[Bug] CUDA tensor validation rejects iqp/iqp-z despite core encoder support

[SuyashParmar]

QdpEngine.encode() supports iqp and iqp-z in qdp-core, but the CUDA tensor validation path in pytorch.rs currently rejects these methods and only allows amplitude, angle, and basis.

This creates inconsistent behavior:

• CPU/list/NumPy paths can use iqp/iqp-z
• CUDA tensor path fails early with an unsupported-method error

Expected behavior
CUDA tensor inputs should allow iqp and iqp-z when dtype/shape/device constraints are satisfied (same style as other supported encodings).

[SuyashParmar]

cc- @ryankert01

[alisha-1000]

Hi @SuyashParmar,

Thanks for reporting this — I took a quick look and it does appear that QdpEngine.encode() supports iqp and iqp-z at the core level, while the CUDA tensor validation path in pytorch.rs restricts the allowed methods to amplitude/angle/basis.

I’d like to work on aligning the CUDA validation logic with the core encoder support, ensuring dtype/shape/device constraints remain consistent. I’ll also add tests to cover the CUDA tensor path for these methods.

Let me know if that matches the intended behavior before I proceed.

[SuyashParmar]

@alisha-1000 Thanks for pointing that out!
I’ve actually already started working on aligning the CUDA validation logic in pytorch.rs with the core QdpEngine.encode() support for iqp and iqp-z, while keeping the dtype/shape/device checks consistent. I’m also planning to add test coverage for the CUDA tensor path for these methods.