ValueError: Found array with 0 sample(s) (shape=(0, 2)) while a minimum of 1 is required by KNeighborsClassifier.

[kdberends]

Describe the bug
After converting 2.3.0 input to 2.4.0 (apparently correctly), user runs into following error:

Traceback (most recent call last):
 File "p:\11211534-008-25mad09a-ovd\C_Work\573_sobek-ovd-j24_6-v2a1\731_FM2Prof\779_Run_FM2Prof\fm2prof_CSS.py", line 17, in <module>
  project.run(overwrite=True)
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\fm2prof\fm2prof_runner.py", line 146, in run
   success = self._run_inifile()
            ^^^^^^^^^^^^^^^^^^^
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\fm2prof\fm2prof_runner.py", line 200, in _run_inifile
  self._initialise_fm2prof()
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\fm2prof\fm2prof_runner.py", line 246, in _initialise_fm2prof
  self._set_fm_model_data(
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\fm2prof\fm2prof_runner.py", line 449, in _set_fm_model_data
  time_independent_data, edge_data = nearest_neighbour.classify_with_regions(
                                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\fm2prof\nearest_neighbour.py", line 47, in classify_with_regions
  css_2d_nodes = neigh.predict(np.array([x_2d_node, y_2d_node]).T)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\sklearn\neighbors\_classification.py", line 274, in predict
   neigh_ind = self.kneighbors(X, return_distance=False)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\sklearn\neighbors\_base.py", line 838, in kneighbors
   X = validate_data(
      ^^^^^^^^^^^^^^
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\sklearn\utils\validation.py", line 2954, in validate_data
   out = check_array(X, input_name="X", **check_params)
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 File "C:\Projecten\Hyridem\OVD\venv\Lib\site-packages\sklearn\utils\validation.py", line 1128, in check_array
   raise ValueError(
ValueError: Found array with 0 sample(s) (shape=(0, 2)) while a minimum of 1 is required by KNeighborsClassifier.

To Reproduce
Bug reproduced with 2.4.0

Expected behavior
Clear error message instead of exception

[kdberends]

Error caused by null values in the region cache. Null values probably exist because region polygon file is incorrect: it does not have the "region attribute".

Issue: during import, polygon is not validated on having this attribute.
Solution: on import, validate that each polygon has the region attribute.

[kdberends]

Related: also fix section polygon validation:

Zal zo dan even bezig gaan met het aanpassen van de region polygon. Voor de section polygon had ik op basis van een foutmelding al eerder wel wat aangepast (namelijk dat de attribute SECTION een tekst string moet zijn, wat hiervoor niet nodig was, maar dat kon ik op basis van de foutmelding en wat googlen wel oplossen), maar zou inderdaad handig zijn als dit een goeie foutmelding krijgt.