Repository navigation
Implementing estimator interface #145
Description
What should we add?
Qiskit C++ currently only has sampler interface to run circuit to get sampling bitstrings. Estimator interface is also demanding to get estimation on Pauli strings.
Implementing a BackendEstimatorV2 class as well as BackendSamplerV2 class
(https://github.com/Qiskit/qiskit-cpp/blob/main/src/primitives/backend_sampler_v2.hpp)
and implementing estimator.run() function that will accept an EstimatorPub class as well as SamplerPub (https://github.com/Qiskit/qiskit-cpp/blob/main/src/primitives/containers/sampler_pub.hpp)
An EstimatorPub will have ISA circuit and observables (see https://quantum.cloud.ibm.com/docs/en/api/qiskit-ibm-runtime/estimator-v2)
estimator.run() function will return BackendEstimatorJob class as well as BackendSamplerJob class (https://github.com/Qiskit/qiskit-cpp/blob/main/src/primitives/backend_sampler_job.hpp) containing estimation values.
[added 9 June]
Using SparseObservable class for input pubs.
Implement Backend::run function that accepts std::vector<EstimatorPub> (at least for QRMI backend) to run as estimator program id.
Activity
1ucian0 commentedon Jun 3, 2026
If you want to participate, please follow these guidelines:
#unitaryhack-26in the Unitary Foundation DiscordThe winners will be announced on June 24th.
Happy hacking!
Add BackendEstimatorV2 interface for issue Qiskit#145
rosspeili commentedon Jun 4, 2026
Hey Jun Doi (@doichanj) opened a PR that addresses the issue, including ripple effects and complementary files, docs, code fixes that short the need for follow up PRs.
jsaroni commentedon Jun 4, 2026
Hi Luciano Bello (@1ucian0), Jun Doi (@doichanj),
Thank you for the check! I submitted the PR first complete with quickly passing test designs, did precision analyses from hardware runs that are often overlooked, and filing compliance that addressed the needed implementations. Please let me know if you have more feedback, I'm happy to discuss!
Polarnova commentedon Jun 7, 2026
1ucian0 commentedon Jun 8, 2026
Thank you to everyone who has submitted a PR for this issue! We currently have several promising fixes under review, so please hold off on submitting new PRs until the Qiskit team posts a follow-up comment here.
We will review the existing submissions in due course. Thank you for your patience!
doichanj commentedon Jun 9, 2026
Thank you for submitting PR's for this PR but, I'm very sorry but I missed to add implementation guides in this issue.
I added two instructions:
SparseObservableclassbackend::runfunction for a vector ofEstimatorPubto run with estimator program id on IQPdoichanj commentedon Jun 9, 2026
I'm now reviewing #160 this PR well follows the instruction and work on the actual device
rosspeili commentedon Jun 9, 2026
Thanks for adding the implementation notes Jun Doi (@doichanj) , that helped a lot. I have updated our branch to use
SparseObservableand nativeBackendV2::runforEstimatorPubvia QRMI (program_id = "estimator"). Looked at #160 while refactoring, and kept the same idea but dropped the sampler workaround and trimmed it to what was needed. Pushed on same branch/pr.Issue #145 Implement BackendEstimatorV2 runtime support(with IBM Quan…
1ucian0 commentedon Jun 9, 2026
Hi GoiBasia (@GoiBasia)
Congratulations for getting the bounty! please comment in this issue so I can assign it to you.
GoiBasia commentedon Jun 9, 2026
Thank you so much! I’m really glad I could be part of this and really enjoyed working on this quantum project. Let me know if you need any further assistance.