Scale Matrix to a new range

Question

Possible Duplicate:
MATLAB: how to normalize/denormalize a vector to range [-1;1]
matlab rescale matrix data to -1 to 1

Consider a 100x200 matrix A, is there a builtin function (or set of builtin functions) that will set the range of A to an arbitrary scale?

This question is already answered here. [http://stackoverflow.com/questions/5077800/generate-a-random-number-in-a-certain-range-in-matlab][1] [1]: http://stackoverflow.com/questions/5077800/generate-a-random-number-in-a-certain-range-in-matlab — vins, Jan 16 '13 at 17:06
I don't think this is a duplicate. What do you mean by `builtin`? Do you want a single `builtin` function or can you use series of `builtin` functions? — slayton, Jan 16 '13 at 17:08
@RobertHarvey: that scales only to -1...1. I don't see where I could change it to apply another range — , Jan 16 '13 at 17:09
@woodchips: that questions can't be adapted to a custom range and it refers only to vectors — , Jan 16 '13 at 17:23
Sure it is. If you can scale a matrix to one range, you can scale it to any range you wish. All it takes is addition/subtraction and multiplication/division, nothing more than it took for the other operation. — , Jan 16 '13 at 17:34
Why not create your own function? Do you need a builtin function? — slayton, Jan 16 '13 at 17:50

score 3 · Accepted Answer · edited Jan 17 '13 at 09:11

3

Does this qualify?

R = [.5 .9];
dR = diff( R );

A =  A - min( A(:)); % set range of A between [0, inf)
A =  A ./ max( A(:)) ; % set range of A between [0, 1]
A =  A .* dR ; % set range of A between [0, dRange]
A =  A + R(1); % shift range of A to R

edited Jan 17 '13 at 09:11

dynamic

46,985
55
154
231

answered Jan 16 '13 at 17:16

slayton

20,123
10
60
89

+1 I tought it was easier... – dynamic Jan 16 '13 at 17:17
I have tested it, and the matrix A has totally wrong values :( – dynamic Jan 16 '13 at 17:26
@link, yea I just noticed that, fixed the bug :) – slayton Jan 16 '13 at 17:56
there is an error on your last line, also i don't understand why you use function like minus instead of the simple operator – dynamic Jan 16 '13 at 18:58
Also `rtimes` doesn't seem to be a function, could you run your code before posting please? – dynamic Jan 16 '13 at 19:04
with rtimes do you mean `.*` ? – dynamic Jan 16 '13 at 19:10
I understand why you used function instead of sign, because I said builtin... LOL I believe you misunderstood me :D... I meant with builtin if there was a function like matrixScale builtin in Matlab... But you could have used simply used the math operator. I have edited the answer – dynamic Jan 17 '13 at 09:10

Scale Matrix to a new range

1 Answers1

Linked

Related