pyfemtet 0.4.14__tar.gz → 0.4.16__tar.gz

{pyfemtet-0.4.14 → pyfemtet-0.4.16}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pyfemtet
-Version: 0.4.14
+Version: 0.4.16
 Summary: Design parameter optimization using Femtet.
 Home-page: https://github.com/pyfemtet/pyfemtet
 License: BSD-3-Clause

pyfemtet-0.4.16/pyfemtet/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.4.16"

{pyfemtet-0.4.14 → pyfemtet-0.4.16}/pyfemtet/opt/interface/_femtet.py

@@ -619,10 +619,14 @@ class FemtetInterface(FEMInterface):
         self.update_model(parameters)
         self.preprocess(self.Femtet)
         self.solve()
+        self.postprocess(self.Femtet)
 
     def preprocess(self, Femtet):
         pass
 
+    def postprocess(self, Femtet):
+        pass
+
     def quit(self, timeout=1, force=True):
         """Force to terminate connected Femtet."""
         major, minor, bugfix = 2024, 0, 1

{pyfemtet-0.4.14 → pyfemtet-0.4.16}/pyfemtet/opt/prediction/single_task_gp.py

@@ -52,6 +52,9 @@ class SingleTaskGPModel(PredictionModelBase):
         train_x = torch.tensor(x).double()
         train_y = torch.tensor(y).double()
 
+        # check y shape (if single objective problem, output dimension is (n,) )
+        self._is_single_objective = len(y[0]) == 1
+
         # Normalize the input data to the unit cube
         self.scaler_x = MyMinMaxScaler()
         train_x = self.scaler_x.fit_transform(train_x)
@@ -73,10 +76,14 @@ class SingleTaskGPModel(PredictionModelBase):
             scaled_x = self.scaler_x.transform(x)
             # predict
             pred = self.gp(scaled_x)
-            scaled_mean = torch.permute(pred.mean, (1, 0))
-            scaled_var = torch.permute(pred.variance, (1, 0))
-        # unscaling
-        mean = self.scaler_y.inverse_transform_mean(scaled_mean).numpy()
-        var = self.scaler_y.inverse_transform_var(scaled_var).numpy()
+            if self._is_single_objective:
+                scaled_mean = pred.mean.reshape((-1, 1))
+                scaled_var = pred.variance.reshape((-1, 1))
+            else:
+                scaled_mean = torch.permute(pred.mean, (1, 0))
+                scaled_var = torch.permute(pred.variance, (1, 0))
+            # unscaling
+            mean = self.scaler_y.inverse_transform_mean(scaled_mean).numpy()
+            var = self.scaler_y.inverse_transform_var(scaled_var).numpy()
         std = np.sqrt(var)
         return mean, std

{pyfemtet-0.4.14 → pyfemtet-0.4.16}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pyfemtet"
-version = "0.4.14"  # ignored by versioning plugin
+version = "0.4.16"  # ignored by versioning plugin
 description = "Design parameter optimization using Femtet."
 authors = ["kazuma.naito <kazuma.naito@murata.com>"]
 readme = "README.md"

pyfemtet-0.4.14/pyfemtet/__init__.py

@@ -1 +0,0 @@
-__version__ = "0.4.14"