Make apply! work when used with any order of subsystem indices

[ba2tro]

apply! needs to be able to permute the operator based on the order of the indices of the qubits/subsystems, e.g., doing CNOT_2->1. Resolves QuantumSavory/QuantumSavory.jl/issues/39 cc @Krastanov

[ba2tro]

Would it be better to just have an optional boolean keyword arg in apply! like perm instead of running the for loop for every multi-qubit system?

[Krastanov]

Thanks for starting this!

The apply! on Operator, and the apply! with SuperOperator both suffer from the same bug and should be fixed.

I would suggest not adding any keyword arguments to apply!. For now we can just pay the cost of this inefficient implementation. The proper fix for the slowness would be to use a sufficiently general stridded array multiplication.

[Krastanov]

Also, I would suggest having a test that has three operators and a permutation that is not its own inverse, e.g. apply!(a⊗b⊗c, (2,3,1),A⊗B⊗C) ≈ (C*a)⊗(A*b)⊗(B*c). The current CNOT tests do not cover such cases (because permutations of 2 elements are their own inverses).

[codecov]

Codecov Report

Merging #133 (3c7e645) into master (fe00cb9) will increase coverage by 0.04%.
Report is 3 commits behind head on master.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #133      +/-   ##
==========================================
+ Coverage   93.62%   93.66%   +0.04%     
==========================================
  Files          25       25              
  Lines        3105     3125      +20     
==========================================
+ Hits         2907     2927      +20     
  Misses        198      198              

Files Changed	Coverage Δ
src/apply.jl	100.00% <100.00%> (ø)

... and 2 files with indirect coverage changes

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

[Krastanov]

I did some minor cleanup in the last commit. Thanks for taking care of this!