m3,m4,m5,m6,m7 and m8.

[dpb]

Do NOT do this -- use the array indices instead...

[Stephen23]

"I want to store these data of M columnwise into a 8 different column matrix, m1,m2,m3,m4,m5,m6,m7 and m8."

You can certainly do that, if you want to force yourself into writing slow, complex, inefficient, obfuscated, insecure, buggy code that is hard to debug:

https://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

In contrast, indexing is simple and very efficient. So far you have not given any reason why you cannot use indexing.

Answer 1

[Johan]

You can use structure to do that, you can read more about it in the matlab help https://fr.mathworks.com/help/matlab/ref/struct.html

here is an example for your specific question:

M = rand(3,8)

M = 3×8

0.9703 0.0578 0.7766 0.9685 0.2685 0.3171 0.4461 0.7332 0.5987 0.5110 0.3488 0.3910 0.5700 0.3421 0.7714 0.2140 0.0907 0.6547 0.8126 0.9746 0.1461 0.2658 0.0747 0.8221

m(1:8) = struct('data',[]);

m = 1×8 struct array with fields:
    data

for i_col = 1:size(M,2)
    m(i_col).data = M(:,i_col);
end
m.data

ans = 3×1

0.9703 0.5987 0.0907

ans = 3×1

0.0578 0.5110 0.6547

ans = 3×1

0.7766 0.3488 0.8126

ans = 3×1

0.9685 0.3910 0.9746

ans = 3×1

0.2685 0.5700 0.1461

ans = 3×1

0.3171 0.3421 0.2658

ans = 3×1

0.4461 0.7714 0.0747

ans = 3×1

0.7332 0.2140 0.8221

% Note that splitting that little data in structured arrays is not very
% efficient
whos M

  Name      Size            Bytes  Class     Attributes

  M         3x8               192  double              

whos m

  Name      Size            Bytes  Class     Attributes

  m         1x8              1088  struct              

[dpb]

"data in structured arrays is not very efficient"

In storage effeciency, no, but in coding efficiency and use as compared to the alternative posited, the savings are immense.

An even less efficient memory use but sitll more convenient programmatically, would be the table

>> array2table(M)
ans =
  3×8 table
      M1         M2         M3          M4         M5         M6         M7          M8   
    _______    _______    _______    ________    _______    _______    _______    ________
    0.78085     0.4528    0.18342     0.10877    0.93074    0.16397    0.37596    0.012267
    0.83353     0.1733    0.57372     0.81383    0.33064    0.54623    0.99263     0.46515
    0.24278    0.31354    0.37909    0.061772    0.83759    0.19313    0.31973      0.1851
>> whos ans
  Name      Size            Bytes  Class    Attributes

  ans       3x8              2656  table              

>>

But, the extra tools that come along with the table generally far outweigh the overhead cost.

[Johan]

Ah yes I forgot to add the memory wise, thank you for the precision :)

[dpb]

And, for good measure, with the table, OP gets the desired variable names coming along "for free"...

Answer 2

[Steven Lord]

Can you dynamically create variables with numbered names like x1, x2, x3, etc.? Yes.

Should you do this? The general consensus is no. That Answers post explains why this is generally discouraged and offers several alternative approaches.