Scale/Normalize values in matrix between 10^-6 and 10^-5

Hello,

I have a matrix Data (90X150) and I want its values to be normalized. I wrote the code below:

% Normalization
min_Data = min(Data,[],2); % min for each row
max_Data = max(Data,[],2); % max for each row
Data_Norm=(Data - min_Data)./(max_Data- min_Data);

Based on the above code, the normalization is done between 0-1, but I want the normalization to be done between 10^-6 and 10^-5. How is this done? Your help is valuable.

Answer 1

[Askic V]

Hello,

this is a matter of a simple linear transformation y = m*x+n.

if you want to normalize data in interval [a,b] to interval [c,d], then the following code would work:

x = [-1 2 4 0 5 6] % input data

x = 1×6

-1 2 4 0 5 6

a = min(x);
b = max(x);
c = 3; % new interval limit
d = 4; % new interval limit

m = (d-c)/(b-a);
n = (c*b-d*a)/(b-a);

y = m*x+n

y = 1×6

3.0000 3.4286 3.7143 3.1429 3.8571 4.0000

Of course, the most preffered way is to use built in functions:

x = [-1 2 4 0 5 6] % input data

x = 1×6

-1 2 4 0 5 6

c = 3; % new interval limit
d = 4; % new interval limit
y2 = rescale (x,c,d) % rescaled data

y2 = 1×6

3.0000 3.4286 3.7143 3.1429 3.8571 4.0000

[stelios loizidis]

Okay. Thank you for your help!!