Julien Gacon

Some comments and questions: * The docstring code examples are indeed already fixed on main 🙂 * The attributes `num_parameters`, `parameters` and `data` are coming from the `BlueprintCircuit`, which inherits...

Edit: I didn't see the issue explaining the problem with `inverse` (thanks @ElePT!), in this case it would make sense to add the pass manager 👍🏻 ~~Thanks for opening a...

Thanks for the contribution @shravanpatel30! This is indeed a missing feature or wrong documentation. However, I'm not sure this is the right approach to solving it, as this selection might...

> If we switch to using a dictionary for ZZFeatureMap and PauliFeatureMap, could that introduce inconsistencies or issues, given that these classes inherit from NLocal? However, if you believe that...

Thanks for the update! Could you update the code such that there's no default `full` entanglement when the number of qubits is not specified? I think it would be better...

I looked a bit in more detail and it seems the problem is not that `HighLevelSynthesis` doesn't know the basis gates (@alexanderivrii your snippet only works because it contains `u`...

Do you already have some profilings for this? 🚀

> but the depth for the qft example has regressed That could maybe point to different handling of gates with minuscule rotation angles -- maybe the old code considered things...

[9235004](https://github.com/Qiskit/qiskit/pull/12870/commits/92350045eb1d22f1c886f153c7d4d108f1a07631) adds backward compatible handling of the tolerances, and some further error propagation instead of eager unwrapping (in spirit of @raynelfss previous comments). If CI passes, I think it should...

Now that #12870 is almost ready, it would probably be best to directly implement these changes in Rust, once it is merged 🙂