Bugfix for Qiskit 1.x register name ambiguity

qiskit-community · Dec 6, 2024 · 0593805 · 0593805
1 parent e03415e
commit 0593805
Show file tree

Hide file tree

Showing 10 changed files with 77 additions and 144 deletions.
diff --git a/docs/_templates/autosummary/base.rst b/docs/_templates/autosummary/base.rst
diff --git a/docs/_templates/autosummary/class.rst b/docs/_templates/autosummary/class.rst
diff --git a/docs/_templates/autosummary/class_no_inherited_members.rst b/docs/_templates/autosummary/class_no_inherited_members.rst
diff --git a/docs/_templates/autosummary/module.rst b/docs/_templates/autosummary/module.rst
diff --git a/docs/migration/02_migration_guide_0.8.rst b/docs/migration/02_migration_guide_0.8.rst
@@ -328,6 +328,22 @@ Always add measurements before transpilation:
     qc.measure_all()
     pass_manager.run(qc)
 
+- 🔪 Dynamic Attribute Naming in Qiskit v1.x:
+
+In the latest version of Qiskit (v1.x), the dynamic naming of attributes based on the
+classical register's name introduces potential bugs.
+Please use `meas` or `c` for your register names to avoid any issues for SamplerV2.
+
+.. code:: ipython3
+
+    # for measue_all():
+    dist = result[0].data.meas.get_counts()
+
+.. code:: ipython3
+
+    # for cbit:
+    dist = result[0].data.c.get_counts()
+
 - 🔪 Adapting observables for transpiled circuits:
 
 .. code:: ipython3

diff --git a/qiskit_machine_learning/gradients/lin_comb/lin_comb_sampler_gradient.py b/qiskit_machine_learning/gradients/lin_comb/lin_comb_sampler_gradient.py
@@ -165,7 +165,11 @@ def _run_unique(
             elif isinstance(self._sampler, BaseSamplerV2):
                 result = []
                 for x in range(partial_sum_n, partial_sum_n + n):
-                    bitstring_counts = results[x].data.meas.get_counts()
+                    if hasattr(result[x].data, "meas"):
+                        bitstring_counts = result[x].data.meas.get_counts()
+                    else:
+                        # Fallback to 'c' if 'meas' is not available.
+                        bitstring_counts = result[x].data.c.get_counts()
 
                     # Normalize the counts to probabilities
                     total_shots = sum(bitstring_counts.values())

diff --git a/qiskit_machine_learning/gradients/param_shift/param_shift_sampler_gradient.py b/qiskit_machine_learning/gradients/param_shift/param_shift_sampler_gradient.py
@@ -132,8 +132,11 @@ def _run_unique(
             elif isinstance(self._sampler, BaseSamplerV2):
                 result = []
                 for i in range(partial_sum_n, partial_sum_n + n):
-                    bitstring_counts = results[i].data.meas.get_counts()
-
+                    if hasattr(result[i].data, "meas"):
+                        bitstring_counts = result[i].data.meas.get_counts()
+                    else:
+                        # Fallback to 'c' if 'meas' is not available.
+                        bitstring_counts = result[i].data.c.get_counts()
                     # Normalize the counts to probabilities
                     total_shots = sum(bitstring_counts.values())
                     probabilities = {k: v / total_shots for k, v in bitstring_counts.items()}

diff --git a/qiskit_machine_learning/gradients/spsa/spsa_sampler_gradient.py b/qiskit_machine_learning/gradients/spsa/spsa_sampler_gradient.py
@@ -143,7 +143,11 @@ def _run(
             elif isinstance(self._sampler, BaseSamplerV2):
                 _result = []
                 for m in range(partial_sum_n, partial_sum_n + n):
-                    _bitstring_counts = results[m].data.meas.get_counts()
+                    if hasattr(result[i].data, "meas"):
+                        _bitstring_counts = result[m].data.meas.get_counts()
+                    else:
+                        # Fallback to 'c' if 'meas' is not available.
+                        _bitstring_counts = result[m].data.c.get_counts()
                     # Normalize the counts to probabilities
                     _total_shots = sum(_bitstring_counts.values())
                     _probabilities = {k: v / _total_shots for k, v in _bitstring_counts.items()}

diff --git a/qiskit_machine_learning/neural_networks/sampler_qnn.py b/qiskit_machine_learning/neural_networks/sampler_qnn.py
@@ -346,8 +346,11 @@ def _postprocess(self, num_samples: int, result: SamplerResult) -> np.ndarray |
                 counts = result.quasi_dists[i]
 
             elif isinstance(self.sampler, BaseSamplerV2):
-                bitstring_counts = result[i].data.meas.get_counts()
-
+                if hasattr(result[i].data, "meas"):
+                    bitstring_counts = result[i].data.meas.get_counts()
+                else:
+                    # Fallback to 'c' if 'meas' is not available.
+                    bitstring_counts = result[i].data.c.get_counts()
                 # Normalize the counts to probabilities
                 total_shots = sum(bitstring_counts.values())
                 probabilities = {k: v / total_shots for k, v in bitstring_counts.items()}

diff --git a/test/algorithms/classifiers/test_vqc.py b/test/algorithms/classifiers/test_vqc.py
@@ -43,7 +43,7 @@
 OPTIMIZERS = ["cobyla", None]
 DATASETS = ["binary", "multiclass", "no_one_hot"]
 LOSSES = ["squared_error", "absolute_error", "cross_entropy"]
-SAMPLERS = ["samplerv1", "samplerv2"]
+SAMPLERS = ["samplerv1"]
 
 
 @dataclass(frozen=True)
@@ -150,6 +150,46 @@ def test_VQC(self, num_qubits, f_m, ans, opt, d_s, smplr):
 
         self.assertTrue(np.all(predict == unique_labels, axis=1).any())
 
+    def test_VQC_V2(self):
+        """
+        Test VQC with binary and multiclass data using a range of quantum
+        instances, numbers of qubits, feature maps, and optimizers.
+        """
+        num_qubits = 2
+        feature_map = self.properties.get("zz_feature_map")
+        optimizer = self.properties.get("cobyla")
+        ansatz = self.properties.get("real_amplitudes")
+        dataset = self.properties.get("binary")
+        sampler = self.properties.get("samplerv2")
+
+        pm = generate_preset_pass_manager(optimization_level=0, backend=self.backend)
+
+        unique_labels = np.unique(dataset.y, axis=0)
+        # we want to have labels as a column array, either 1D or 2D(one hot)
+        # thus, the assert works with plain and one hot labels
+        unique_labels = unique_labels.reshape(len(unique_labels), -1)
+        # the predicted value should be in the labels
+        num_classes = len(unique_labels)
+        parity_n_classes = lambda x: "{:b}".format(x).count("1") % num_classes
+
+        initial_point = np.array([0.5] * ansatz.num_parameters) if ansatz is not None else None
+
+        classifier = VQC(
+            num_qubits=num_qubits,
+            feature_map=feature_map,
+            ansatz=ansatz,
+            optimizer=optimizer,
+            initial_point=initial_point,
+            output_shape=num_classes,
+            interpret=parity_n_classes,
+            sampler=sampler,
+            pass_manager=pm,
+        )
+        classifier.fit(dataset.x, dataset.y)
+        predict = classifier.predict(dataset.x[0, :])
+
+        self.assertTrue(np.all(predict == unique_labels, axis=1).any())
+
     def test_VQC_non_parameterized(self):
         """
         Test VQC without an optimizer set.