raw_confidence_score property for predictions

src/ansys/simai/core/data/predictions.py

@@ -86,13 +86,25 @@ def post(self) -> PredictionPostProcessings:
         return self._post_processings
 
     @property
-    def confidence_score(self) -> str:
+    def confidence_score(self) -> Optional[str]:
         """Confidence score, which is either ``high`` or ``low``.
 
         This method blocks until the confidence score is computed.
         """
         self.wait()
-        return self.fields["confidence_score"]
+        confidence_score = self.fields["confidence_score"]
+        if confidence_score not in ["high", "low", None]:
+            raise ValueError("Must be None or one of: 'high', 'low', None.")
+        return confidence_score
+
+    @property
+    def raw_confidence_score(self) -> Optional[float]:
+        """Raw confidence score, a float.
+
+        This method blocks until the confidence score is computed.
+        """
+        self.wait()
+        return self.fields["raw_confidence_score"]
 
     def delete(self) -> None:
         """Remove a prediction from the server."""

tests/conftest.py

@@ -117,6 +117,8 @@ def _factory(post_processings=None, geometry=None, **kwargs) -> Prediction:
             kwargs["geometry_id"] = geometry.id
         kwargs.setdefault("boundary_conditions", {"Vx": 10.01, "Vy": 0.0009})
         kwargs.setdefault("state", "successful")
+        kwargs.setdefault("confidence_score", None)
+        kwargs.setdefault("raw_confidence_score", None)
         prediction = simai_client._prediction_directory._model_from(kwargs)
         if post_processings is not None:
             # If we passed post-processings as parameter,

tests/test_core_sse.py

@@ -135,15 +135,15 @@ def test_sse_event_prediction_success(sse_mixin, prediction_factory):
     assert pred.is_pending
     # Mock a SSE success event
     updated_record = pred.fields.copy()
-    updated_record.update({"state": "successful", "confidence_score": "abysmal"})
+    updated_record.update({"state": "successful", "confidence_score": "high"})
     sse_mixin._handle_sse_event(
         create_sse_event(
             f'{{"target": {{"id": "{pred.id}", "type": "prediction"}}, "type": "job", "record": {json.dumps(updated_record)}}}'
         )
     )
     assert not pred.is_pending
     assert pred.is_ready
-    assert pred.confidence_score == "abysmal"
+    assert pred.confidence_score == "high"
 
 
 def test_sse_event_update_prediction_failure(sse_mixin, prediction_factory):

tests/test_predictions.py

@@ -20,6 +20,7 @@
 # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 # SOFTWARE.
 
+import pytest
 import responses
 
 from ansys.simai.core.data.geometries import Geometry
@@ -190,3 +191,21 @@ def test_run_no_bc(simai_client, geometry_factory):
     )
     geometry = geometry_factory(id="geom-0")
     simai_client.predictions.run(geometry.id)
+
+
+@responses.activate
+def test_confidence_score(prediction_factory):
+    """WHEN accessing a Prediction's confidence score properties
+    THEN the corresponding values are returned
+    """
+    prediction = prediction_factory(confidence_score="high", raw_confidence_score=0.94107)
+    empty_prediction = prediction_factory()
+    bad_prediction = prediction_factory(confidence_score="abysmal")
+
+    assert prediction.confidence_score == "high"
+    assert prediction.raw_confidence_score == 0.94107
+    assert empty_prediction.confidence_score is None
+    assert empty_prediction.raw_confidence_score is None
+    with pytest.raises(ValueError) as exc:
+        assert bad_prediction.confidence_score
+    assert str(exc.value) == "Must be None or one of: 'high', 'low', None."