Tanuj Khattar

One use case of this compiler would be to support merging parameterized single qubit rotations into a parameterized `PhXZ` gate, as suggested by @Wuggins

Reopening the issue since we have another request from internal team members for this compiler.

From cirq sync: If it's possible to do analytically, let's do it. The result might end up looking very complicated but that's upto users using the analytical compiler.

Also consider moving `Scalar` type from `linear_dict.py` to `type_alias.py`.

Filed a bug in asv: https://github.com/airspeed-velocity/asv/issues/964

What is the expected behavior for putting two measurements with the same key on two different qubits? Do we support repeated measurements on different qubits? My understanding was that a...

> I think allowing the same key on different qubits within a moment is fine though, because operations in a moment are ordered, and the hardware interface layer does support...

So then the original issue raised is not a bug (because measurements with same key in the same moment are well defined) and we should instead fix the sampler error...

@mpharrigan As discussed offline, can you write a brief comment / code piece describing how the simulator using the protocol to act on a numpy array and falling back to...

Discussion from cirq sync: We should do it. In terms of implementations, it's probably best to add a flag to `DensePauliString` which can control whether the internal representation is sparse...