Add AttributeError as suspect for dependency issue (openml#1121)

* Add AttributeError as suspect for dependency issue

Happens for example when loading a 1.3 dataframe with a 1.0 pandas.

Author: PGijsbers

openml/datasets/dataset.py

@@ -544,15 +544,23 @@ def _load_data(self):
                     data, categorical, attribute_names = pickle.load(fh)
         except FileNotFoundError:
             raise ValueError(f"Cannot find file for dataset {self.name} at location '{fpath}'.")
-        except (EOFError, ModuleNotFoundError, ValueError) as e:
+        except (EOFError, ModuleNotFoundError, ValueError, AttributeError) as e:
             error_message = e.message if hasattr(e, "message") else e.args[0]
             hint = ""
 
             if isinstance(e, EOFError):
                 readable_error = "Detected a corrupt cache file"
-            elif isinstance(e, ModuleNotFoundError):
+            elif isinstance(e, (ModuleNotFoundError, AttributeError)):
                 readable_error = "Detected likely dependency issues"
-                hint = "This is most likely due to https://github.com/openml/openml-python/issues/918. "  # noqa: 501
+                hint = (
+                    "This can happen if the cache was constructed with a different pandas version "
+                    "than the one that is used to load the data. See also "
+                )
+                if isinstance(e, ModuleNotFoundError):
+                    hint += "https://github.com/openml/openml-python/issues/918. "
+                elif isinstance(e, AttributeError):
+                    hint += "https://github.com/openml/openml-python/pull/1121. "
+
             elif isinstance(e, ValueError) and "unsupported pickle protocol" in e.args[0]:
                 readable_error = "Encountered unsupported pickle protocol"
             else: