[WIP] Reviving optimizing nuggets

Author: Leguark

gempy_engine/API/interp_single/_interp_single_feature.py

@@ -24,8 +24,7 @@ def interpolate_feature(interpolation_input: InterpolationInput,
                         solver_input: SolverInput,
                         external_interp_funct: Optional[CustomInterpolationFunctions] = None,
                         external_segment_funct: Optional[Callable[[np.ndarray], float]] = None,
-                        stack_number: Optional[int] = None,
-                        clean_buffer: bool = True) -> ScalarFieldOutput:
+                        stack_number: Optional[int] = None) -> ScalarFieldOutput:
 
     if BackendTensor.engine_backend is not AvailableBackends.PYTORCH and NOT_MAKE_INPUT_DEEP_COPY is False:
         grid = copy.deepcopy(interpolation_input.grid)

gempy_engine/core/data/internal_structs.py

@@ -22,8 +22,10 @@ def __init__(self, sp_internal: SurfacePointsInternals, ori_internal: Orientatio
                  xyz_to_interpolate: np.ndarray=None, fault_internal=None):
         self.sp_internal = sp_internal
         self.ori_internal = ori_internal
-        if xyz_to_interpolate is not None:
+        if xyz_to_interpolate is not None and xyz_to_interpolate.dtype != BackendTensor.dtype_obj:
             self.xyz_to_interpolate = xyz_to_interpolate.astype(BackendTensor.dtype)
+        else:
+            self.xyz_to_interpolate = xyz_to_interpolate
         self._fault_internal = fault_internal
 
     def __hash__(self):