Add GF2Multiplication bloq for multiplication over GF($2^m$)
#1436
Conversation
| qubits in each of the two input registers $a$ and $b$ that should be multiplied. | ||
|
|
||
| Registers: | ||
| - $x$: Input THRU register of size $m$ that stores elements from $GF(2^m)$. |
There was a problem hiding this comment.
format this section like Args so it can be parsed by the autgen thing. The register names shouldn't be latex anywho since they are indeed the python strings 'x', 'y', and 'result' that we are documenting.
| ctrl, result[i] = bb.add(Toffoli(), ctrl=ctrl, target=result[i]) | ||
| x[j], y[i - j] = ctrl[0], ctrl[1] | ||
|
|
||
| # Done :) |
| @@ -69,7 +69,7 @@ def adjoint(self) -> 'Bloq': | |||
| return Free(self.dtype, self.dirty) | |||
|
|
|||
| def on_classical_vals(self) -> Dict[str, int]: | |||
| return {'reg': 0} | |||
| return {'reg': self.dtype.from_bits([0] * self.dtype.num_qubits)} | |||
There was a problem hiding this comment.
This is fine for now, but let's add this to the forthcoming issue to track "friction related to adding new data types that aren't integers". Specifically: having allocate always allocate a bunch of zero bits that get interpreted in whatever encoding the dtype specifies is a fine choice but it's a choice nonetheless that we should probably discuss in the broader context of datatypes and it should be documented somewhere.
There was a problem hiding this comment.
Added to #1437
| def test_gf2_multiplication_classical_sim(m): | ||
| bloq = GF2Multiplication(m) | ||
| GFM = GF(2**m) | ||
| for x in GFM.elements: |
There was a problem hiding this comment.
I think you should be able to use @NoureldinYosri 's helper function for writing tests like these
There was a problem hiding this comment.
What's the helper function? assert_consistent_classical_action ?
There was a problem hiding this comment.
refactored using assert_consistent_classical_action
| for y in GFM.elements: | ||
| xout, yout, resout = bloq.call_classically(x=x, y=y) | ||
| assert x == xout and y == yout | ||
| expected_res = x * y |
There was a problem hiding this comment.
Usually, we can put the "reference behavior" as the bloq's on_classical_vals method and test call_classically's outputs on bloq vs bloq.decompose_bloq().
There was a problem hiding this comment.
added on_classical_vals to the bloq
| bloq_autotester(_gf2_multiplication_symbolic) | ||
|
|
||
|
|
||
| @pytest.mark.parametrize('m', [2, 3, 4, 5]) |
There was a problem hiding this comment.
This is a borderline case, but consider factoring into a slow, exhaustive test marked with slow and a quick check that finishes in < 2 seconds.
|
Addressed all nits, merging now. |
* Add GF2Multiplication for multiplication over GF(2^m) * Fix formatting * Address nits
Part of a larger effort to support galois field arithmetic in qualtran. xref #1419