Restrict mixer calibration to target qubits' channels by jevillegasd · Pull Request #1581 · qiboteam/qibocal

jevillegasd · 2026-07-02T09:12:29Z

Previously calibrate_mixers configured and calibrated every channel in the platform regardless of targets. Now only the drive/probe channels of the requested qubits are touched (all targets together, to preserve intermodulation products), the report table is split per qubit, and offset_i/offset_q updates propagate to sibling channels sharing a frequency-multiplexed readout port, since those get recalibrated in hardware too.

To do this I patch the _channels_by_module method from qibolab.instruments.qblox.cluster to keep only the channels associated to the target qubit(s).

Previously calibrate_mixers configured and calibrated every channel in the platform regardless of `targets`. Now only the drive/probe channels of the requested qubits are touched (all targets together, to preserve intermodulation products), the report table is split per qubit, and offset_i/offset_q updates propagate to sibling channels sharing a frequency-multiplexed readout port, since those get recalibrated in hardware too.

codecov · 2026-07-02T09:17:35Z

Codecov Report

❌ Patch coverage is 0% with 61 lines in your changes missing coverage. Please review.
✅ Project coverage is 91.54%. Comparing base (85a7b6f) to head (8dea4d5).
⚠️ Report is 295 commits behind head on main.

Files with missing lines	Patch %	Lines
src/qibocal/protocols/calibrate_mixers.py	0.00%	61 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1581      +/-   ##
==========================================
- Coverage   91.84%   91.54%   -0.30%     
==========================================
  Files         147      147              
  Lines       11635    11679      +44     
==========================================
+ Hits        10686    10692       +6     
- Misses        949      987      +38

Flag	Coverage Δ
unittests	`91.54% <0.00%> (-0.30%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
src/qibocal/protocols/calibrate_mixers.py	`1.89% <0.00%> (-0.50%)`	⬇️

... and 5 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

jevillegasd · 2026-07-02T09:20:38Z

How I handle the update is a bit dummy running first in a reduced version of the hardware (only the target channels) and later in the full cluster to update offsets (and make sure all mixers that share an LO get the same offset update).

RoyStegeman · 2026-07-09T08:08:49Z

Hi @jevillegasd, to for me to understand, what is the main motivation for this? Is it the time saved or something else?

jevillegasd · 2026-07-09T11:49:17Z

Hi @jevillegasd, to for me to understand, what is the main motivation for this? Is it the time saved or something else?

Hey @RoyStegeman , you may recall we talked about needing this when you have more pulses than sequencers. This may happen for example if in a platform we defined the ge, ef, and cr pulses, for two qubits, this already exhausts the available sequencers, even if they are not used simultaneously in a practical application. The mixer calibration would fail in such cases because it loads by default all existing sequencers.

Additionally, when you re-calibrate the mixer in practice you need to re-calibrate your pulse amplitudes and drag coefficients, so so it does make sense to have some level of control of which ports are being re calibrated.

RoyStegeman · 2026-08-21T08:17:03Z

            sequencer.sideband_cal()


+def _target_channels(platform: CalibrationPlatform, targets: list[str]) -> set[str]:


Suggested change

def _target_channels(platform: CalibrationPlatform, targets: list[str]) -> set[str]:

def _target_channels(platform: CalibrationPlatform, targets: list[QubitId]) -> set[ChannelId]:

Using these type aliases improves readability. I point out just here, but of course it should be done for all instances where there is now str that can be a more specific type alias.

RoyStegeman · 2026-08-21T08:26:11Z

+
+
 def _acquisition(
    params: CalibrateMixersParameters,


Suggested change

params: CalibrateMixersParameters,

_params: CalibrateMixersParameters,

to indicate it's not consumed

RoyStegeman · 2026-08-21T09:19:13Z

+    ``Cluster.configure`` always configures every channel in
+    ``Cluster._channels_by_module`` (the whole platform, by default), so the map is
+    narrowed temporarily and restored right after, to avoid touching unrelated modules
+    or leaking the restriction into later experiments.


Please mention that this handling is needed because Cluster._channels_by_module is a cached property.

Also mention that try is there not because you expect it to fail, but just to be absolutely certain that if it does then the previous cluster._channels_by_module is still restored. Or if there is some other reason, explain that one.

RoyStegeman · 2026-08-21T09:45:24Z

+        target_channels = _target_channels(platform, [target])
+        mapping[target] = sorted(
+            {
+                cluster._modules[slot].short_name


cluster._modules is not cached so you're generating the same dict it for each target, it can be moved out of the loop.

RoyStegeman · 2026-08-21T10:51:39Z

+    # round2
+    # offset_i/offset_q are per physical port, shared by every qubit multiplexed on it
+    # Calibrating one recalibrates all of them in hardware, so propagate the update to
+    # update to every channel on a sahred calibrated port, not just the targets.


Suggested change

# update to every channel on a sahred calibrated port, not just the targets.

# update to every channel on a shared calibrated port, not just the targets.

RoyStegeman · 2026-08-21T10:57:22Z

+        cal = final_cal[mod_name]
+        for ch_id, seq_id in channels.items():
+            ch = channel_mixer(ch_id)
+            port = PortAddress.from_path(cluster.channels[ch_id].path).ports[0] - 1


Suggested change

port = PortAddress.from_path(cluster.channels[ch_id].path).ports[0] - 1

port = PortAddress.from_path(ch.path).ports[0] - 1

RoyStegeman · 2026-08-21T10:59:15Z

+    for slot, channels in results.sequencer_map.items():
+        mod_name = cluster._modules[slot].short_name
+        if mod_name not in final_cal:
+            continue  # Skip if no calibration data for this module


This is duplicated in the second round. The mod_name handling can be done only once

RoyStegeman · 2026-08-21T11:02:53Z

        data: Acquisition data

    Returns:
        Empty results


These are not empty

Copilot

Pull request overview

Restricts Qblox mixer calibration and reporting to channels associated with requested qubits.

Changes:

Filters configured channels by target.
Produces per-target module reports.
Propagates port-level offsets to multiplexed sibling channels.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Co-authored-by: Roy Stegeman <roystegeman@live.nl>

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

for more information, see https://pre-commit.ci

jevillegasd requested review from a team July 2, 2026 09:12

RoyStegeman self-requested a review August 19, 2026 10:52

RoyStegeman reviewed Aug 21, 2026

View reviewed changes

RoyStegeman requested a balanced review from Copilot August 21, 2026 11:19

Copilot started reviewing on behalf of RoyStegeman August 21, 2026 11:20 View session

Copilot AI reviewed Aug 21, 2026

View reviewed changes

Comment thread src/qibocal/protocols/calibrate_mixers.py

Comment thread src/qibocal/protocols/calibrate_mixers.py

Comment thread src/qibocal/protocols/calibrate_mixers.py Outdated

jevillegasd and others added 5 commits September 4, 2026 11:53

Filter channels by RF type in mapping

b474f00

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Update src/qibocal/protocols/calibrate_mixers.py

d38e862

Co-authored-by: Roy Stegeman <roystegeman@live.nl>

Update src/qibocal/protocols/calibrate_mixers.py

6c5f413

Co-authored-by: Roy Stegeman <roystegeman@live.nl>

Improve comments in calibrate_mixers.py

2fba261

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

[pre-commit.ci] auto fixes from pre-commit.com hooks

8dea4d5

for more information, see https://pre-commit.ci

		sequencer.sideband_cal()


		def _target_channels(platform: CalibrationPlatform, targets: list[str]) -> set[str]:

	params: CalibrateMixersParameters,
	_params: CalibrateMixersParameters,

	# update to every channel on a sahred calibrated port, not just the targets.
	# update to every channel on a shared calibrated port, not just the targets.

	port = PortAddress.from_path(cluster.channels[ch_id].path).ports[0] - 1
	port = PortAddress.from_path(ch.path).ports[0] - 1

Uh oh!

Conversation

jevillegasd commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov Bot commented Jul 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

jevillegasd commented Jul 2, 2026

Uh oh!

RoyStegeman commented Jul 9, 2026

Uh oh!

jevillegasd commented Jul 9, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

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!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

jevillegasd commented Jul 2, 2026 •

edited

Loading

codecov Bot commented Jul 2, 2026 •

edited

Loading