ENH: Improve error message when rolling over NaT value (pandas-dev#46087)

Author: simonjayhawkins

doc/source/whatsnew/v1.4.2.rst

@@ -32,6 +32,7 @@ Bug fixes
 - Fix some cases for subclasses that define their ``_constructor`` properties as general callables (:issue:`46018`)
 - Fixed "longtable" formatting in :meth:`.Styler.to_latex` when ``column_format`` is given in extended format (:issue:`46037`)
 - Fixed incorrect rendering in :meth:`.Styler.format` with ``hyperlinks="html"`` when the url contains a colon or other special characters (:issue:`46389`)
+- Improved error message in :class:`~pandas.core.window.Rolling` when ``window`` is a frequency and ``NaT`` is in the rolling axis (:issue:`46087`)
 
 .. ---------------------------------------------------------------------------
 

pandas/core/window/rolling.py

@@ -837,12 +837,6 @@ def _gotitem(self, key, ndim, subset=None):
             subset = self.obj.set_index(self._on)
         return super()._gotitem(key, ndim, subset=subset)
 
-    def _validate_monotonic(self):
-        """
-        Validate that "on" is monotonic; already validated at a higher level.
-        """
-        pass
-
 
 class Window(BaseWindow):
     """
@@ -1687,7 +1681,7 @@ def _validate(self):
             or isinstance(self._on, (DatetimeIndex, TimedeltaIndex, PeriodIndex))
         ) and isinstance(self.window, (str, BaseOffset, timedelta)):
 
-            self._validate_monotonic()
+            self._validate_datetimelike_monotonic()
 
             # this will raise ValueError on non-fixed freqs
             try:
@@ -1712,18 +1706,24 @@ def _validate(self):
         elif not is_integer(self.window) or self.window < 0:
             raise ValueError("window must be an integer 0 or greater")
 
-    def _validate_monotonic(self):
+    def _validate_datetimelike_monotonic(self):
         """
-        Validate monotonic (increasing or decreasing).
+        Validate self._on is monotonic (increasing or decreasing) and has
+        no NaT values for frequency windows.
         """
+        if self._on.hasnans:
+            self._raise_monotonic_error("values must not have NaT")
         if not (self._on.is_monotonic_increasing or self._on.is_monotonic_decreasing):
-            self._raise_monotonic_error()
+            self._raise_monotonic_error("values must be monotonic")
 
-    def _raise_monotonic_error(self):
-        formatted = self.on
-        if self.on is None:
-            formatted = "index"
-        raise ValueError(f"{formatted} must be monotonic")
+    def _raise_monotonic_error(self, msg: str):
+        on = self.on
+        if on is None:
+            if self.axis == 0:
+                on = "index"
+            else:
+                on = "column"
+        raise ValueError(f"{on} {msg}")
 
     @doc(
         _shared_docs["aggregate"],
@@ -2630,13 +2630,3 @@ def _get_window_indexer(self) -> GroupbyIndexer:
             indexer_kwargs=indexer_kwargs,
         )
         return window_indexer
-
-    def _validate_monotonic(self):
-        """
-        Validate that on is monotonic;
-        """
-        if (
-            not (self._on.is_monotonic_increasing or self._on.is_monotonic_decreasing)
-            or self._on.hasnans
-        ):
-            self._raise_monotonic_error()

pandas/tests/window/test_groupby.py

@@ -651,7 +651,7 @@ def test_groupby_rolling_nans_in_index(self, rollings, key):
         )
         if key == "index":
             df = df.set_index("a")
-        with pytest.raises(ValueError, match=f"{key} must be monotonic"):
+        with pytest.raises(ValueError, match=f"{key} values must not have NaT"):
             df.groupby("c").rolling("60min", **rollings)
 
     @pytest.mark.parametrize("group_keys", [True, False])

pandas/tests/window/test_timeseries_window.py

@@ -5,6 +5,7 @@
     DataFrame,
     Index,
     MultiIndex,
+    NaT,
     Series,
     Timestamp,
     date_range,
@@ -134,7 +135,7 @@ def test_non_monotonic_on(self):
 
         assert not df.index.is_monotonic
 
-        msg = "index must be monotonic"
+        msg = "index values must be monotonic"
         with pytest.raises(ValueError, match=msg):
             df.rolling("2s").sum()
 
@@ -757,3 +758,12 @@ def test_rolling_on_multi_index_level(self):
             {"column": [0.0, 1.0, 3.0, 6.0, 10.0, 15.0]}, index=df.index
         )
         tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("msg, axis", [["column", 1], ["index", 0]])
+def test_nat_axis_error(msg, axis):
+    idx = [Timestamp("2020"), NaT]
+    kwargs = {"columns" if axis == 1 else "index": idx}
+    df = DataFrame(np.eye(2), **kwargs)
+    with pytest.raises(ValueError, match=f"{msg} values must not have NaT"):
+        df.rolling("D", axis=axis).mean()