Question

I’m trying to analyze data sets that are obtained from CSV files. After the data is read into matlab, I am left with a variable of my data only. The number of columns and rows changes between each file. Is there a way to average each column and then create a variable for the one with the closest average to a certain value? and then also select the columns directly before and after this middle column and create variables for them, as well as create a variable for the column with the lowest average? Currently, I am selecting the columns manually and creating a variables for them that way.

For example:

I have this table of numbers. (I used the same number in each column for sake of easy averaging in this example.

    1     2     3     4     5     

    1     2     3     4     5

    1     2     3     4     5

    1     2     3     4     5

    1     2     3     4     5

Let’s say I want the column whose average is closest to 3.2

That column would be column 3 whose average is 3. Then I would want the code to select the column before (column 2) and the column after (column 4). As well as the column with the lowest average (column 1)

Answer 1

First get the averages (I assume the data matrix is in variable X):

Xmns = mean(X);

Then to find the minimum, use “min”:

[val,ind] = min(Xmns);

“val” holds the minimum value, “ind” the corresponding index in Xmns, which is the corresponding column.

To find the column mean closest to a particular value, again you can use min:

[val,ind] = min(abs(Xmns-key_val));

Now “ind” holds the column index with mean closest to “key_val”. The next column is just “ind+1” and the previous “ind-1” – just be sure to check you are not beyond the ends of the matrix (i.e. ind may already be 1 or size(X,2)).

Also, given the column index “ind”, to create a new variable with that column, you just use:

sc= X(:,ind);

and if you want to remove that column from X:

X(:,ind) = [];

and that is all.