default tags in BaseForecaster; added some new tags (#1013)

Adds some default tag values in `BaseForecaster` - these are automatically inherited via `_all_tags`.

Also adds two new tags that I anticipate we will need:
* `"X-y-must-have-same-index"` - do `X` and `y` need to have the same index?
* `"enforce-index-type"` argument to `check_X`/`check_y` if needed - this has to be a tag now, see the problem in #1005 .

I've also added these new tags to the extension template.

This PR also:
* corrects a minor bug: mis-spelled tags in the extension template
* adds the new tags to the allowed tags in tests
* removes the requirement for tags to be boolean

extension_templates/forecasting.py

@@ -61,12 +61,15 @@ class MyForecaster(BaseForecaster):
     """
 
     # todo: fill out estimator tags here
+    #  delete the tags that you *didn't* change - these defaults are inherited
     _tags = {
-        "fh_in_fit": True,  # is the forecasting horizon already required in fit?
-        "handles_missing_data": False,  # can the estimator handle missing data?
-        "univariate_only": True,  # can the estimator deal with multivariate series?
+        "requires-fh-in-fit": True,  # is forecasting horizon already required in fit?
+        "handles-missing-data": False,  # can estimator handle missing data?
+        "univariate-only": True,  # can estimator deal with multivariate series y?
+        "X-y-must-have-same-index": True,  # can estimator handle different X/y index?
+        "enforce-index-type": None,  # index type that needs to be enforced in X/y
     }
-    # in case of inheritance, concrete class should set all tags
+    # in case of inheritance, concrete class should typically set tags
     #  alternatively, descendants can set tags in __init__ (avoid this if possible)
 
     # todo: add any hyper-parameters and components to constructor

sktime/forecasting/base/_base.py

@@ -58,6 +58,15 @@ class BaseForecaster(BaseEstimator):
     forecasters.
     """
 
+    # default tag values - these typically make the "safest" assumption
+    _tags = {
+        "requires-fh-in-fit": True,  # is forecasting horizon already required in fit?
+        "handles-missing-data": False,  # can estimator handle missing data?
+        "univariate-only": True,  # can estimator deal with multivariate series y?
+        "X-y-must-have-same-index": True,  # can estimator handle different X/y index?
+        "enforce-index-type": None,  # index type that needs to be enforced in X/y
+    }
+
     def __init__(self):
         self._is_fitted = False
 

sktime/tests/_config.py

@@ -287,6 +287,8 @@
     "handles-missing-data",
     "skip-inverse-transform",
     "requires-fh-in-fit",
+    "X-y-must-have-same-index",
+    "enforce-index-type",
 )
 
 # These methods should not change the state of the estimator, that is, they should

sktime/utils/_testing/estimator_checks.py

@@ -129,12 +129,7 @@ def check_estimator_tags(Estimator):
     assert hasattr(Estimator, "_all_tags")
     all_tags = Estimator._all_tags()
     assert isinstance(all_tags, dict)
-    assert all(
-        [
-            isinstance(key, str) and isinstance(value, bool)
-            for key, value in all_tags.items()
-        ]
-    )
+    assert all([isinstance(key, str) for key, _ in all_tags.items()])
 
     if hasattr(Estimator, "_tags"):
         tags = Estimator._tags