Properly implement fidelity analysis with SCF

[david-pl]

Closes #592 .

We are now mostly re-using the AddressAnalysis for walking the IR. Except for IfElse statements, where we need to take care when merging results from branches.
To do this, I had to make a small change to AddressAnalysis, but that shouldn't affect anything else.

Probably needs more tests, but basics seem to work quite well.

Also, we might need some wrapper around running this analysis to have a decent UX when trying to study the fidelity of a given program.

@weinbe58 I'm also not sure if I'm being to strict about expecting an AddressReg for qubtis: IList[Qubit] arguments.

[codecov]

Codecov Report

❌ Patch coverage is 91.12150% with 19 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/bloqade/squin/analysis/fidelity/impls.py	86.17%	13 Missing ⚠️
src/bloqade/analysis/fidelity/impls.py	92.10%	3 Missing ⚠️
src/bloqade/qasm2/dialects/noise/fidelity.py	88.88%	3 Missing ⚠️

📢 Thoughts on this report? Let us know!

[github-actions]

☂️ Python Coverage

current status: ✅

Overall Coverage

Lines	Covered	Coverage	Threshold	Status
10391	9132	88%	0%	🟢

New Files

File	Coverage	Status
src/bloqade/analysis/fidelity/impls.py	92%	🟢
src/bloqade/squin/analysis/fidelity/_init_.py	100%	🟢
src/bloqade/squin/analysis/fidelity/impls.py	86%	🟢
TOTAL	93%	🟢

Modified Files

File	Coverage	Status
src/bloqade/analysis/address/analysis.py	90%	🟢
src/bloqade/analysis/fidelity/_init_.py	100%	🟢
src/bloqade/analysis/fidelity/analysis.py	100%	🟢
src/bloqade/qasm2/dialects/noise/fidelity.py	92%	🟢
src/bloqade/squin/_init_.py	100%	🟢
TOTAL	96%	🟢

updated for commit: 053e798 by action🐍

[weinbe58]

Overall looks good, just one comment on the implementation of IfElse. Also I would add a test to make sure that you execute only one branch if the branch value is known at compile time.

[weinbe58]

@weinbe58 I'm also not sure if I'm being to strict about expecting an AddressReg for qubtis: IList[Qubit] arguments.

It depends on the situation, the short has is AddressReg is an IList of all qubits with known addresses PartialIList if some of the qubit in the list have a known address and some have UnknownQubit or if the size of the array is known at compile time and finally UnknownReg you know its an IList of qubit values but you have no idea the size of the list.

[david-pl]

@weinbe58 Thanks for the review. The IfElse impls was somewhat left over from a previous version. It was also missing the return values. I fixed that, added the constant condition evaluation and a test.

It depends on the situation, the short has is AddressReg is an IList of all qubits with known addresses PartialIList if some of the qubit in the list have a known address and some have UnknownQubit or if the size of the array is known at compile time and finally UnknownReg you know its an IList of qubit values but you have no idea the size of the list.

I suppose it's fine then. Basically, the fidelity analysis will give up whenever there is any qubit without a known address. Technically, we could update the fidelity at least for the known qubits, but something's bound to go wrong in this case anyway since at least for one qubit the fidelity will be off.

[weinbe58]

LGTM