QAOA MaxCut

[NadavClassiq]

PR Description

Some notes

• Please make sure that you placed the files in an appropriate folder

• And that the files have indicative names.

• Please note that Classiq runs automatic code linting, which may minorly alter some files.

  • If you're familiar with pre-commit, you may run pre-commit install, and then at each commit, your files will be altered in a similar way

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[orsa-classiq]

1 - Why the mixing operation is Digital? and the Phase Analog?

2 - "toward a partition that maximizes the number of cut edges.":

*towards

*maybe minimizes the hamiltonian? instead of the cut. and put the f_obj in negative sign

Reply via ReviewNB

[NadavClassiq]

1—The phase rotation is considered "analog" because it continuously modifies the state's phase in proportion to the cost function, while the mixer is "digital" because it is implemented using discrete quantum gate operations (e.g., RX rotations with discrete parameters) that drive transitions between basis states.

Should I eliminate this terminology? You have used it similarly in knasack

[orsa-classiq]

It was a bit different - the digital was a digital transformation (arithmetrics). The phase is indeed an analog operation. I wouldn't use this terminology because it is a bit confusing, and not needed for the example, but that's up to you

[orsa-classiq]

Line #4.        num_layers: CInt,

maybe you don't need this - use the gammas\betas size instead

Reply via ReviewNB

[NadavClassiq]

I get:

AttributeError: 'CParamList' object has no attribute 'size'

Am I missing anything? Maybe in generative mode?...

I left it as it was for now.

[orsa-classiq]

maybe you should use .len

[orsa-classiq]

Line #6.        params: CArray[CReal, NUM_LAYERS * 2],

Maybe worth an explanation - this is an execution parameter, to be set by the sample method later on in the classical optimization loop

Reply via ReviewNB

[NadavClassiq]

I tried clarifying, please resolve if it is clear.

[orsa-classiq]

Line #9.        allocate(v.size, v)

no need to provide v.size, because it was declared in the main signature

Reply via ReviewNB

[orsa-classiq]

Line #10.        hadamard_transform(v)  # here or aside?

remove the comment

Reply via ReviewNB

[orsa-classiq]

"can be applied to graph-based optimization tasks" - I would change to "can be applied to unconstrained combinatorial optimization tasks"

Reply via ReviewNB

[orsa-classiq]

Line #6.    G.add_nodes_from(graph_nodes)

no need to add nodes, enough to add edges

Reply via ReviewNB

[orsa-classiq]

hadamard_transform() - remove brackets

"As in vanilla QAOA" - what do you mean by that? isn't vanilla QAOA what you implement here?

Reply via ReviewNB

[orsa-classiq]

@NadavClassiq Please also do the following:

1. fix the address of max-cut here: https://github.com/Classiq/classiq-library/blob/main/algorithms/README.md#L4
2. add a 'metadata.json` file so the algorithm will be able to render on the platform
3. remove the old max-cut implementation for applications dir

[github-actions]

🌈 Incredible, @NadavClassiq! You've merged your 16th PR! 🎯🎊

Your ongoing commitment to classiq-library is truly remarkable. You're a driving force in our community! 🚀
Your contributions are helping to shape the future of quantum computing! What exciting features or improvements do you envision next? 🔮

We are grateful for your dedication! 💫