Mean of mutiple variables

I have 5 '.mat' data files where each datafile contains 1 by 19 cell (each cell size = 512 by 9). I wanted to find the mean of the 4th column of 1st cell from all 5 data files.

Here is a look at the code.

 load('data1a.mat');load('data2.mat');load('data3.mat');load('data4.mat');load('data5.mat');...
% Extracting the 4th column
a=data1a{1,1}(:,4);b=data2{1,1}(:,4) and so on...
    m=(a+b+c+d+e)/5

Now people in stoneage might used to do this kind of laborious exercise. I am wondering if anyone could give a one/two liner answer for finding mean which is not naive like mine.

[KALYAN ACHARJYA]

Please attach one mat sample file.

[dpb]

NB: If all have the same number of observations, then the overall mean of x is the mean of means so don't actually have to have all the data at once.

[Stephen23]

Copy and pasting code like this is a sign that you are doing something wrong:

load('data1a.mat');load('data2.mat');load('data3.mat');load('data4.mat');load('data5.mat');...

Learn to use arrays and indexing! For example:

https://www.mathworks.com/help/matlab/import_export/process-a-sequence-of-files.html

[ali akbar]

please see attached.

Answer 1

[dpb]

d=dir('data*.mat');        % or similar list. Name files so can use appropriate wildcard expression
total=0;
for i=1:numel(d)
  data=load d(i).name;
  total=total+structfun(@(x)sum(x{1}(:,4)),data);
end
mn=mean(total);

Above does the summation; same result if were to use mean in the anonymous function as well if same number elements in each cell array.

If the variables were named consistently would be easier to code; above handles whatever they are inside the .mat files as long as just one variable per file.

[ali akbar]

thank you. matlab throwing a syntax error on the second line of the 'for' loop.

[dpb]

so, debug it. (Exercise for Student :) )

Accidentally wrote it as command syntax instead of function form needed.