Merge pull request #174 from UCL-CCS/change_default_sqp

Changed target_sqp in QubitTapering from X to Z

symmer/projection/__init__.py

@@ -1,6 +1,6 @@
 """init for projection."""
 from .utils import *
-from .base import S3_projection
+from .base import S3Projection
 from .qubit_tapering import QubitTapering
 from .contextual_subspace import ContextualSubspace
 from .qubit_subspace_manager import QubitSubspaceManager

symmer/projection/base.py

@@ -4,7 +4,7 @@
 from symmer.evolution import trotter, Had
 from functools import reduce
 
-class S3_projection:
+class S3Projection:
     """ 
     Base class for enabling qubit reduction techniques derived from
     the Stabilizer SubSpace (S3) projection framework, such as tapering

symmer/projection/contextual_subspace.py

@@ -4,11 +4,11 @@
     update_eigenvalues, StabilizerIdentification, ObservableBiasing, stabilizer_walk,
     # get_noncon_generators_from_commuting_stabilizers
 )
-from symmer.projection import S3_projection
+from symmer.projection import S3Projection
 from symmer.evolution import trotter
 from typing import List, Union, Optional
 
-class ContextualSubspace(S3_projection):
+class ContextualSubspace(S3Projection):
     """ 
     Class for performing contextual subspace methods as per https://quantum-journal.org/papers/q-2021-05-14-456/.
     Reduces the number of qubits in the problem while aiming to control the systematic error incurred along the way.
@@ -21,11 +21,13 @@ class ContextualSubspace(S3_projection):
     NOTE: the order in which (1) and (2) are performed depends on the noncontextual strategy specified
     3. Apply unitary partitioning (either sequence of rotations or linear combination of unitaries) to collapse noncontextual cliques
     
-    The remaining steps are handled by the parent S3_projection class:
+    The remaining steps are handled by the parent S3Projection class:
     4. rotate each stabilizer onto a single-qubit Pauli operator, 
     5. drop the corresponding qubits from the Hamiltonian whilst
     6. fixing the +/-1 eigenvalues
     """
+    name = 'contextual_subspace' # for reference in QubitSubspaceManager
+
     def __init__(self,
             operator: PauliwordOp,
             noncontextual_strategy: str = 'diag',
@@ -320,7 +322,7 @@ def project_onto_subspace(self, operator_to_project:PauliwordOp=None) -> Pauliwo
         # if there are no stabilizers, return the input operator
         if self.stabilizers is None:
             return operator_to_project   
-        # instantiate the parent S3_projection class that handles the subspace projection
+        # instantiate the parent S3Projection class that handles the subspace projection
         super().__init__(self.stabilizers)
         self.S3_initialized = True
         # perform unitary partitioning

symmer/projection/qubit_tapering.py

@@ -3,10 +3,10 @@
 import numpy as np
 from typing import List, Union
 from cached_property import cached_property
-from symmer.projection import S3_projection
+from symmer.projection import S3Projection
 from symmer.operators import PauliwordOp, IndependentOp, QuantumState
 
-class QubitTapering(S3_projection):
+class QubitTapering(S3Projection):
     """ 
     Class for performing qubit tapering as per https://arxiv.org/abs/1701.08213.
     Reduces the number of qubits in the problem whilst preserving its energy spectrum by:
@@ -17,11 +17,13 @@ class QubitTapering(S3_projection):
     4. dropping the corresponding qubits from the Hamiltonian whilst
     5. fixing the +/-1 eigenvalues
 
-    Steps 1-2 are handled in this class whereas we defer to the parent S3_projection for 3-5.
+    Steps 1-2 are handled in this class whereas we defer to the parent S3Projection for 3-5.
     """
+    name = 'qubit_tapering' # for reference in QubitSubspaceManager
+
     def __init__(self,
             operator: PauliwordOp, 
-            target_sqp: str = 'X'
+            target_sqp: str = 'Z'
         ) -> None:
         """ 
         Input the PauliwordOp we wish to taper.
@@ -57,7 +59,7 @@ def taper_it(self,
         """ 
         Finally, once the symmetry generators and sector have been identified, 
         we may perform a projection onto the corresponding stabilizer subspace via 
-        the parent S3_projection class.
+        the parent S3Projection class.
 
         This method allows one to input an auxiliary operator other than the internal
         operator itself to be tapered consistently with the identified symmetry. This is 
@@ -90,7 +92,7 @@ def taper_it(self,
         else:
             operator_to_taper = self.operator.copy()
 
-        # taper the operator via S3_projection.perform_projection
+        # taper the operator via S3Projection.perform_projection
         tapered_operator = self.perform_projection(
             operator=operator_to_taper,
             ref_state=ref_state,

tests/test_projection/test_contextual_subspace.py

@@ -150,7 +150,7 @@ def test_project_state_onto_subspace(up_method):
     CS.update_stabilizers(3, aux_operator=CC_taper, strategy='aux_preserving')
     CS.project_onto_subspace()
     projected_state = CS.project_state_onto_subspace(QT.tapered_ref_state)
-    assert projected_state == QuantumState([[0,0,0]], [1])
+    assert projected_state == QuantumState([[0,0,0]], [-1])
 
 def test_project_state_onto_subspace_before_operator():
     CS = ContextualSubspace(H_taper, noncontextual_strategy='StabilizeFirst')