| Model Predictive Control Toolbox | |
Scales a matrix automatically or by specified mean and standard deviation.
Syntax
[ax,mx,stdx] = autosc(x)
sx = scal(x,mx)
sx = scal(x,mx,stdx)
rx = rescal(x,mx)
rx = rescal(x,mx,stdx)
Description
autosc scales an input matrix or vector x by its column means (mx) and standard deviations (stdx) automatically and outputs mx and stdx as options. By using scal, the input can also be scaled by some specified means and/or standard deviations. rescal converts scaled data back to original data.
Output mx is a row vector containing the mean value for each column of x while stdx is a row vector containing the standard deviation for each column. Outputs ax and sx are obtained by dividing the difference of each column of x and the mean for the column by the standard deviation for the column, i.e., ax(:,i) = (x :,i) - mx(i)/stdx(i). Output rx is determined by multiplying each column of x by the corresponding standard deviation and adding the corresponding mean to that product.
If only two arguments are specified in scal or rescal, x is scaled by specified means (mx) only.
Example
See mlr for an example of the use of these functions.
| appmod | cmpc | |