Fix startpoint sampling for hierarchical optimization

pypesto/petab/importer.py

@@ -631,20 +631,37 @@ def create_prior(self) -> Union[NegLogParameterPriors, None]:
         else:
             return None
 
-    def create_startpoint_method(self, **kwargs) -> StartpointMethod:
+    def create_startpoint_method(
+        self, x_ids: Sequence[str] = None, **kwargs
+    ) -> StartpointMethod:
         """Create a startpoint method.
 
         Parameters
         ----------
+        x_ids:
+            If provided, create a startpoint method that only samples the
+            parameters with the given IDs.
         **kwargs:
             Additional keyword arguments passed on to
             :meth:`pypesto.startpoint.FunctionStartpoints.__init__`.
         """
 
         def startpoint_method(n_starts: int, **kwargs):
-            return petab.sample_parameter_startpoints(
+            startpoints = petab.sample_parameter_startpoints(
                 self.petab_problem.parameter_df, n_starts=n_starts
             )
+            if x_ids is None:
+                return startpoints
+
+            # subset parameters according to the provided parameter IDs
+            from petab.C import ESTIMATE
+
+            parameter_df = self.petab_problem.parameter_df
+            pars_to_estimate = list(
+                parameter_df.index[parameter_df[ESTIMATE] == 1]
+            )
+            x_idxs = [pars_to_estimate.index(x_id) for x_id in x_ids]
+            return startpoints[:, x_idxs]
 
         return FunctionStartpoints(function=startpoint_method, **kwargs)
 
@@ -740,7 +757,9 @@ def create_problem(
             x_names=x_ids,
             x_scales=x_scales,
             x_priors_defs=prior,
-            startpoint_method=self.create_startpoint_method(),
+            startpoint_method=self.create_startpoint_method(
+                x_ids=np.delete(x_ids, x_fixed_indices)
+            ),
             **problem_kwargs,
         )