How can normalize matrix in range [-1,1]?

the matrix have dimension 8192*51975 contains positive and negative values

Answer 1

[Matt Macaulay]

Divide the matrix by it's largest absolute value. For example, take the matrix:

A = 200*(.5-rand(3))

Now normalise it:

A = A/max(max(abs(A)))

Answer 2

[KSSV]

To normalize between the limits [r0,r1] i.e [-1,1]. Do the following:

r0 = -1 ; r1 = +1 ;
% First normalize to [0 ,1] 
a = rand(10,1) ;
range = max(a) - min(a);
a = (a - min(a)) / range;
% Then scale to [-1,1]
range2 = r1-r0;
a = (a * range2) + r0;