Using mapminmax simulink tool in my neural network

Hello, I am new to neural network toolbox in Simulink. In MAPMINMAX function, it asks for parameters MAXIMUMx [Nx1] and MINIMUMx [Nx1]. As far as I know, when I provide a input given in user guide X = [1 2 4; 1 1 1; 3 2 2; 0 0 0], it being a 4x3 matrix, max_x is 4, and min_x is 0. But when I provide the parameters as 4 and 0 resp, I dont see the correct output as supposed to be by [Y,PS] = mapminmax(X). Also, I tried to provide Minimum_X = [1;1;2;0] and Maximum_X = [4;1;3;0], i.e max and min from each row, but even that doesnt give the same results. It seems I dint understand the function correctly, in Simulink. Can someone please explain what is going wrong here?

[Olatunji Omisore]

Hi, I have similar experience. Presently, I am working on pneumatic teleoperative control using MATLAB Simulink. I want to normalize input signals to -1 and +1, for effective controller tuning however, i couldnt understand how mapminmax block works yet. Anyone to guide?

Answer 1

[Greg Heath]

In the MATLAB NN ToolBox, ROWS are variables and COLUMNS are data vectors.

MINMAX finds the extrema of rows

 X = [1 2 4; 1 1 1; 3 2 2; 0 0 0], 
 minmaxX = minmax(X)

 minmaxX =  1     4
            1     1
            2     3
            0     0

If the row mins and row maxes are different, MAPMINMAX scales the rows between -1 and 1. For example, see rows 1 and 3 of the transformed X:

mapminmaxX = mapminmax(X)

mapminmaxX =

           -1     -0.33333            1
            1            1            1
            1           -1           -1
            0            0            0

Hope this helps.

Thank you for formally accepting my answer

Greg