Deprecate qiskit-ibm-provider in favor of qiskit-ibm-runtime (#1371)

### Summary Closes #1346 by adding logic to handle the HGP and service for a `qiskit-ibm-runtime` backend object and replacing mentions of `qiskit-ibm-provider` in docs with `qiskit-ibm-runtime` syntax. ### Details and comments - Updates sessions howto with the new sessions syntax supported by `qiskit-ibm-runtime` (requires >=15.0) - Removes old code to handle `qiskit-ibmq-runtime` backends - Replaces non-free/deprecated backends in docs with `ibm-osaka` --------- Co-authored-by: Will Shanks <[email protected]>
qiskit-community · Jan 25, 2024 · 4ea9cf6 · 4ea9cf6
1 parent e648bc5
commit 4ea9cf6
Show file tree

Hide file tree

Showing 8 changed files with 95 additions and 61 deletions.
diff --git a/docs/conf.py b/docs/conf.py
@@ -171,7 +171,6 @@
     "matplotlib": ("https://matplotlib.org/stable/", None),
     "qiskit": ("https://qiskit.org/documentation/", None),
     "uncertainties": ("https://pythonhosted.org/uncertainties", None),
-    "qiskit_ibm_provider": ("https://qiskit.org/ecosystem/ibm-provider/", None),
     "qiskit_aer": ("https://qiskit.org/ecosystem/aer", None),
     "qiskit_dynamics": ("https://qiskit.org/documentation/dynamics", None),
     "qiskit_ibm_runtime": ("https://qiskit.org/ecosystem/ibm-runtime/", None),

diff --git a/docs/howtos/cloud_service.rst b/docs/howtos/cloud_service.rst
@@ -18,21 +18,22 @@ Saving
 ~~~~~~
 
 .. note::
-    This guide requires :mod:`qiskit-ibm-provider`. For how to migrate from the deprecated :mod:`qiskit-ibmq-provider` to :mod:`qiskit-ibm-provider`,
-    consult the `migration guide <https://qiskit.org/ecosystem/ibm-provider/tutorials/Migration_Guide_from_qiskit-ibmq-provider.html>`_.\
+    This guide requires :mod:`qiskit-ibm-runtime` version 0.15 and up, which can be installed with ``python -m pip install qiskit-ibm-runtime``.
+    For how to migrate from the older :mod:`qiskit-ibm-provider` to :mod:`qiskit-ibm-runtime`,
+    consult the `migration guide <https://docs.quantum.ibm.com/api/migration-guides/qiskit-runtime-from-provider>`_.\
 
 You must run the experiment on a real IBM
 backend and not a simulator to be able to save the experiment data. This is done by calling
 :meth:`~.ExperimentData.save`:
 
 .. jupyter-input::
 
-    from qiskit_ibm_provider import IBMProvider
+    from qiskit_ibm_runtime import QiskitRuntimeService
     from qiskit_experiments.library.characterization import T1
     import numpy as np
 
-    provider = IBMProvider()
-    backend = provider.get_backend("ibmq_lima")
+    service = QiskitRuntimeService(channel="ibm_quantum")
+    backend = service.backend("ibm_osaka")
 
     t1_delays = np.arange(1e-6, 600e-6, 50e-6)
 
@@ -142,7 +143,7 @@ The :meth:`~.ExperimentData.auto_save` feature automatically saves changes to th
 .. jupyter-output::
 
     You can view the experiment online at https://quantum.ibm.com/experiments/cdaff3fa-f621-4915-a4d8-812d05d9a9ca
-    <ExperimentData[T1], backend: ibmq_lima, status: ExperimentStatus.DONE, experiment_id: cdaff3fa-f621-4915-a4d8-812d05d9a9ca>
+    <ExperimentData[T1], backend: ibm_osaka, status: ExperimentStatus.DONE, experiment_id: cdaff3fa-f621-4915-a4d8-812d05d9a9ca>
 
 Setting ``auto_save = True`` works by triggering :meth:`.ExperimentData.save`.
 

diff --git a/docs/howtos/rerun_analysis.rst b/docs/howtos/rerun_analysis.rst
@@ -12,9 +12,9 @@ Solution
 --------
 
 .. note::
-    Some of this guide uses the :mod:`qiskit-ibm-provider` package. For how to migrate from 
-    the deprecated ``qiskit-ibmq-provider`` to ``qiskit-ibm-provider``, consult the
-    `migration guide <https://qiskit.org/ecosystem/ibm-provider/tutorials/Migration_Guide_from_qiskit-ibmq-provider.html>`_.\
+    This guide requires :mod:`qiskit-ibm-runtime` version 0.15 and up, which can be installed with ``python -m pip install qiskit-ibm-runtime``.
+    For how to migrate from the older :mod:`qiskit-ibm-provider` to :mod:`qiskit-ibm-runtime`,
+    consult the `migration guide <https://docs.quantum.ibm.com/api/migration-guides/qiskit-runtime-from-provider>`_.\
 
 Once you recreate the exact experiment you ran and all of its parameters and options,
 you can call the :meth:`.add_jobs` method with a list of :class:`Job

diff --git a/docs/howtos/runtime_sessions.rst b/docs/howtos/runtime_sessions.rst
@@ -10,32 +10,35 @@ You want to run experiments in a `Runtime session
 Solution
 --------
 
-Use the :class:`~qiskit_ibm_provider.IBMBackend` object in ``qiskit-ibm-provider``, which supports sessions.
+.. note::
+    This guide requires :mod:`qiskit-ibm-runtime` version 0.15 and up, which can be installed with ``python -m pip install qiskit-ibm-runtime``.
+    For how to migrate from the older :mod:`qiskit-ibm-provider` to :mod:`qiskit-ibm-runtime`,
+    consult the `migration guide <https://docs.quantum.ibm.com/api/migration-guides/qiskit-runtime-from-provider>`_.\
+
+Use the :class:`~qiskit_ibm_runtime.IBMBackend` object in :mod:`qiskit-ibm-runtime`, which supports sessions.
 
 In this example, we will set the ``max_circuits`` property to an artificially low value so that the experiment will be
 split into multiple jobs that run sequentially in a single session. When running real experiments with a
 large number of circuits that can't fit in a single job, it may be helpful to follow this usage pattern:
 
 .. jupyter-input::
 
-    from qiskit_ibm_provider import IBMProvider
+    from qiskit_ibm_runtime import QiskitRuntimeService
     from qiskit_experiments.library.tomography import ProcessTomography
     from qiskit import QuantumCircuit
 
-    provider = IBMProvider()
-    backend = provider.get_backend("ibm_nairobi")
+    service = QiskitRuntimeService(channel="ibm_quantum")
+    backend = service.backend("ibm_osaka")
     qc = QuantumCircuit(1)
     qc.x(0)
 
-    with backend.open_session() as session:
-        exp = ProcessTomography(qc)
-        exp.set_experiment_options(max_circuits=3)
-        exp_data = exp.run(backend)
-        exp_data.block_for_results()
-        # Calling cancel because session.close() is not available for qiskit-ibm-provider<=0.7.2.
-        # It is safe to call cancel since block_for_results() ensures there are no outstanding jobs 
-        # still running that would be canceled.
-        session.cancel()
+    backend.open_session()
+    exp = ProcessTomography(qc)
+    # Artificially lower circuits per job, adjust value for your own application
+    exp.set_experiment_options(max_circuits=3)
+    exp_data = exp.run(backend)
+    # This will prevent further jobs from being submitted without terminating current jobs
+    backend.close_session()
 
 Note that runtime primitives are not currently supported natively in Qiskit Experiments, so  
 the ``backend.run()`` path is required to run experiments.
diff --git a/qiskit_experiments/framework/experiment_data.py b/qiskit_experiments/framework/experiment_data.py
@@ -32,6 +32,7 @@
 import sys
 import json
 import traceback
+import warnings
 import numpy as np
 import pandas as pd
 from dateutil import tz
@@ -652,16 +653,15 @@ def _set_backend(self, new_backend: Backend, recursive: bool = True) -> None:
         provider = self._backend_data.provider
         if provider is not None:
             self._set_hgp_from_provider(provider)
+        # qiskit-ibm-runtime style
+        elif hasattr(self._backend, "_instance"):
+            self.hgp = self._backend._instance
         if recursive:
             for data in self.child_data():
                 data._set_backend(new_backend)
 
     def _set_hgp_from_provider(self, provider):
         try:
-            # qiskit-ibmq-provider style
-            if hasattr(provider, "credentials"):
-                creds = provider.credentials
-                self.hgp = f"{creds.hub}/{creds.group}/{creds.project}"
             # qiskit-ibm-provider style
             if hasattr(provider, "_hgps"):
                 for hgp_string, hgp in provider._hgps.items():
@@ -2528,21 +2528,31 @@ def __getstate__(self):
     @staticmethod
     def get_service_from_backend(backend):
         """Initializes the service from the backend data"""
-        return ExperimentData.get_service_from_provider(backend.provider)
+        # qiskit-ibm-runtime style
+        try:
+            if hasattr(backend, "service"):
+                token = backend.service._account.token
+                return IBMExperimentService(token=token, url=backend.service._account.url)
+            return ExperimentData.get_service_from_provider(backend.provider)
+        except Exception:  # pylint: disable=broad-except
+            return None
 
     @staticmethod
     def get_service_from_provider(provider):
         """Initializes the service from the provider data"""
-        db_url = "https://auth.quantum.ibm.com/api"
         try:
-            # qiskit-ibmq-provider style
-            if hasattr(provider, "credentials"):
-                token = provider.credentials.token
             # qiskit-ibm-provider style
             if hasattr(provider, "_account"):
-                token = provider._account.token
-            service = IBMExperimentService(token=token, url=db_url)
-            return service
+                warnings.warn(
+                    "qiskit-ibm-provider has been deprecated in favor of qiskit-ibm-runtime. Support"
+                    "for qiskit-ibm-provider backends will be removed in Qiskit Experiments 0.6.",
+                    DeprecationWarning,
+                    stacklevel=2,
+                )
+                return IBMExperimentService(
+                    token=provider._account.token, url=provider._account.url
+                )
+            return None
         except Exception:  # pylint: disable=broad-except
             return None
 

diff --git a/releasenotes/notes/runtime-provider-support-5358b72ec0035419.yaml b/releasenotes/notes/runtime-provider-support-5358b72ec0035419.yaml
@@ -0,0 +1,11 @@
+---
+features:
+  - |
+    Experiments run via the ``qiskit-ibm-runtime`` provider can now be saved
+    to the cloud service.
+upgrade:
+  - |
+    With the impending deprecation of the ``qiskit-ibm-provider`` package, support for
+    ``qiskit-ibm-provider`` is now deprecated and will be removed
+    in the next release. Users should migrate to ``qiskit-ibm-runtime`` following the
+    `migration guide <https://docs.quantum.ibm.com/api/migration-guides/qiskit-runtime-from-provider>`.
diff --git a/requirements-extras.txt b/requirements-extras.txt
@@ -1,4 +1,3 @@
-qiskit-ibm-provider>=0.6.1 # for submitting experiments to backends through the IBM provider
 cvxpy>=1.3.2 # for tomography
 scikit-learn # for discriminators
 qiskit-aer>=0.13.2