`backend.calculate_overlap` should not use `np.abs`

[andrea-pasquale]

I've looked at the implementation of backend.calculate_overlap

qibo/src/qibo/backends/numpy.py

Lines 748 to 751 in 5b3b6da

	def calculate_overlap(self, state1, state2):
	return self.np.abs(
	self.np.sum(self.np.conj(self.cast(state1)) * self.cast(state2))
	)

it is not clear to me why we need also np.abs the inner product for a Hilbert space doesn't require to take the norm usually. Unless I am missing something I think that we should remove np.abs for better clarity.

[renatomello]

To avoid ambiguity, I propose renaming this method as something like overlap_pure_states or inner_product_pure_states, because the definition of overlap becomes non-unique when dealing with mixed states and density matrices.