Separate normalized and unnormalized objects

[ba2tro]

Separated the symbolic objects into normalized and unnormalized, added a dispatch of tr from LinearAlgebra.jl for computing the trace of the unnormalized objects and fixed tests according to the new naming convention

[codecov[bot]]

Codecov Report

Merging #50 (1b07957) into master (60d60e2) will increase coverage by 0.39%. The diff coverage is 82.35%.

@@            Coverage Diff             @@
##           master      #50      +/-   ##
==========================================
+ Coverage   63.29%   63.69%   +0.39%     
==========================================
  Files          27       27              
  Lines        1079     1099      +20     
==========================================
+ Hits          683      700      +17     
- Misses        396      399       +3     

Files	Coverage Δ
src/StatesZoo/StatesZoo.jl	50.00% <ø> (ø)
src/StatesZoo/zalm_pair/zalm_pair.jl	27.11% <83.33%> (+4.89%)	:arrow_up:
...ngle_dual_rail_midswap/single_dual_rail_midswap.jl	87.87% <81.81%> (-3.43%)	:arrow_down:

... and 2 files with indirect coverage changes

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more

[Krastanov]

I have been trying to think about names, specifically suffixes, that will lead to the least amount of confusion. My main worry is that U frequently stands for Unitary, and N frequently stands for number of qubits or for dimensionality.

Maybe changing U to W would make sense (for "weighted", given that it is weighted by the probability for success) and just removing the N (given that people usually expect if something is a density matrix for it to be normalized).

Do you have any thoughts on this? Happy to merge as is and then do a renaming PR or to discuss naming here before merging.