Matthew Harrigan

Add n-bit Identity, implement C[I] = I

3

We have a one qubit `Identity` bloq. This should probably let you do an n-bit identity operation and the controlled version of it should be an `n+m` bit identity for...

Propagate rotation epsilon in gate cost

4

The `t_complexity` protocol ignores the `eps` parameter when counting rotations. The solution is to enhance the `QECGatesCost` cost and associated dataclass `GateCounts` to count rotation/eps pairs.

Convenience methods for using dot2tex to make latex figures

3

We use graphviz. There exists a package `dot2tex` that can generate latex code and/or pdf figures from graphviz. This has the benefit of supporting latex math. Here's a little code...

TextbookQPE bloq examples throw error in t_complexity

3

These bloq examples throw an error from `cirq.unitary` when trying to count t gates in textbook qpe - `textbook_qpe_using_m_bits` - `textbook_qpe_from_standard_deviation_eps` - `textbook_qpe_from_precision_and_delta` cc @tanujkhattar ``` ../../cirq_interop/t_complexity_protocol.py:243: in t_complexity ret...

Controlled versions of non-THRU bloqs

14

With the discussion in #1272 #1273 , I have been reminded of how uncomfortable I feel about controlled versions of bloqs with RIGHT or LEFT (allocating / de-allocating) registers. There...

mean_estimation.complex_phase_oracle_test.test_phase_oracle is very slow because of Fxp

3

This is the change: https://github.com/quantumlib/Qualtran/pull/1041/files#r1638968757 Somehow, this has turned the parameterized unit test from a 1 second to a 21 second ordeal on my machine. Is it worth resuscitating our...

I'm trying to find a minimal example of this, but flattening the bloq used in `phasing_via_cost_function_test.test_square_phasing_via_phase_gradient` causes a data type error: ``` E qualtran._infra.composite_bloq.BloqError: AddIntoPhaseGrad register dtypes are not consistent...

Design discussion: Introduce zero-sized registers to support quantum interfaces

9

In https://github.com/quantumlib/Qualtran/pull/1094, the desire to have zero-size registers is introduced. The motivation is to have a consistent set of register names for a "quantum interface" and just set the size...

Provide functionality for determining whether a bloq can be tensor-simulated in theory and in practice. - Add a function that takes any bloq (or `BloqExample`), decomposes it as much as...

In the qubit counting protocol, we take the max of the various widths of all the moments in a circuit. This can give you expressions like `Max(1*n, 2*n)`. If you...