[BUG] fix faulty BaseObject.__eq__ and deep_equals if an attribut…

…e or nested structure contains `float` (#4109)

This fixes `deep_equals` logic and reverse dependency `BaseObject.__eq__` for the case that parameters or nested objects contained an `float`.

Incorrectly, any two floats were considered equal. This is a consequence of erroneous logic to cover the case of `np.nan`.

For more context, also see sktime/skbase#111

Tests were also added to ensure that the `float` and `np.nan` case are both covered correctly.

sktime/utils/_testing/deep_equals.py

@@ -120,10 +120,8 @@ def ret(is_equal, msg):
         return ret(*_tuple_equals(x, y, return_msg=True))
     elif isinstance(x, dict):
         return ret(*_dict_equals(x, y, return_msg=True))
-    elif isinstance(x, type(np.nan)):
-        return ret(
-            isinstance(y, type(np.nan)), f"type(x)={type(x)} != type(y)={type(y)}"
-        )
+    elif _is_np_nan(x):
+        return ret(_is_np_nan(y), f"type(x)={type(x)} != type(y)={type(y)}")
     elif isclass(x):
         return ret(x == y, f".class, x={x.__name__} != y={y.__name__}")
     elif type(x).__name__ == "ForecastingHorizon":
@@ -139,6 +137,11 @@ def ret(is_equal, msg):
     return ret(True, "")
 
 
+def _is_np_nan(x):
+
+    return isinstance(x, float) and np.isnan(x)
+
+
 def _tuple_equals(x, y, return_msg=False):
     """Test two tuples or lists for equality.
 

sktime/utils/_testing/tests/test_deep_equals.py

@@ -17,6 +17,9 @@
     [([([([()])])])],
     np.array([2, 3, 4]),
     np.array([2, 4, 5]),
+    3.5,
+    4.2,
+    np.nan,
     pd.DataFrame({"a": [4, 2]}),
     pd.DataFrame({"a": [4, 3]}),
     (np.array([1, 2, 4]), [pd.DataFrame({"a": [4, 2]})]),