created solovay kitaev algorithm #133
Conversation
created SV algorith implementation with temporary testing scaffolding at the bottom, doesn't work right now -> approx matrix is considerably off from target
created SV algorithm implementation with temporary testing scaffolding at the bottom, doesn't work right now -> approx matrix is considerably off from target
using updated code with proper testing scaffolding (prints out approx and target array)
| best_dist = dist | ||
| best_gate = gate | ||
|
|
||
| return best_gate |
There was a problem hiding this comment.
The approximating sequence length should not be 1 and is set by us. The user might have a requirement for more depth, but we will set this initially as <= 16. Ref - https://arxiv.org/pdf/quant-ph/0505030
...
I also believe @PranavTupe2000's code could be beneficial here, as discussed in our weekly meeting
| for gate in w_sequence[1:]: | ||
| w_approx = w_approx * gate | ||
|
|
||
| result_sequence.extend([best_v, best_w, v_approx.dagger(), w_approx.dagger()]) |
There was a problem hiding this comment.
I believe the result_sequence variable should also contain the
|
|
||
| return result_sequence, final_error | ||
|
|
||
| def solovay_kitaev(target: np.ndarray, basic_gates: List[np.ndarray], depth: int = 3) -> List[np.ndarray]: |
There was a problem hiding this comment.
For a smoother interface, I believe basis_gates parameter should accept the BasisSet enum as a param and internally derive the basis gates from the BASIS_GATE_MAP
Fixes #126
Summary of changes