Using for-loop to extract data from every column in a matrix

Hi,

I'm new to using for-loops - can anyone help with my problem.

I have a matrix, 18000X31. I want to be able to write a for loop that for each column, a new variable is created that includes the data from the rows 3000-13600. Any help would be gratefully appreciated.

Thanks in advance!

[Stephen23]

"I have a matrix, 18000X31."

So you have a matrix, the basic data format for easily storing and efficiently processing data within MATLAB (in fact even the name comes from "MATrix LABoratory", giving a big hint that using matrices is probably a very good idea).

"I want to be able to write a for loop that for each column, a new variable is created that includes the data from the rows 3000-13600."

So you want to replace your simple matrix (which is trivial and efficient to access using indexing) with a whole bunch of dynamically created variables, thus making accessing your data slow, complex, inefficient, obfuscated, and difficult to debug.

Why would you want to do that?

"Any help would be gratefully appreciated."

https://xyproblem.info/

https://meta.stackexchange.com/questions/66377/what-is-the-xy-problem

https://en.wikipedia.org/wiki/XY_problem

Explain what you are actually trying to achieve (not how you think you might achieve it).

[Edward Keavney]

Hi Stephen,

My question was in relation to a larger project that I'm working on and I thought using a for-loop would help to allow me to step-by-step work through my problem.

I'm trying to achieve isolating the data in the rows 3000-13600 for all of the 31 columns so that I can then take the mean average of every 100 readings in the newly created matrix so that I eventully produce a 106X31 matrix.

I obvioulsy thought breaking it down into steps using a for-loop would be the best way to achieve that, but I'm happy to learn more about what the best practise would be.

Thanks, Ed.

[Walter Roberson]

mean100 = reshape(mean(reshape(M(3000:13599, :), 100, [], size(M,2)), 1), [], size(M,2));

Watch out for that endpoint !

[Edward Keavney]

@Walter Roberson Fantastic! Thanks for your help, this works perfectly!

[Stephen23]

"I thought using a for-loop would help to allow me to step-by-step work through my problem."

There is no problem with using a FOR loop. Or reshaping, as Walter Roberson showed you.

But you should avoid creating "new variables" as you described in your question.

Answer 1

[Walter Roberson]

M = randi([0 9], 180, 31);   %example

letters = ['A':'Z' 'a' : 'z'];
numletters = length(letters);

for col = 1 : size(M, 2)
    randname = letters( randi(numletters, 1, 32) );
    tosave.(randname) = M(:,col);
end

tname = tempname() + ".mat";
save(tname, '-struct', 'tosave');
load(tname)

whos

  Name                                    Size            Bytes  Class     Attributes

  BBvQHvyAMsruyGmPBCAtGFwBICGcDREY      180x1              1440  double              
  FBWVzSBqpSgEroLkseFBwLDtWWXsxlxD      180x1              1440  double              
  HkpsAkMtzUJwGFquzeVpqhJCkyMnwQIJ      180x1              1440  double              
  HwsuyfFDnyaOprLHwveGAvIAPaahStOH      180x1              1440  double              
  JpXqzfTZKfscSQxjQpzcyAXkGFrdNDaY      180x1              1440  double              
  KwPHJsOOHqmeABtRqJNRPuIoWXKmptiq      180x1              1440  double              
  KyHyAlLpoEqjClqVDHQSfSSChHZRxIQz      180x1              1440  double              
  LuwGMmEbgbVnTxKWysgVqWQUBSAUOqNU      180x1              1440  double              
  M                                     180x31            44640  double              
  PSlMFIADpFbonEerutSludFKtrdVvpkd      180x1              1440  double              
  RwdXMqsNjgGaWvbTWCkAXgcSUxHaxqiz      180x1              1440  double              
  TcYHMTGMNhqxILMMhOcCxpJLfgcQbxJn      180x1              1440  double              
  WYetiFJgfEdvjNNcCWMMjtSyxWXsBnCn      180x1              1440  double              
  WcIWJBEpPcFTAtrQDARnBpINiiOzsLOO      180x1              1440  double              
  YeOazlBpsMqHbPwnFUZpDFJEWqyGpuJi      180x1              1440  double              
  YgdZRTlEDdUWbOfquxxuYplYuXjddeKX      180x1              1440  double              
  ZmrbXkYCgqFjfNDsbNviFeyUPjKTgapb      180x1              1440  double              
  aWsXNSxgZZheMVhkGZzcmtvqTJXBlaUv      180x1              1440  double              
  abPmUvvAaIQHsRIYxmNhdeNWRjWimFLe      180x1              1440  double              
  aiZCTrGsddtyrlpMiLJCMHVPxDQLyfex      180x1              1440  double              
  asBxWZIEHuIPHhTSTepsOeRqpIjoOdcI      180x1              1440  double              
  cmdout                                  1x33               66  char                
  col                                     1x1                 8  double              
  jPXOPvgNgBpTvPeIdtZdzMzpYTDARNkb      180x1              1440  double              
  kHVdaqCsycDjQeAhzsrxzpUwbyFWWsnc      180x1              1440  double              
  lZsnQggmqCwhgprroPTIEXNYZDsljuKo      180x1              1440  double              
  letters                                 1x52              104  char                
  lxCLtgkKcHWuxyvObWQOvoesroyYGegT      180x1              1440  double              
  numletters                              1x1                 8  double              
  ooOGyGzeKmhUMXXGYDjehlkqaWsIaIzc      180x1              1440  double              
  qujcfWwyEHTAhwEspdcUqxPxVjiFBYhb      180x1              1440  double              
  randname                                1x32               64  char                
  sReqMWIVCMwWweFeFOJCcijOdpnHeNHi      180x1              1440  double              
  tname                                   1x1               308  string              
  tosave                                  1x1             49848  struct              
  vIljIfxErWOCoshTfilYvCWfCDtGestN      180x1              1440  double              
  vpvYyvIjGYuBFtMTrAiywIlmjtUMQrNq      180x1              1440  double              
  wFiDXZhNtaSNBGFScNPHNCbgKrZkmrTf      180x1              1440  double              
  xmPutkbkEHAkWqvJqyZljSFohnXBrQyn      180x1              1440  double              

Now what? What are you going to do with sReqMWIVCMwWweFeFOJCcijOdpnHeNHi now that you have it?

Answer 2

[Image Analyst]

@Edward Keavney, do this for your matrix m:

[rows, columns] = size(m)
for col = 1 : columns
    fprintf('Processing column %d of %d.\n', col, columns);
    % Create a new variable with only this particular column,
    % but only with rows 3000 to 13600:
    thisColumn = m(3000 : 13600, col);    
    % Now process this column somehow, 
    % like getting its mean or whatever you want to do.
end
fprintf('All done!\n');