feat: add qubit discard/measure methods
#580
Conversation
I think this could be nicer to use in examples than the current inout usage. benefits: - importing qubits gets you all the methods - with an LSP you get auto complete on `qubit.` - Extends nicely to other qubit types that might have different interfaces
ss2165
requested review from
qartik and
mark-koch
and removed request for
qartik
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #580 +/- ##
==========================================
- Coverage 91.64% 91.58% -0.06%
==========================================
Files 59 59
Lines 6269 6311 +42
==========================================
+ Hits 5745 5780 +35
- Misses 524 531 +7 ☔ View full report in Codecov by Sentry. |
| @@ -17,7 +17,7 @@ def test_basic(validate): | |||
| def foo(q: qubit) -> None: ... | |||
|
|
|||
| @guppy(module) | |||
| def test(q: qubit @owned) -> qubit: | |||
| def test(q: qubit @ owned) -> qubit: | |||
There was a problem hiding this comment.
it appears my formatter has been at all the decorators. I can undo this if necessary
guppylang/prelude/quantum.py
Outdated
|
|
||
| @guppy | ||
| @no_type_check | ||
| def cx(self: "qubit", target: "qubit") -> None: |
There was a problem hiding this comment.
I would suggesttarget.cx(control) instead,
so the n-controlled cases keep the same structure.
target.toffoli(control0, control1)
target.ncx(*controls)
There was a problem hiding this comment.
Whichever way round it is, I see a lot of potential for bugs caused by people getting it the wrong way round.
In general I'm not so keen on forcing multi-qubit operations to have a "distinguished" qubit in this way: in general there's no natural way to select which qubit it should be.
There was a problem hiding this comment.
I think that's very valid. How about we restrict it a specific set of special single qubit operations: alloc (already the case), discard, measure (non destructive), reset, and any variants of those above (e.g. measure_reset, destructive measure). These special set can be implemented in terms of the existing inout functions as in this PR currently.
ss2165
changed the title
qubit discard/measure methods
🤖 I have created a release *beep* *boop* --- ## [0.13.0](v0.12.2...v0.13.0) (2024-11-12) ### ⚠ BREAKING CHANGES * `prelude` module renamed to `std` ### Features * add `qubit` discard/measure methods ([#580](#580)) ([242fa44](242fa44)) * Add `SizedIter` wrapper type ([#611](#611)) ([2e9da6b](2e9da6b)) * conventional results post processing ([#593](#593)) ([db96224](db96224)) * Improve compiler diagnostics ([#547](#547)) ([90d465d](90d465d)), closes [#551](#551) [#553](#553) [#586](#586) [#588](#588) [#587](#587) [#590](#590) [#600](#600) [#601](#601) [#606](#606) * restrict result tag sizes to 256 bytes ([#596](#596)) ([4e8e00f](4e8e00f)), closes [#595](#595) ### Bug Fixes * Mock guppy decorator during sphinx builds ([#622](#622)) ([1cccc04](1cccc04)) ### Documentation * Add DEVELOPMENT.md ([#584](#584)) ([1d29d39](1d29d39)) * Fix docs build ([#639](#639)) ([bd6011c](bd6011c)) ### Miscellaneous Chores * Manually set last release commit ([#643](#643)) ([b2d569b](b2d569b)) ### Code Refactoring * rename prelude to std ([#642](#642)) ([1a68e8e](1a68e8e)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). --------- Co-authored-by: Agustín Borgna <agustin.borgna@quantinuum.com>
Extends the fact that allocation is tied to the
qubittype by implementing measure/discard methods.Note this includes a drive-by
measure_resetoperation which avoids resetting the qubit and just flip, might not be something to add in this PR.