Qiskit session #551
Merged
Qiskit session #551
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
[sc-54393] |
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## new_device_feature_branch #551 +/- ##
=============================================================
Coverage ? 100.00%
=============================================================
Files ? 8
Lines ? 795
Branches ? 0
=============================================================
Hits ? 795
Misses ? 0
Partials ? 0 ☔ View full report in Codecov by Sentry. |
…ssion options on device initialization and when using the session manager. We use the options in the device for things that are generally not updateable for the device e.g. backends; we use session options for everything else
…structor, and verifying that an error is raised due to such behavior
austingmhuang
requested review from
obliviateandsurrender and
lillian542
and removed request for
obliviateandsurrender
austingmhuang
commented
Jun 7, 2024
pennylane_qiskit/qiskit_device2.py
Outdated
Comment on lines
106
to
123
| for k, v in kwargs.items(): | ||
| # Options like service and backend should be tied to the settings set on device | ||
| if k in session_options and k != "max_time": | ||
| warnings.warn(f"Using '{k}' set in device, {device.backend}", UserWarning) | ||
|
|
||
| # Need "_" since `max_time` attribute on Session is `_max_time` | ||
| # When there is overlap between the Session options on the device and the session options | ||
| # passed in via qiskit_session, we prefer the ones passed in via qiskit_session | ||
| elif existing_session and ( | ||
| hasattr(existing_session, "_" + k) or hasattr(existing_session, k) | ||
| ): | ||
| warnings.warn( | ||
| f"`{k}` was set in the Session passed to the device. Using `{k}` '{v}' set in `qiskit_session`.", | ||
| UserWarning, | ||
| ) | ||
| session_options[k] = v | ||
| else: | ||
| session_options[k] = v |
There was a problem hiding this comment.
If Qiskit releases new Session options, this code would still capture those options and raise appropriate warnings if overlapped (unless they make one of the session attributes have extremely weird naming convention)
lillian542
reviewed
Jun 10, 2024
pennylane_qiskit/qiskit_device2.py
Outdated
| # Need "_" since `max_time` attribute on Session is `_max_time` | ||
| # When there is overlap between the Session options on the device and the session options | ||
| # passed in via qiskit_session, we prefer the ones passed in via qiskit_session | ||
| elif existing_session and getattr(existing_session, "_" + k): |
There was a problem hiding this comment.
Do we need to worry about setting a default for the getattr here?
There was a problem hiding this comment.
I don't think so. existing_session, if it exists, should always be some Session() which has all those attributes. I suppose if the User themselves wrote their own version of Session (e.g. our mock) this could be a problem, but I don't see why a User would want to do that...? Then again, if Qiskit significantly changes the attributes of Session or adds an attribute without an "_" in front of it this would be a problem. I guess it wouldn't hurt to set a default 🤷 but not sure how much value it provides.
obliviateandsurrender
approved these changes
Jun 20, 2024
lillian542
reviewed
Jun 20, 2024
pennylane_qiskit/qiskit_device2.py
Outdated
| session = Session(backend=device.backend) | ||
|
|
||
| # When an existing session exists, we want to use its settings unless overwritten | ||
| # by settings in the qiskit_session |
There was a problem hiding this comment.
I still consider the behaviour quite unintuitive. If we want to go for this, can we include information in the docstring about what will happen if the device that is passed already has a session on it?
There was a problem hiding this comment.
No I think it is kind of unintuitive to do it this way as well.
The best way is probably to just make the docstring more clear that if you initialize qiskit_session you will not be using your existing session set in the device.
The logic is that I think the existing session set in device should only have value in a very specific case: batched circuits. In all other cases, the session is just necessary to access the primitives but doesn't necessarily do anything.
Therefore if a user has gone out of their way to use qiskit_session, they probably want to do some kind of like VQE or QAOA or whatever iterative thing, so they probably want those settings. I'll make it clearer that a User doesn't accidentally think they should set the Session in the device.
Co-authored-by: Utkarsh <[email protected]>
Co-authored-by: Utkarsh <[email protected]>
lillian542
approved these changes
Jul 4, 2024
tests/test_base_device.py
Outdated
Comment on lines
283
to
286
| """Test that warnings are raised when there are is an overlap between the Session options | ||
| on the device and the session options passed in via qiskit_session. Also ensures that the | ||
| session options passed in from the qiskit_session take precedence except for `backend` or | ||
| `service`""" |
There was a problem hiding this comment.
docstring could use clarification following updates in the PR
Co-authored-by: lillian542 <[email protected]>
austingmhuang
added a commit
that referenced
this pull request
Jul 25, 2024
* New Qiskit device prototype (#350) * initial prototype device * add use_primitives kwarg * reorganize circuit conversion part 1 * move circuit translation out of device * estimator execution * some small improvements for codefactor * allow circuits with mixed MP types * move translation functions * add support for broadcast_expand and session * add kwargs for options * tidy up options and session * warn if using non-primitive measurements * remove context manager for device session * refactor handling of Options and kwargs * cleaning up * fix 'c register already exists' hardware error * change Options update interface * don't allow options to override shots * don't use classical reg in estimator circuits * update docstring * add conversion tests * add observable conversion test * fix bug in PauliOps converter * add more conversion test functions * remove Adjoint from supported ops and tidy up * tests and little fixes * add tests and black formatting * fix wire order bug for Estimator returns * add tests * remove error if device doesn't initialize * try to get CI to run * more CI stuff * try a thing * black formatting * fix typo * add missing skip-if-no-account * update converter tests * add mockers to allow tests to run in CI * temporarily comment out integration tests * get service from backend * add mock service to mock backend * mock calls to Session in unit testing * uncomment the other tests again * black formatting * newer black formatting * add mocked tests for main execute method * add MockSession to mocked execution tests * pylint * add backend to MockSession calls * mock tests for _execute methods * black formatting * mock transpile for execute_runtime_service * add name to MockedBackend * Apply suggestions from code review Co-authored-by: Matthew Silverman <[email protected]> Co-authored-by: Astral Cai <[email protected]> * Add barrier to ops list * apply suggestion from code review Co-authored-by: Astral Cai <[email protected]> * revert adding barrier for now * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Matthew Silverman <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Matthew Silverman <[email protected]> * Small performance change * Revert * pin qiskit for now * Pin qiskit-ibm-runtime * Move function in init to update_kwargs * Delete print statement lol * fix for shot information * name property change * fixes * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Matthew Silverman <[email protected]> * Edited docstrings * Function signature * Shots are now with context manager * black reformat * black reformat * test changes * tests * fixed conflicts * Change mp_to_pauli to accept mps that affect more than 1 qubit * Revert to fix CI tests * black reformat * Update setup.py Co-authored-by: lillian542 <[email protected]> * Update requirements-ci.txt Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/converter.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/converter.py Co-authored-by: lillian542 <[email protected]> * tests * Fixed a test * maybe this works? * placeholder settings * fixed CI tests, bandaid fix * gitignore for 1.0 qiskit versioning * delete venv1 * tests, shots information, shot vector case, ibm_run_time compatibility * woops * docstring for operation_to_qiskit * black * appease codecov * mock test * Deleted comment * remove some notebooks * Update .gitignore Co-authored-by: Astral Cai <[email protected]> * tests and docstrings --------- Co-authored-by: Matthew Silverman <[email protected]> Co-authored-by: Astral Cai <[email protected]> Co-authored-by: Austin Huang <[email protected]> Co-authored-by: Austin Huang <[email protected]> Co-authored-by: Astral Cai <[email protected]> * Add compile_backend kwarg (#398) Adds the compile_backend kwarg to the Qiskit device. It is useful when you want to do circuit transpilation when using the old Qiskit API (e.g. use_primitives = False). * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> --------- Co-authored-by: Austin Huang <[email protected]> Co-authored-by: Austin Huang <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Support both V1 and V2 (#399) * add name to MockedBackend * support both V1 and V2 syntax for retrieving backend name and num_qubits * test relevant methods with both V1 and V2 MockBackends * tests updated for old device api as per other PR * tests * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Apply suggestions from code review Co-authored-by: Utkarsh <[email protected]> * merge conf --------- Co-authored-by: Austin Huang <[email protected]> Co-authored-by: Austin Huang <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Scalar tests passing * Change mp_to_pauli to be more general * Hamiltonian testing * tests * Revert "tests" This reverts commit 4f21e7ddf354cc36efee55c5d50e4f978c3d2729. * Revert "Hamiltonian testing" This reverts commit 17c3e674235709a7c5e1baf037ebf37b42f3efbf. * Revert "Change mp_to_pauli to be more general" This reverts commit a1ff60638f19d302a802f46a4d586a616e0be95c. * revert * Install PL dev intead of PL stable on CI (#516) (#518) * install PL dev intead of PL stable on CI * trigger ci * also a couple of other places Co-authored-by: lillian542 <[email protected]> * Delete ds store * import change * black * some changes * linters * black * pylinting * black * small change * change conftest * black linter * fix * Generalize mp_to_pauli so that it can work with any observable that has a Pauli Rep (#517) * tests * converter * fixes to tests * refactor * delete print * black * pylint changes * deleted tests/pylintrc * one additional test * black * tests * redo commits * clean up * black * formatting * black * Update pennylane_qiskit/converter.py Co-authored-by: Utkarsh <[email protected]> * rewrite tests to all manual cases * [skip-ci] black * single qubit operations * value error for no pauli rep * list comprehension * removed one list comprehension * ehh honestly this is fine too * delete some useless lines * [skip-ci] refactor * [skip ci] accidentally deleted something...? * Update tests/test_converter.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/converter.py Co-authored-by: lillian542 <[email protected]> * remove parametrize * [skip ci] black * [skip ci] docstring * [skip ci] added assert statement * [skip-ci] linear comb changes * additional more complicated integration tests * undo some formatting * formats * integration tests * tests * deleted unsupported observables * remove sparsehamiltonian for now * rollback pylint * revert observable stopping condition * black * just a commit * black * fm * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * linter errors * linter * linter * linters * try finally block better? * revert * let's just do this... * trailing whitespace * linter * black * formats * black * black * clean up commits * reformatting fixed * clean up commit --------- Co-authored-by: Utkarsh <[email protected]> Co-authored-by: lillian542 <[email protected]> * Use old _execute_runtime_service for observables that are not compatible with SPO representation (#525) * Functionality * syntax error * additional test * additional test cases * edit comment * comment * additional comments * added a warning * added a warning * delete print statement * warning test * delete a test * black * Minor adjustment to sort observables. Modified tests to accomodate. Doc strings edit * Modified tests and warning * edit * fixes * moved test to mocked * mockedbackend update * black * changes * test * uncomment * docstring * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * black * fix --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Move off of "ibmq_qasm_simulator" and towards using AerSimulator for qiskit tests (#521) * import aersimulator * commit * black * added functionality for local simulators when use_primitves = False * [skip ci] changes to AerSim * removal of ibm service * tests * woops * temp fixes * unfinished changes * ignore tests due to version errors * ignore tests due to version errors * black * cleanup * delete comment * revert a change * replaced a test case * shouldn't need to skip if no acc anymore * added comments * delete comment * minor refactor * Added additional comments * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * [skip ci] delete comment * add comment --------- Co-authored-by: Utkarsh <[email protected]> * Update reqs to 1.0 (#536) * removed legacy ci * Delete .github/workflows/tests.yml * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * removed devices * pylint * pylint * docs fix * setup.py changes * delete * revert * reqs change * setup * change to reqs to match ci * removed a test * pylint * put ibmq.rst back * delete ibmq * remove ibmq * deletion * codecov * lint * changes to tests * Revert "Remove devices that will not be supported in the new release" (#544) * changelogs * Remove basicaer (#546) * removed legacy ci * Delete .github/workflows/tests.yml * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * removed devices * pylint * pylint * docs fix * setup.py changes * delete * revert * reqs change * setup * change to reqs to match ci * removed a test * pylint * put ibmq.rst back * delete ibmq * remove ibmq * deletion * codecov * lint * changes to tests * Revert "Remove devices that will not be supported in the new release" (#544) * docs * remove basicaer * reqs to build docs * docs * pylint * tests fix * path change * Changelog and doc * changelogs * Update CHANGELOG.md * deleted a test file * removed error * basic sim pylint * Update CHANGELOG.md Co-authored-by: lillian542 <[email protected]> * remove ifelse block * pylint --------- Co-authored-by: lillian542 <[email protected]> * Remove use primitives and everything that depends on it (#538) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Remove ibmq devices (#550) * Removing ibmq devices from the docs and relevant files * missed something in docs * Changelog updates * Update CHANGELOG.md Co-authored-by: Utkarsh <[email protected]> --------- Co-authored-by: Utkarsh <[email protected]> * Migrate to v2 primitives (#539) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * deleting unused tests * line change * pylint * yay * docstring * refactoring of estimator and sampler * process_estimator_job tests * comment for clarity --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * pylint * Process kwargs (#547) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * We delete the Options Handling class because there are no more Options() to handle. Additionally, process_kwargs is left as a stub as a temporary measure while we figure out what to do with kwargs * changed warnings due to difference in UI for setting shots between Qiskit and Pennylane. Tracking shots has also been updated due to estimatorV2 syntax change * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * due to the fact that shots are not tracked in the estimator's metadata anymore, variances are calculated a different way using the precision instead * black * lint * docstring changes * backend options? * deleting unused tests * line change * [skip ci] changed test to be more readable * New tests for options functionality and edge case * pylint * We make sure that transpilation options are not passed to the primitive and that no errors are raised as a result * Due to changing the signature of get_transpile_args(), we need to fix one of the tests * warning message for default_shots was unclear. changed to be more clear * add more comments * yay * docstring * Testing to ensure that options and kwargs combine properly * edit tests for pylint * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * edit test regex matching due to changes earlier * refactoring of estimator and sampler * generate samples tested * process_estimator_job tests * comment for clarity * pylint * pylint * [skip ci] minor formatting * Fix unintended additonal dimensionality and added test for res != 1 testcase * fix to transpiles * comments to explain some stuff * merge conflicts * docstring * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * clean up * refactor * pylint * formatting of docstring * revert to tuple(res) * [skip ci] fix to dimensions of sampler * docstrings * some docstrings changes * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * black * linter --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Docstrings for converter functions and new qiskit device (#552) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * We delete the Options Handling class because there are no more Options() to handle. Additionally, process_kwargs is left as a stub as a temporary measure while we figure out what to do with kwargs * changed warnings due to difference in UI for setting shots between Qiskit and Pennylane. Tracking shots has also been updated due to estimatorV2 syntax change * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * due to the fact that shots are not tracked in the estimator's metadata anymore, variances are calculated a different way using the precision instead * black * lint * docstring changes * backend options? * deleting unused tests * line change * [skip ci] changed test to be more readable * New tests for options functionality and edge case * pylint * We make sure that transpilation options are not passed to the primitive and that no errors are raised as a result * Due to changing the signature of get_transpile_args(), we need to fix one of the tests * warning message for default_shots was unclear. changed to be more clear * add more comments * yay * docstring * Testing to ensure that options and kwargs combine properly * edit tests for pylint * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * edit test regex matching due to changes earlier * refactoring of estimator and sampler * generate samples tested * process_estimator_job tests * comment for clarity * pylint * pylint * [skip ci] minor formatting * [skip ci] docstrings for converter functions * Fix unintended additonal dimensionality and added test for res != 1 testcase * fix to transpiles * comments to explain some stuff * merge conflicts * [skip ci] examples of QiskitDevice2 added to docstring * docstring changes * docstring * more docstrings * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * clean up * refactor * pylint * formatting of docstring * docstrings * docstrings * formatting * changes * some examples * Update pennylane_qiskit/converter.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * examples and links * change inheritance for remote device * build docs * better docs * fix docs a little * remove redundant docstrings * revert * import fix * build sphinx * revert change to QiskitDev2 * Update pennylane_qiskit/remote.py Co-authored-by: Utkarsh <[email protected]> * reformat to within 100 chars --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Diagonalize gates (#558) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * We delete the Options Handling class because there are no more Options() to handle. Additionally, process_kwargs is left as a stub as a temporary measure while we figure out what to do with kwargs * changed warnings due to difference in UI for setting shots between Qiskit and Pennylane. Tracking shots has also been updated due to estimatorV2 syntax change * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * due to the fact that shots are not tracked in the estimator's metadata anymore, variances are calculated a different way using the precision instead * black * lint * docstring changes * backend options? * deleting unused tests * line change * [skip ci] changed test to be more readable * New tests for options functionality and edge case * pylint * We make sure that transpilation options are not passed to the primitive and that no errors are raised as a result * Due to changing the signature of get_transpile_args(), we need to fix one of the tests * warning message for default_shots was unclear. changed to be more clear * add more comments * yay * docstring * Testing to ensure that options and kwargs combine properly * edit tests for pylint * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * edit test regex matching due to changes earlier * refactoring of estimator and sampler * generate samples tested * process_estimator_job tests * comment for clarity * pylint * pylint * [skip ci] minor formatting * Fix unintended additonal dimensionality and added test for res != 1 testcase * fix to transpiles * comments to explain some stuff * merge conflicts * docstring * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * clean up * refactor * pylint * formatting of docstring * revert to tuple(res) * [skip ci] fix to dimensions of sampler * docstrings * some docstrings changes * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * black * [skip ci] import split_non_commuting * diagonalize tests for Hadamard * changed stopping condition to reflect reality of what's supported and added tests and changed tests to fit new stopping condition * [skip ci] added comment about qml.var not providing matching answers * some tests * [skip ci] linter * interesting changes * linter * split non commuting test cases * sprod * sampler tested as well * linter * black * docstrings and comments * docstrings and comments black * comment regarding magic number * todo * add np * more concise way of testing * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * fix black * diagonalize for edge case * linter * pylint * clean up * fix docstring --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Qiskit session (#551) * functionality implemented * minor adjustments to tests * [skip-ci] Qiskit Sessions now test many warnings since you can set session options on device initialization and when using the session manager. We use the options in the device for things that are generally not updateable for the device e.g. backends; we use session options for everything else * [skip-ci] tests that we are passing on kwargs to Qiskit's session constructor, and verifying that an error is raised due to such behavior * [skip ci] pylint * small comments * Generalization of the session options * delicious docstrings * [skip ci] tests and clarification * [skip ci] better session options * comments for clarity * changes to the tests & the warning message * docstrings * type error changes * docstrings * add qiskit_session to docs * a little more consistency in comments * for docs * fix ci * black * revert * revert * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Qiskit Session update * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * docstring update --------- Co-authored-by: Utkarsh <[email protected]> Co-authored-by: lillian542 <[email protected]> * Tests with fakehardware (#553) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * We delete the Options Handling class because there are no more Options() to handle. Additionally, process_kwargs is left as a stub as a temporary measure while we figure out what to do with kwargs * changed warnings due to difference in UI for setting shots between Qiskit and Pennylane. Tracking shots has also been updated due to estimatorV2 syntax change * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * due to the fact that shots are not tracked in the estimator's metadata anymore, variances are calculated a different way using the precision instead * black * lint * docstring changes * backend options? * deleting unused tests * line change * [skip ci] changed test to be more readable * New tests for options functionality and edge case * pylint * We make sure that transpilation options are not passed to the primitive and that no errors are raised as a result * Due to changing the signature of get_transpile_args(), we need to fix one of the tests * warning message for default_shots was unclear. changed to be more clear * add more comments * yay * docstring * Testing to ensure that options and kwargs combine properly * edit tests for pylint * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * edit test regex matching due to changes earlier * refactoring of estimator and sampler * generate samples tested * process_estimator_job tests * comment for clarity * pylint * pylint * [skip ci] minor formatting * Fix unintended additonal dimensionality and added test for res != 1 testcase * fix to transpiles * comments to explain some stuff * merge conflicts * docstring * [skip ci] some tests with fakehardware * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * clean up * refactor * pylint * formatting of docstring * revert to tuple(res) * [skip ci] fix to dimensions of sampler * docstrings * some docstrings changes * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * black * [skip ci] import split_non_commuting * diagonalize tests for Hadamard * changed stopping condition to reflect reality of what's supported and added tests and changed tests to fit new stopping condition * [skip ci] added comment about qml.var not providing matching answers * some tests * [skip ci] linter * interesting changes * linter * split non commuting test cases * sprod * sampler tested as well * linter * black * docstrings and comments * docstrings and comments black * comment regarding magic number * todo * add np * more concise way of testing * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * fix black * diagonalize for edge case * linter * pylint * clean up * fix docstring * flaky * flaky and fake * rename bakcend to aer_backend --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * fix merge conf * fix setup.py * reqs.txt * changelog changes * Use new qiskit device as the base for remote (#566) * import changes * Added TODOs for tests * changes * this should pass * pylint * circular import * observables update * remerge * import from qiskitdevice2 * Delete unnecessary tests and mocks * black/pylint * Add tests back in for codecov. * refactor tests * delete legacy device only functionality * change around imports * add assertion * fix * fix * fix setup * clean up * fix reqs.txt * fix * changelog changed * maybe this works? * a docstring? * a docstring? * reverts * does this break * revert * fix * fix * attempt a doc fix * Update tests/test_integration.py Co-authored-by: Utkarsh <[email protected]> * some docstrings --------- Co-authored-by: obliviateandsurrender <[email protected]> * Plugin page update (#563) * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * We delete the Options Handling class because there are no more Options() to handle. Additionally, process_kwargs is left as a stub as a temporary measure while we figure out what to do with kwargs * changed warnings due to difference in UI for setting shots between Qiskit and Pennylane. Tracking shots has also been updated due to estimatorV2 syntax change * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * due to the fact that shots are not tracked in the estimator's metadata anymore, variances are calculated a different way using the precision instead * black * lint * docstring changes * backend options? * deleting unused tests * line change * [skip ci] changed test to be more readable * New tests for options functionality and edge case * pylint * We make sure that transpilation options are not passed to the primitive and that no errors are raised as a result * Due to changing the signature of get_transpile_args(), we need to fix one of the tests * warning message for default_shots was unclear. changed to be more clear * add more comments * yay * docstring * Testing to ensure that options and kwargs combine properly * edit tests for pylint * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * edit test regex matching due to changes earlier * refactoring of estimator and sampler * generate samples tested * process_estimator_job tests * comment for clarity * pylint * pylint * [skip ci] minor formatting * [skip ci] docstrings for converter functions * Fix unintended additonal dimensionality and added test for res != 1 testcase * fix to transpiles * comments to explain some stuff * merge conflicts * [skip ci] examples of QiskitDevice2 added to docstring * docstring changes * docstring * more docstrings * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * clean up * refactor * pylint * formatting of docstring * docstrings * docstrings * formatting * changes * some examples * Update pennylane_qiskit/converter.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * examples and links * change inheritance for remote device * build docs * better docs * fix docs a little * remove redundant docstrings * revert * import fix * prelim changes to aer * changes to build sphinx * revert change * delete section on ibmq devices * add examples for remote.rst * plugin updates to the remote device and basicsim * doc fixes for sphinx build * docs * small fi * fix weird spacing * [skip ci] small fix * [skip ci] error in codeblock * delete extra the * [skip ci] changelog * merge confs * formatting * black * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * address comments * undo * readme * plugin page fixes * fixed documentation * fix * Update doc/index.rst Co-authored-by: lillian542 <[email protected]> * doc changes * typo * weird change didn't go through * change to doc * small change * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * Update doc/devices/remote.rst Co-authored-by: Utkarsh <[email protected]> * change name of iqp token * small fix --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Tracker functionality (#533) * is this all? * doc strings * tests for tracker * trackers * trackers * fixed tests * removed legacy ci * Delete .github/workflows/tests.yml * Deleted use_primitives from tests * remove use_primitive kwarg and things that depend on it * fix tests and split_exec * fixed test * pylint * change to v2 prims, del Options * del Options * temp changes to options * naming and 0.46 test * rename * update qiskit device * dep warnings * pylint * changes to tests * deleted options for now * small changes * access sampler results * Sampler tests and functionality * estimator multi measurement works * estimator now gives variances * comments * removed backend.run() and _execute_runtime * remove additional stuff * linter * docstring changes * skip additional test * [skip ci] format is correct, checks probs as well * [skip ci] docstring * docstrings * We delete the Options Handling class because there are no more Options() to handle. Additionally, process_kwargs is left as a stub as a temporary measure while we figure out what to do with kwargs * changed warnings due to difference in UI for setting shots between Qiskit and Pennylane. Tracking shots has also been updated due to estimatorV2 syntax change * un did confusing change that didnt do anything * rerun ci * Syntax changes due to version change * Delete for codecov * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * docstrings * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * finishing touches * comments * [skip ci] formatting * [skip ci] refactor * due to the fact that shots are not tracked in the estimator's metadata anymore, variances are calculated a different way using the precision instead * black * lint * docstring changes * backend options? * deleting unused tests * line change * [skip ci] changed test to be more readable * New tests for options functionality and edge case * pylint * We make sure that transpilation options are not passed to the primitive and that no errors are raised as a result * Due to changing the signature of get_transpile_args(), we need to fix one of the tests * warning message for default_shots was unclear. changed to be more clear * add more comments * yay * docstring * Testing to ensure that options and kwargs combine properly * edit tests for pylint * woops * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * Update pennylane_qiskit/qiskit_device2.py Co-authored-by: Utkarsh <[email protected]> * edit test regex matching due to changes earlier * print out some stuff for tests * refactoring of estimator and sampler * generate samples tested * process_estimator_job tests * comment for clarity * pylint * pylint * [skip ci] temp * [skip ci] There is a bug due to the post processing of results that is causing some of the assertion statements to fail. We can ignore these assertions for now and address how to rework reorder_fn to avoid this bug * [skip ci] * [skip ci] minor formatting * Fix unintended additonal dimensionality and added test for res != 1 testcase * fix to transpiles * comments to explain some stuff * merge conflicts * pylint * formatting * tracker comments * black * comments about the tracker * bet * fix to imports * black * temp * baller implementation * increase shot number to reduce error * edit function * black * better tests * refactor * black * delete print statement * refactor * Delete unneeded import * fix assertion * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * Update tests/test_base_device.py Co-authored-by: lillian542 <[email protected]> * new tests * removed simulations keyword * some xfails pending discussion --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Utkarsh <[email protected]> * Update CHANGELOG.md Co-authored-by: lillian542 <[email protected]> * CHANGELOG --------- Co-authored-by: lillian542 <[email protected]> Co-authored-by: Matthew Silverman <[email protected]> Co-authored-by: Astral Cai <[email protected]> Co-authored-by: Astral Cai <[email protected]> Co-authored-by: Utkarsh <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Context:
Session options like
max_timewere not supported. If the Users wanted to useqiskit_session, they wouldn't be able to set any settings for that session.What this PR does:
This PR allows users to set options that the Session initializer accepts. This is to make this relatively future-proof, where if Qiskit decides that the Session should accept other keyword arguments then we would be able to simply pass that along. We do not allow users to modify two specific settings when using
qiskit_session, namely thebackendand theservice. This is because we force the user to specify a backend when we initialize the device in the first place, by proxy, what service. Basically, it doesn't make sense to let them modify this on session initialization. We do wantmax_timeand potentially other keyword arguments in the future to be able to be modified when initializing theqiskit_sessionthough.Any session that the user set on device initialization is overwritten by this context manager (unless the context manager is trying to do something strange, like change the backend/service). The examples/docstrings try to make this as clear as possible.