[CQT-421] Implement QGymMapper mapper pass #642

rares1609 · 2025-11-12T14:31:25Z

No description provided.

…pyproject.toml

juanboschero

Good stuff

juanboschero · 2025-11-21T09:29:11Z

docs/compilation-passes/mapping/qgym-mapper.md

@@ -0,0 +1,69 @@
+The [Qgym](https://github.com/QuTech-Delft/qgym) package functions in a manner similar 
+to the well known gym package, in the  sense that it provides a number of environments 


Suggested change

to the well known gym package, in the sense that it provides a number of environments

to the well known gym package, in the sense that it provides a number of environments

juanboschero · 2025-11-21T09:30:10Z

docs/compilation-passes/mapping/qgym-mapper.md

+well as a `TRPO.zip` file containing the weights of a trained agent in your working 
+directory. 
+
+```python


The example below should be in the test_docs file

juanboschero · 2025-11-21T09:31:05Z

opensquirrel/passes/mapper/qgym_mapper.py

+        return interaction_graph
+
+    def _get_mapping(self, mapping_data: Any, qubit_register_size: int) -> Mapping:
+        """Convert QGym's physical-to-logical mapping to OpenSquirrel's logical-to-physical mapping."""


Missing args and return in docstring

juanboschero · 2025-11-21T09:31:39Z

opensquirrel/passes/mapper/qgym_mapper.py

+        mapping_data = self._extract_mapping_data(last_obs)
+        return self._get_mapping(mapping_data, qubit_register_size)
+
+    def _ir_to_interaction_graph(self, ir: IR) -> nx.Graph:


docstring args + return

elenbaasc

Looks good, thanks for the effort 😄!

Some small refactors required, but no major changes of functionality (I think)

elenbaasc · 2025-11-21T10:21:26Z

docs/compilation-passes/mapping/index.md

+- [Hardcoded Mapper](hardcoded-mapper.md) (`HardcodedMapper`)
+- [Identity Mapper](identity-mapper.md) (`IdentitiyMapper`)
+- [Random Mapper](random-mapper.md) (`RandomMapper`)
+- [Qgym Mapper](qgym-mapper.md) (`QGymMapper`)


Is it Qgym or QGym? We should be consistent. If it is Qgym, then we should also call the pass the QgymMapper, or leave it as is if it is QGym, but then adjust the text.

elenbaasc · 2025-11-21T10:22:13Z

docs/compilation-passes/mapping/qgym-mapper.md

@@ -0,0 +1,69 @@
+The [Qgym](https://github.com/QuTech-Delft/qgym) package functions in a manner similar 
+to the well known gym package, in the  sense that it provides a number of environments 


Suggested change

to the well known gym package, in the sense that it provides a number of environments

to the well-known gym package, in the sense that it provides a number of environments

elenbaasc · 2025-11-21T10:22:45Z

docs/compilation-passes/mapping/qgym-mapper.md

@@ -0,0 +1,69 @@
+The [Qgym](https://github.com/QuTech-Delft/qgym) package functions in a manner similar 
+to the well known gym package, in the  sense that it provides a number of environments 
+on which reinforcement learning (RL) agents can be applied. The main purpose of qgym is 


New sentences are always on new lines. (change here and in other places)

Suggested change

on which reinforcement learning (RL) agents can be applied. The main purpose of qgym is

on which reinforcement learning (RL) agents can be applied.

The main purpose of qgym is

elenbaasc · 2025-11-21T12:45:25Z

docs/compilation-passes/mapping/qgym-mapper.md

+with open('connectivities.json', 'r') as file:
+    connectivities = json.load(file)


Does this needs to be a JSON, can you not simply provide a connectivity (dict[str, list[int]]) as we defined it for example for the routers?

elenbaasc · 2025-11-21T13:28:01Z

docs/compilation-passes/mapping/qgym-mapper.md

+connection_graph = nx.Graph()
+connection_graph.add_edges_from(hardware_connectivity)
+
+qgym_mapper = QGymMapper(agent_class = "TRPO", agent_path = "TRPO.zip", 
+                        connection_graph=connection_graph)


I would rather pass the connectivity and then define what a connection_graph is in in the QgymMapper.

elenbaasc · 2025-11-26T15:08:14Z

tests/passes/mapper/test_qgym_mapper.py

+    assert str(circuit1) != initial_circuit
+
+
+def test_check_not_many_logical_as_physical_qubits(mapper1: QGymMapper, circuit2: Circuit) -> None:


Suggested change

def test_check_not_many_logical_as_physical_qubits(mapper1: QGymMapper, circuit2: Circuit) -> None:

def test_unequal_number_logical_and_physical_qubits(mapper1: QGymMapper, circuit2: Circuit) -> None:

elenbaasc · 2025-11-26T15:08:49Z

tests/passes/mapper/test_qgym_mapper.py

+
+@pytest.mark.parametrize(
+    "mapper, circuit, expected_mapping_length",  # noqa: PT006
+    [("mapper1", "circuit1", 5), ("mapper2", "circuit2", 7)],


Suggested change

[("mapper1", "circuit1", 5), ("mapper2", "circuit2", 7)],

[

("mapper1", "circuit1", 5),

("mapper2", "circuit2", 7)

],

elenbaasc · 2025-11-26T15:10:23Z

tests/passes/mapper/test_qgym_mapper.py

+@pytest.mark.parametrize(
+    "mapper, circuit, expected_mapping_length",  # noqa: PT006
+    [("mapper1", "circuit1", 5), ("mapper2", "circuit2", 7)],
+)


Suggested change

)

ids=["tuna-5-mapping", "starmon-7-mapping"],

)

elenbaasc · 2025-11-26T15:11:39Z

tests/passes/mapper/test_qgym_mapper.py

+    mapper = request.getfixturevalue(mapper)  # type: ignore[arg-type]
+    mapping = mapper.map(circuit.ir, circuit.qubit_register_size)
+    assert isinstance(mapping, Mapping)
+    assert len(mapping) == expected_mapping_length


Can we test more than just the length? Or will the resultant mapping be different each time?

elenbaasc · 2025-11-26T15:26:53Z

opensquirrel/passes/mapper/TRPO_starmon7_5e5.zip

We should put these (*.zips) and the qgym_mapper.py module in a separate folder qgym_mapper.

Rares Oancea added 4 commits November 12, 2025 15:30

[CQT-421]Implement QGymMapper mapper pass

c8f7a71

Resolve merge issues

c07d5fd

Fix typing issue

796ff55

Fix linting issue

24e83da

elenbaasc changed the title ~~[CQT-421]Implement QGymMapper mapper pass~~ [CQT-421] Implement QGymMapper mapper pass Nov 13, 2025

Rares Oancea added 6 commits November 15, 2025 14:50

downgrade qgym, fix action space handling, retrain agents and update …

39b1da4

…pyproject.toml

remove unused type ignore

8aba2c4

format qgym_mapper.py

214e2c1

update __init__.py

6b7d2b5

update __init__.py again

3f376b2

skip QGymMapper tests if qgym-mapper dependency group is not installed

2ef82ed

rares1609 requested a review from elenbaasc November 15, 2025 14:26

juanboschero reviewed Nov 21, 2025

View reviewed changes

elenbaasc reviewed Nov 26, 2025

View reviewed changes

		@@ -0,0 +1,69 @@
		The [Qgym](https://github.com/QuTech-Delft/qgym) package functions in a manner similar
		to the well known gym package, in the sense that it provides a number of environments

	to the well known gym package, in the sense that it provides a number of environments
	to the well-known gym package, in the sense that it provides a number of environments

	on which reinforcement learning (RL) agents can be applied. The main purpose of qgym is
	on which reinforcement learning (RL) agents can be applied.
	The main purpose of qgym is

		with open('connectivities.json', 'r') as file:
		connectivities = json.load(file)

		assert str(circuit1) != initial_circuit


		def test_check_not_many_logical_as_physical_qubits(mapper1: QGymMapper, circuit2: Circuit) -> None:

	def test_check_not_many_logical_as_physical_qubits(mapper1: QGymMapper, circuit2: Circuit) -> None:
	def test_unequal_number_logical_and_physical_qubits(mapper1: QGymMapper, circuit2: Circuit) -> None:

[CQT-421] Implement QGymMapper mapper pass #642

Are you sure you want to change the base?

[CQT-421] Implement QGymMapper mapper pass #642

Uh oh!

Conversation

rares1609 commented Nov 12, 2025

Uh oh!

juanboschero left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

elenbaasc left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants