You'll have to explain that in more detail.
I have 15 matrix called error!m where m goes from 2 to 16.
Each error!m matrix has 84 columns and 97 rows.
What I need is something like:
Create !M matrices with 97 rows and 84 columns called for example error_x_!M where m goes from 2 to 16.
For the matrix error_x_2, it is necessary that the observation (1,1) be equal to the observation (1, 1) of the matrix error1 minus the average of the first column of the matrix error1. For the observation (2, 1) that is equal to the observation (2, 1) of the matrix error1 minus the first column of the matrix of error1 (remains constant). and so on