Implementation of complex numbers with real floating precision

[jgalan]

This is a new class that will allow to work with complex numbers using real precision thanks to the implementation of MPFR library types. The first intention is to be used for the definition of the axion field and propagation calculations.

root [0] TRestComplex::SetPrecision(30)
root [1] TRestComplex a
(TRestComplex &) (1,0i)
root [2] TRestComplex b(1.e-15)
(TRestComplex &) (1e-15,0i)
root [3] TRestComplex c(0,1)
(TRestComplex &) (0,1i)
root [4] TRestComplex d(0,1.e-25)
(TRestComplex &) (0,1e-25i)
root [5] a + b + c + d
(TRestComplex) (1.000000000000001,1.0000000000000000000000001i)
root [6] TRestComplex e = (a+b)/(c+d)
(TRestComplex &) (0,-1.0000000000000009999999999i)

[jgalan]

Approuve plz