Catch edge cases for clustering and plotting

pydmd/costs.py

@@ -704,7 +704,8 @@ def cluster_hyperparameter_sweep(
         """
         if n_components_range is None:
             n_components_range = np.arange(
-                np.max((self.svd_rank // 4, 2)), self.svd_rank
+                np.max((self.svd_rank // 4, 2)),
+                self.svd_rank // 2,
             )
         score = np.zeros_like(n_components_range, float)
 
@@ -1359,7 +1360,9 @@ def to_xarray(self):
             },
             attrs={
                 "svd_rank": self.svd_rank,
-                "omega_transformation": self._transform_method,
+                "omega_transformation": self._xarray_sanitize(
+                    self._transform_method
+                ),
                 "n_slides": self._n_slides,
                 "window_length": self._window_length,
                 "num_frequency_bands": self.n_components,
@@ -1411,6 +1414,8 @@ def from_xarray(self, ds):
                 self._pydmd_kwargs[new_attr_name] = self._xarray_unsanitize(
                     ds.attrs[attr]
                 )
+            elif "omega_transformation" in attr:
+                self._transform_method = self._xarray_unsanitize(ds.attrs[attr])
 
         return self
 

pydmd/mrcosts.py

@@ -526,6 +526,9 @@ def plot_local_reconstructions(
             elif level > 0:
                 x_iter, _ = self.costs_array[level - 1].scale_separation()
 
+        if not x_iter.shape == (self._n_data_vars, self._n_time_steps):
+            raise ValueError("Input data has the wrong shape.")
+
         if kwargs is None:
             kwargs = {}