frexp — dissect floating-point numbers into base 2 exponent and mantissa
[f,e]=frexp(x)
real vector or matrix
array of real values, usually in the range 0.5 <= abs(f) < 1.
array of integers that satisfy the equation: x = f.*2.^e
x = f.*2.^e
This function corresponds to the ANSI C function frexp(). Any zeros in x produce f=0 and e=0.
x
f=0
e=0
[f,e]=frexp([1,%pi,-3,%eps])