How can a convert a data set to percentage of maximum absolute value?

i have a rather large data set and would like to have it converted to percent of maximaum value in each group.

ie. for the following example:

TR02 0 0.44 -0.40

TR02 22.5 0.66 0.55

TR02 67.5 0.92 1.75

TR02 90 1.06 3.25

TR02 112.5 1.72 5.37

TR02 157.5 2.91 6.96

TR02 180 1.00 2.57

TR02 202.5 1.94 0.65

TR02 247.5 0.47 -0.45

TR02 270 0.39 -1.48

TR02 292.5 0.81 -1.86

TR02 337.5 0.43 -1.60

TR03 0 0.25 -0.38

TR03 22.5 2.83 1.21

TR03 67.5 1.19 2.85

TR03 90 0.68 2.79

TR03 112.5 1.81 4.86

TR03 157.5 2.76 7.24

TR03 180 2.55 3.26

TR03 202.5 0.90 1.65

TR03 247.5 2.20 -1.21

TR03 270 0.73 -1.12

TR03 292.5 0.66 -1.60

TR03 337.5 1.21 -1.63

TR04 0 1.00 -0.99

TR04 22.5 0.84 0.19

TR04 67.5 0.92 0.91

TR04 90 1.09 2.37

TR04 112.5 2.27 4.41

TR04 157.5 1.13 3.82

TR04 180 1.27 1.59

TR04 202.5 1.42 0.63

TR04 247.5 1.12 0.84

TR04 270 0.86 -1.01

TR04 292.5 0.91 -1.22

I would like to calculate the maximaum absolute value for each "TR" (must be absolute as I do not want percentages greater than and absolute value of 100) and convert all the values for that "TR" to percent of that maximum retaining the +/- sign.

Answer 1

[Fangjun Jiang]

a=rand(5)-0.5;
b=max(abs(a(:)));
c=a/b
d=c*100