CircuitOperation: change use_repetition_ids default to False by maffoo · Pull Request #6910 · quantumlib/Cirq

maffoo · 2025-01-02T22:00:16Z

This changes CircuitOperation so that use_repetition_ids defaults to False rather than True. If explicit repetition_ids are provided (to the constructor or to the replace method), we set repetition_ids to True automatically. The motivation for this is that internally we have never used repetition_ids and in fact we have helper methods for creating looped circuit operations since it is easy to forget to pass use_repetition_ids=False to override the default. The history here is that we started implementing repetition_ids at about the same time as we started implementing support for repeated measurement keys, and unfortunately I think we picked the wrong default for the keys. All of our internal experiments such as QEC and looped sampling experiments have set use_repetition_ids=False because repeated keys make it much more efficient to retrieve and manipulate data (one can get a single 3D array of bits for the repeated key instead of separate 2D arrays for each key instance with a different id) and so this better aligns cirq defaults with actual usage.

codecov · 2025-01-02T22:12:54Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.86%. Comparing base (83a8e0e) to head (8e28d22).
Report is 2 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #6910   +/-   ##
=======================================
  Coverage   97.86%   97.86%           
=======================================
  Files        1084     1084           
  Lines       94290    94299    +9     
=======================================
+ Hits        92280    92289    +9     
  Misses       2010     2010

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

pavoljuhas · 2025-01-03T02:41:06Z

        # The eventual number of repetitions of the returned CircuitOperation.
        final_repetitions = protocols.mul(self.repetitions, repetitions)
-        return self.replace(repetitions=final_repetitions, repetition_ids=repetition_ids)
+        print(f"repeat: {repetitions=}. {repetition_ids=}, {use_repetition_ids=}")


Intentional or a debug print out?

Ah yes, this was unintentional. Removed.

pavoljuhas

LGTM, but please see comments before merging. Thank you!

pavoljuhas · 2025-01-04T02:17:34Z

        parent_path: Tuple[str, ...] = (),
        extern_keys: FrozenSet['cirq.MeasurementKey'] = frozenset(),
-        use_repetition_ids: bool = True,
+        use_repetition_ids: Optional[bool] = None,


Nit - consider updating the use_repetition_ids docstring below.

pavoljuhas · 2025-01-04T02:29:26Z

            repetition_ids: List of IDs, one for each repetition. If unset,
                defaults to `default_repetition_ids(repetitions)`.
+            use_repetition_ids: If given, this specifies the value for `use_repetition_ids`
+                of the resulting circuit operation. If the not given, we enable


Suggested change

of the resulting circuit operation. If the not given, we enable

of the resulting circuit operation. If not given, we enable

pavoljuhas · 2025-01-04T02:47:18Z

-      [       [ a: ═══@═══^═══ ](loops=2)    ](loops=2)
+      [       [ 0: ───M───X─── ]                               ]
+0: ───[ 0: ───[       ║   ║    ]────────────────────────────── ]──────────────────────────────
+      [       [ a: ═══@═══^═══ ](repetition_ids=['0', '1'])    ](repetition_ids=['0', '1'])


This might get quite long for large repetition count.
Should we perhaps use (loops=2, use_repetition_ids=True) instead?

PS: No strong preference on my side.

Agreed. Changed to use (loops=2, use_repetition_ids=True) as suggested.

…uantumlib#6910)" Put back the default `use_repetition_ids=True` so we do not make API change without deprecation warning. This reverts commit 5ffb3ad.

…7237) * Revert "CircuitOperation: change use_repetition_ids default to False (#6910)" Put back the default `use_repetition_ids=True` so we do not make API change without deprecation warning. This reverts commit 5ffb3ad. * Add FutureWarning for upcoming change of use_repetition_ids default * Adjust unit tests for default `use_repetition_ids=True`

…n_ids default to False Revert "Flip back to default `use_repetition_ids=True` in CircuitOperation (quantumlib#7237)" This reverts commit 58d9619. This also finalizes quantumlib#7232

…ult to False (#7265) Revert "Flip back to default `use_repetition_ids=True` in CircuitOperation (#7237)" This reverts commit 58d9619. This also finalizes #7232

@pavoljuhas

…lib#6910) Review: @pavoljuhas

…uantumlib#7237) * Revert "CircuitOperation: change use_repetition_ids default to False (quantumlib#6910)" Put back the default `use_repetition_ids=True` so we do not make API change without deprecation warning. This reverts commit 5ffb3ad. * Add FutureWarning for upcoming change of use_repetition_ids default * Adjust unit tests for default `use_repetition_ids=True`

…n_ids default to False (quantumlib#7265) Revert "Flip back to default `use_repetition_ids=True` in CircuitOperation (quantumlib#7237)" This reverts commit 58d9619. This also finalizes quantumlib#7232

CircuitOperation: change use_repetition_ids default to False

fb4bc11

maffoo requested review from a team and vtomole as code owners January 2, 2025 22:00

maffoo requested a review from NoureldinYosri January 2, 2025 22:00

CirqBot added the size: M 50< lines changed <250 label Jan 2, 2025

pavoljuhas reviewed Jan 3, 2025

View reviewed changes

Remove debug print

219bdaf

pavoljuhas approved these changes Jan 4, 2025

View reviewed changes

maffoo added 3 commits January 6, 2025 10:26

Fix circuit diagrams in tests

15fda18

Clarify comment

3ea9493

Merge branch 'main' into u/maffoo/rep-ids

8e28d22

maffoo enabled auto-merge (squash) January 6, 2025 18:31

maffoo merged commit 5ffb3ad into main Jan 6, 2025

maffoo deleted the u/maffoo/rep-ids branch January 6, 2025 18:48

daxfohl mentioned this pull request Jan 23, 2025

Fix mapped measurement keys of repeat_until fields in CircuitOperations #6881

Merged

daxfohl mentioned this pull request Apr 6, 2025

1.5.0 is making a breaking change to the results returned by simulator #7232

Closed

pavoljuhas mentioned this pull request Apr 8, 2025

Flip back to default use_repetition_ids=True in CircuitOperation #7237

Merged

mhucka mentioned this pull request Apr 18, 2025

Improve reporting of errors & efficiency of pr-size-labeler #7288

Merged

This was referenced May 30, 2025

Support serializing sweeps with numpy values #7398

Merged

Update RB to reduce the size of the generated circuits #7411

Merged

BichengYing pushed a commit to BichengYing/Cirq that referenced this pull request Jun 20, 2025

CircuitOperation: change use_repetition_ids default to False (quantum…

1dd6d2c

…lib#6910) Review: @pavoljuhas

NoureldinYosri mentioned this pull request Aug 5, 2025

Add a new transformer that performs random pauli insertion #7558

Closed

NoureldinYosri mentioned this pull request Aug 7, 2025

update the RB fitting method to compensate for the standard deviation induced from the random number generator #7562

Merged

NoureldinYosri mentioned this pull request Sep 11, 2025

Create a gauge for idle moments #7648

Merged

NoureldinYosri mentioned this pull request Sep 25, 2025

fix bug in idle moments gauge #7669

Merged

NoureldinYosri mentioned this pull request Oct 21, 2025

Remove xfail from QSD tests #7712

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

CircuitOperation: change use_repetition_ids default to False#6910

CircuitOperation: change use_repetition_ids default to False#6910
maffoo merged 5 commits intomainfrom
u/maffoo/rep-ids

maffoo commented Jan 2, 2025

Uh oh!

codecov bot commented Jan 2, 2025 •

edited

Loading

Uh oh!

pavoljuhas Jan 3, 2025

Uh oh!

maffoo Jan 3, 2025

Uh oh!

pavoljuhas left a comment

Uh oh!

pavoljuhas Jan 4, 2025

Uh oh!

maffoo Jan 6, 2025

Uh oh!

pavoljuhas Jan 4, 2025

Uh oh!

maffoo Jan 6, 2025

Uh oh!

pavoljuhas Jan 4, 2025

Uh oh!

maffoo Jan 6, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

	of the resulting circuit operation. If the not given, we enable
	of the resulting circuit operation. If not given, we enable

Conversation

maffoo commented Jan 2, 2025

Uh oh!

codecov bot commented Jan 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pavoljuhas left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

codecov bot commented Jan 2, 2025 •

edited

Loading