Average or sum specific-non-sequential columns

Is there a straight forward quick way to do calculations (sum, mean, etc) on columns that are not in a sequence by avoiding unwanted columns?

For example, if I have 100 rows x10 columns matrix, I would like to do calculations (mean or sum) of all rows in column 1-2, 4, 6-8, and 10. I want to skip column 3,5, and 9.

The matrix of interest looks something like this: Matrix(1:100,1:10), and the columns to be skipped are in an array of their own such as this BadColumns=[3,5,9], I need help of how to avoid the bad columns by implementing the BadColumns array in the code

TheMean=mean(Matrix(:,:),2) %but avoid [3,5,9] in second dimension

Answer 1

[KSSV]

Let A be your 100x10 matrix. You can remove the said columns from he matrix and use the functions to get what you want.

idx = [3 5 9] ;  % index of columns which you dont want to consider 
A(:,idx) = [] ;    % Remove those columns 

Now A has only your required columns. You can get what you want.

[Lina Koronfel]

Perfect Thanks alot!!

Answer 2

[Image Analyst]

I would have done it somewhat differently than KSSV. You don't need to change your variable - you can keep it the same, just tell mean() what columns to include in the sum:

data = randi(9, 100, 10); % Create sample data
% I have 100 rows x10 columns matrix, 
% I would like to do calculations (mean or sum) of all rows in column 1-2, 4, 6-8, and 10. 
% I want to skip column 3,5, and 9.
columnsToInclude = [1, 2, 4, 6:8, 10]; % Whatever you want.
theColumnMeans = mean(data(:, columnsToInclude), 1)

[Lina Koronfel]

Thank you for the solution. In my situation I think KSSV is more straight forward because I already have the array idx of the columns to ignore. In reality it is maybe 100 out of 250 columns that need to be removed, so creating another array of columns to include maybe an extra step for me. But I'm sure this solution is more straight forward in another situation. Thanks again and cheers :)